April 02, 2020

Essential services: 3000-MPE/iX computing

Is You Trip Necessary
HP 3000 and MPE customers have long felt they were unique. It may have been a feeling that flowed from HP's special treatment of the 3000. The server that earned HP's place at the business computing table was under-served during its final years. That felt special in a troubling way. 

Now, with the COVID-19 crisis changing the world, the 3000 and MPE have a confirmed position. These servers, built for legacy datacenters, are essential services. You can look it up at a US government website.

The 15 pages of the Department of Homeland Security advisory memorandum on "Identification of essential critical infrastructure workers during COVID-19 response" includes an extensive section on Information Technology.

The document comes from the DHS Cybersecurity and Infrastructure Security Agency (CISA). It's got an Essential Critical Infrastructure Workforce advisory list. Other federal agencies, state and local governments, as well as the private sector all advise the CISA about which jobs are essential.

One paragraph covers nearly everyone who works in IT.

"Suppliers, designers, transporters and other workers support the manufacture, distribution and provision and construction of essential global, national and local infrastructure for computing services. This includes cloud computing services and telework capabilities, business infrastructure, financial transactions/services, web-based services, and critical manufacturing."

In another spot are also "datacenter operators, including system administrators. IT managers and purchasers." There are "engineers for data transfer solutions, software, and hardware." Don't forget "database administrators for all industries (including financial services)."

If you're on the road to work toward these jobs, be certain your driving is essential. However, the HP 3000 and MPE/iX are even more essential. Servers running MPE/iX usually host historical data and track sales and inventories. Manufacturing is managed by legacy. That's infrastructure. 

And the HP 3000 and MPE/iX are more essential now because they require fewer resources. Legacy computing has already proven itself and had its bugs ironed out. It needs fewer IT staff hours. For any MPE/iX system that can be moved to a virtual instance, using Stromasys Charon, the footprint can be even lighter. Newer Intel servers and blades demand less power and take up less space.

Long ago, HP 3000 advocate Wirt Atmar called the server a peaceful device. "It creates invoices, tracks receivables, records contracts," he said. "When those things are in the air, being exchanged, we stay away from war." Wirt started his IT career developing government calculations for nuclear attack throw-weights, plus estimating projected casualties.

It's a war-like feeling in our world as we battle our way back to health. Most IT datacenter work doesn't have to be conducted on the site of the computers. But there are times when travel to a physical location is required. You can feel certain that trip is necessary.

Posted by Ron Seybold at 12:58 PM in Homesteading, Newsmakers | Permalink | Comments (0)

Pivital Solutions: Your complete
HP 3000 resource

March 31, 2020

MPE file equations and Unix equivalents

Blackboard equation
HP 3000s, as well as MPE, employ a unique tool to define the attributes of a file. That tool is file equations, a 3000 speciality. Robelle calls these "commands that redefine the attributes of a file, including perhaps the actual filename."

In any migration away from HP 3000s (an ill-advised move at the moment, considering the COVID-19 Crisis) managers must ensure they don't lose functionality. Unix doesn't have file equations. Customers need to learn how to make Unix's symbolic links report the information that 3000s deliver from a LISTEQ command.

3000 managers are used to checking file equations when something mysterious happens with an MPE file. Dave Oksner of 3000 application vendor Computer And Software Enterprises (CASE) offered the Unix find command as a substitute for file equations. You need to tell find to only process files of type "symbolic link."

Oksner's example of substituting find for LISTEQ:

find /tmp/ -type l -exec ls -l {} \;

which would start from the /tmp directory, look for symbolic links, and execute “ls -l” on the filenames it finds. You could, of course, eliminate the last part if you only wanted to know what the filenames were and get

find /tmp/ -type l

(I believe it’s the same as using ‘-print’ instead of ‘-exec [command]’)

Beware of output to stderr (if you don’t have permission to read a directory, you’ll get errors) getting interspersed.

Jeff Vance added that the command interpreter in MPE also can deliver file information through a listfile command:

Don't forget the CI where you can do:

:listfile @,2;seleq=[object=symlink]

:help listfile all shows other options.

Our former Inside COBOL columnist and product reviewer Shawn Gordon offers his own MPE vs. Unix paper, and Robelle's experts wrote a NewsWire column contrasting Unix shell scripts with MPE tools.

Image by Gerd Altmann from Pixabay

Posted by Ron Seybold at 06:46 PM in Hidden Value, Homesteading, Migration | Permalink | Comments (0)

March 28, 2020

TE turnoff date for 3000s could be 2021

Light switch
For a long time, TE Connectivity has been one of the biggest users of MANMAN software running on HP 3000s. That might be on the way to a permanent quieting — sometime next year.

Terry Simpkins at the firm was patching a Series 918 not long ago. It gave him a chance to check in with the remaining 371 subscribers to the 3000 mailing list, once he got the know-how he needed to patch.

"The force is once again calm," he said. Simpkins found a converter that allowed him to replace his old single-ended 2 Gb disk drives with the newer 36 Gb LVD drives. "I now have more disk space than my little 918 will ever need, plus a few spare drives to ensure I’ll never have a disk fail. Now to dust off the old CSL tapes and see what I want to restore."

TE measures its 3000 footprint by the number of databases online. "We’re down to four active MANMAN databases, from a high of 22. Three will convert to SAP at the end of June, so the last five-plus months will be a single MANMAN DB in Germany. I suspect we are going to be extremely bored at that point."

As to the shutdown date of 3000 operations, Simpkins said, "Right now that looks like somewhere between November 2020 and March 2021." 

Photo by twinsfisch on Unsplash

Posted by Ron Seybold at 11:47 AM in Homesteading, Migration | Permalink | Comments (0)

March 25, 2020

Preparing for a new kind of disaster

Paris coffin
A virus that kills in record numbers is circling the globe. Some who are felled by the disease might be 70 or older. That's the same age range as several HP 3000 experts who still support MPE computing. It might be a disaster if there's no one left at a company who knows details about a 3000 that's still running.

Avoiding that full stop is the business of disaster recovery. Add global fatalities to this list from 3000 consultant Paul Edwards's 2004 disaster recovery white paper. Sixteen years ago he ticked off a big list. "The top ten types of disasters, which have caused the most damage in recent years, are power outage, storm damage, flood, hardware error/failure, bombing, hurricane, fire, software error, power surge/spike, and earthquake."

The full paper remains on Edwards' website at Paul Edwards & Associates. Although, as Edwards notes, the paper doesn't go into the details of writing a disaster recovery plan, it discusses the main points to consider. Another Edwards paper, Homesteading: Plan for the Future, details what should be in a good Systems Manager Notebook. 

Every site should have one because it contains critical hardcopy information to back up the information contained in the system. It is part of the Disaster Recovery Plan that should be in place and is used to manually recreate your environment. You can’t have too much information inside it.

Twenty years ago this month, the 3000 community was already experienced at recovery from the disaster of losing a key staffer.

At that time, a 3000 manager read about a Florida site "where the system manager passed away without much notice. It sounds like documentation is pretty important in that kind of crisis. What do you recommend as a minimum?"

Paul Edwards replied:

The contents of a System Manager Notebook include hardware and software information that is vital to recovering your system in any type of disaster. The rest of the company’s business operating procedures must be combined with the IS plan to form a comprehensive corporate disaster recovery contingency plan.

The Notebook contains hardware model and serial numbers; license agreements for all software and hardware; a copy of all current maintenance agreements, equipment warranty information, complete applications documentation of program logic; data file layouts and system interaction, along with system operator run books and any other appropriate documentation. There is a wealth of information contained in each HP 3000 that can be printed and stored offsite that is critical to a recovery effort.

Image by Hans Rohmann from Pixabay

Posted by Ron Seybold at 10:40 AM in Homesteading | Permalink | Comments (0)

March 20, 2020

Making CI variables more Unix-like

By John Burke

Newswire Classic

How can you make CI variables behave more Unix-like?

For those of us who grew up on plain old MPE, CI variables were a godsend. We were so caught up in the excitement of what we could do with CI variables and command files, it took most of us a while to realize the inadequacy of the implementation. For those coming to MPE/iX from a Unix perspective, CI variables seem woefully inadequate. There were two separate questions from people with such a Unix perspective that highlighted different “problems” with the implementation of CI variables.

But first, how do CI variables work in MPE/iX? Tom Emerson gave a good, concise explanation.

“SETVAR is the MPE/iX command for setting a job/session (local) variable. I use ‘local’ somewhat loosely here because these variables are ‘global’ to your entire job or session and, by extension, are automatically available to any sub-processes within your process tree. There are some more-or-less ‘global’ variables, better known as SYSTEM variables, such as HPSUSAN, HPCPU, etc.”

The first questioner was looking for something like user-definable system variables that could be used to pass information among separate jobs/sessions. Unfortunately, no such animal exists. At least not yet, and probably not for some time if ever.

There is, however, a workaround in the form of UDCs created by John Krussel that implement system, account and user-level variables. The UDCs make use of the hierarchical file system (HFS) to create and maintain “variables.”

The second questioner was looking for something comparable to shell variables which are not automatically available at all levels. You have to export shell variables for them to be available at lower levels. Thus, there is a certain locality to shell variables.

It was at this point that Jeff Vance, HP’s principal CI Architect at CSY, noted that he had worked on a project to provide both true system-wide and local CI variables (in fact, the design was complete and some coding was done). Jeff offered a suggestion for achieving locality.

Variable names can be created using CI depth, PIN, etc. to try to create uniqueness. E.g.,

setvar foo!hppin value

setvar foo!hpcidepth value1

Mark Bixby noted that CI variables are always job/session in scope, while shell variables are local, but inherited by children if the variables have been exported. He suggested that if, working in the CI, some level of locality could be achieved by “making your CI script use unique variable names. If I’m writing a CI script called FOO, all of my variable references will start with FOO also, i.e.

SETVAR FOO_XX “temp value”

SETVAR FOO_YY “another value”

...

DELETEVAR [email protected]

“That way FOO’s variables won’t conflict with any variables in any parent scripts.”

HP has a formally documented recommendation for creating “local-ness.” 

MPE: How to create CI variables with local (command file) scope

Problem Description: I have separate command files that use the same variable names in them. If one of the command files calls the other, then they both affect the same global variable with undesirable results. Is there the concept of a CI variable with its scope local to the command file?

Solution: No. All user-defined CI variables have global (JOB/SESSION) scope. Some HP Defined CI variables (HPFILE, HPPIN, HPUSERCMDEPTH) return a different value depending on the context within the JOB/SESSION when they are called.

HPFILE returns the fully qualified filename of the command file.

HPPIN returns the PIN of the calling process.

HPUSERCMDEPTH returns the command file call nesting level.

To get the effect of local scope using global variables, you need a naming convention to prevent name collisions. There are several cases to consider.

Command file CMDA calls CMDB, both using varname VAR1.

• Use a hardcode prefix in each command file.

In CMDA use: SETVAR CMDA_VAR1 1

In CMDB use: SETVAR CMDB_VAR1 2

• Use HPFILE.

SETVAR ![FINFO(HPFILE,”FNAME”)]_VAR1 1

• Use HPUSERCMDEPTH.

SETVAR D!”HPUSERCMDEPTH”_VAR1 1 (Note: need a leading non digit)

Command file CMDA calls itself, uses varname VAR1.

• Same answer as case 1, the third solution: use HPUSERCMDEPTH.

There are two son processes. Each one calls CMDA which calls CMDB at the same nesting level.

• Same answer as case 1, the third solution: use HPUSERCMDEPTH. Not sure if this will work since not sure if HPUSERCMDEPTH is reset at JSMAIN, CI, or user process level.

• Use HPPIN and HPUSERCMDEPTH.

SETVAR P!”HPPIN”_!”HPUSERCMDEPTH”_VAR1 1

•Use HPPIN, HPUSERCMDEPTH and HPFILE (guaranteed unique, hard to read)

SETVAR P!”HPPIN”_!”HPUSERCMDEPTH”_![FINFO(HPFILE,”FN AME”)]_![FINFO(HPFILE,

“GROUP “)]_![FINFO(HPFILE,”ACCT”)]_VAR1 1

Again, there is no true local scope, only global scope for CI variables within any one session/job. The techniques presented above do provide at least a reasonable workaround for both system-wide and process-local variables.

Posted by Ron Seybold at 07:53 PM in Hidden Value, Homesteading | Permalink | Comments (0)

March 16, 2020

Deep pockets? Maybe not for MPE positions

Pants pocket
Even in the earliest days of 2020, consultants and programmers are hunting down chances to earn money servicing the 3000. When Doug Hagy looked into joining the LinkedIn HP 3000 Community, he wanted to see if the group was a source of related work opportunities. "I developed on the HP 3000 continuously from 1981 to 1999," he said. "At its peak of popularity, it was a pretty solid platform. Companies who chose HP 3000 usually had deep pockets."

Hagy isn't wrong altogether. An HP 3000 investment can be traced to Fortune 100 corporations like Boeing, or a part of the L'Oreal beauty empire. It's far more likely to see an MPE/iX server running as a place like a Texas title insurance company, or a manufacturer of saw blades.

We had to reply that we didn't know of new work opportunities for 3000 experts. Certainly, his 18 years of development experience qualifies Hagy as one of those. From time to time, opportunities surface in places like the 3000-L mailing list. Fresche Legacy has a stable of 3000 people who help in migrations as well as perform some system maintenance for 3000s.

FM Global was looking to hire a Powerhouse developer on a contract basis in January. Pay was $50 an hour for the job in Rhode Island, on a six-month set of contracts "extended for years to come." The company even advised applicants of a $62 a night rate it had for contractors' lodging at Extended Stay America. It didn't look like the lodging was fully compensated, but there was a $23 a day per diem.

Just this week we heard from Birket Foster, whose MB Foster firm is still assisting in migrations of data from 3000s. Some of those are contract-driven, while others can be sometimes project-specific engagements.

LinkedIn has a 3000 Community with 674 members, which makes it twice the size of today's 3000-L membership list. LinkedIn's groups used to have an attached Jobs feature, but by now the Jobs are spread across all of the site's resources. With that said, a post for Programmer/Analyst for the City of Lawton, Oklahoma was listed in November. MPE/iX was among the requirements.

Hagy was enthusiastic. "If I found someone wishing to migrate an app from a 3000, that could be interesting. IMAGE, V3000, COBOL could be an interesting project. Lots was developed for this platform."

The deep pockets are mostly gone from 3000 enterprises. Migrating Image, VPlus, and COBOL II was a project for the previous decade. Companies are migrating data for use with cloud computing and alternatives involving Linux. Archival 3000 systems are running, and some others are managing production on a timeframe to allow companies to migrate.

Hagy, who operates Twin Lakes Consulting, "a nimble micro business" based in Greensburg PA, says he last touched MPE around 1999 or so. "I was doing Y2K prep and providing backfill support for businesses moving to new platforms," he said. I wasn't thinking there'd be any HP 3000 action out there now, until I saw the group you moderate. I'd make time to assist if someone wanted to port a novel app from MPE and needed someone to dissect its inner workings. Before MPE I did RTE work on HP 1000s."

Some companies will need a programmer or consultant whose experience goes back to the days of real-time systems with HP badges on the front of the server. They emerge from the shadows of an era where reliable service, on an unhackable server, that simply worked, could be enough.

Image by ds_30 from Pixabay

Posted by Ron Seybold at 03:51 PM in Homesteading, Migration | Permalink | Comments (0)

March 12, 2020

3000-L newsgroup heads for new future

Jeff Kell shutdown
IT staff at the University of Tennessee at Chattanooga in 2014, switching off their last HP 3000

The manager of the university datacenter that hosts the 3000-L mailing list and newsgroup has told members the list will be moved in some way during the months to come. Without sharing a timeline for the changes, technical director Greg Jackson of the University of Tennessee at Chattanooga (UTC) Information Technology said, "UTC will stop support of the list in its present form, as we move to a different delivery method."

"In the fall [of 2019]," Jackson said, "I contacted the moderator of the hp3000 list and let them know that at some undetermined point in the future, UTC will stop supporting the listserv in its present form as we move to a different delivery method. Since this list is still active, when the time comes, we will work with the group to ensure a smooth transition."

News of the movement and rehosting of the biggest archive of 3000 community messaging surfaced after UK users couldn't access the archives. Robert Mills said that when he contacted Jackson about being locked out of the archives, he was told "Over the past few months, there have been several attacks on the listserv that originated from IP addresses outside the US. Therefore, we decided to restrict the ListServ to the US only."

The list's membership count stands at 371. Donna Hofmeister, a support engineer at Allegro and one of the list's moderators, said the community should decide now how the message service and 3000 knowledge archive can move forward.

"Due to the looming changes at UTC," she said in a message, "hp3000-l needs to do something." The archives of the list, which date back to 1994, will "somehow be made available for searching."

"As one of the said moderators," she added, "I think it's only appropriate to ask — do we want hp3000-l (in whatever form it might take) to continue? The amount of traffic (which is around five messages/month) makes it a question that should be asked."

Hofmeister said that "having access to the archives has real value. So my plans are to <somehow> make the archives available for searching. So what do you say? Keep HP3000-L active in some form (I'm leaning towards making it a Google Group) or let it go away when UTC takes down the listserv? In either case, the archive will be available."

Fourteen list members responded quickly to vote for keeping the list alive. Two alternatives emerged as options when the UTC hosting ends as it exists, using the LISTSERV software. Rehosting on groups.io was suggested by Tracy Johnson, while Keven Miller proposed a free version of another listserve program.

"The Lsoft Lite free version supports up to 10 lists, 500 members each," Miller said. "There are a few other lists [whose UTC] archives might be nice [to preserve]. HP9000-L, OpenMPE, and maybe a few hidden lists. I would think that Lsoft Lite would be the easiest to move archives to. But I'm sure there are other open source solutions."

Posted by Ron Seybold at 05:29 PM in Homesteading, Web Resources | Permalink | Comments (0)

March 05, 2020

Large Disk patch delivers 3000 jumbo limits

Marshmallows
As the HP 3000 was winding its way out of the HP product lineup, it gained a greater footprint. Storage capabilities grew with the rollout of Large Disk. The effort was undertaken because HP's disk module sizes were doubling in size approximately every two years: 4 GB to 9, 18, 36, 73, 146, and then 300 GB.

The disk project might have never seen its limited release without OpenMPE. The advocacy group that was formed after HP's exit announcement saw the same disk size trend. OpenMPE drove the initiative of "Support future large disks" in the Interex 2003 Systems Improvement Ballot.

Just two years later, Interex was dead. The directive from the 3000 community to HP labs lived onward, though. HP said its investigation found the need for more work to be done to support large disk configurations.

The MPE/iX 6.5 Large File enhancement allowed bigger Files. 6.5 also permitted more disk space in each MPE Group and Account. But several CI commands and utilities were limited in their ability to work with the resulting larger Groups and Accounts. All of these inputs were assessed during the Large Disk investigation and as many as possible were addressed by the Large Disk patches.

So what does Large Disk deliver? The patches provide the following enhancements for MPE/iX 7.5:

• Large Disk includes the ability to attach and use SYSGEN to configure any sized SCSI-2 compliant Disk. MPE/iX uses SCSI-2 protocol to connect to SE, HVD and LVD SCSI Disks as well as Fibre Channel over SCSI. The SCSI-2 standard allows for disks of up to 2 Terabytes. SCSI-3 disks may be larger but will only report up to 2 Terabytes of storage for SCSI-2 format inquiries.

• Large Disk includes the ability to initialize an MPE/iX Disk Volume of up to 512 Gigabytes on SCSI-2 compliant disks. SCSI-2 Disks that are larger than 512 GB are truncated at the 512 GB limit. No matter how big the disk, HP reported, the space beyond 512 GB will not be usable by the MPE/iX or any applications.

There are limits to how much Large Disk is available. And MPE/iX disk volume includes disk-resident OS data structures that use some disk space, so no more than 511 GB of user file space should be expected.

• Large Disk includes a number of opportunistic enhancements to MPE Command Interpreter commands and utility programs to 'smooth' user experience when dealing with large disks, large groups and large accounts. These commands and utilities are REPORT, :[ALT|LIST|NEW][GROUP|ACCT], FSCHECK, and DISCFREE.

HP strongly advises installing all of these patches at the same time using Patch/iX. The Large Disk Patches are:

• MPEMXX8(A) -- FSCHECK.MPEXL.TELESUP
• MPEMXU3(B) -- [ALT|LIST|NEW][ACCT|GROUP]
• MPEMXT3 -- SCSI Disk Driver Update
• MPEMXT4 -- SSM Optimization (>87 GB)
• MPEMXT7 -- DISCFREE.PUB.SYS
• MPEMXU3 -- REPORT
• MPEMXV2(A) -- CATALOG.PUB.SYS
• MPEMXW9(A) CIERR.PUB.SYS, CICATERR.PUB.SYS

Image by pixel1 from Pixabay

Posted by Ron Seybold at 08:24 PM in Hidden Value, Homesteading | Permalink | Comments (0)

February 25, 2020

Wayback: 3000s showed a Spectrum of hope

BeyondRisc
Thirty-six years ago this month, HP put a reboot of its business future into orbit. The project called Spectrum was the entry of PA-RISC (originally called "HP High Precision Architecture") when publicly announced in the HP Journal in 1985. HP brought the future into the light by killing its Vision project at the 1984 Interex user conference.

Stan Sieler, one of the founders of Allegro, was working at HP in the years before the HP announcement of what the company called High Performance Precision Architecture RISC. "A year or so later, when it was simply called PA-RISC (or HP PA-RISC), I asked Joel Birnbaum what happened to the "High" and I was ignored. Along with Bill Worley, these were the fathers of RISC inside HP. Birnbaum had been recruited from IBM's RISC project."

Digital was famous for raining on HP's Reduced Instruction Set Computing, as well as Unix, during the time PA-RISC rose up. Ken Olsen, DEC's founder, pulled the plug in 1989 on Prism, Digital's RISC computer design. HP struggled to get its business servers onto PA-RISC, managing to put its Unix onto the new architecture first. Digital tried to make inroads by touting its 32-bit VAX processors versus the 16-bit HP 3000 classic servers. "Digital has it now," the ads in the trade weeklies proclaimed.

Sieler says that several other companies were incensed at HP having a product called Spectrum, including Chevrolet. "I remember hearing reports of some legal actions against HP, which were reportedly dropped after HP promised to never use that term externally. That is apparently why we titled our book about PA-RISC Beyond RISC instead of Beyond Spectrum. We were told HP wouldn't buy any copies if we had "Spectrum" on the cover. But we did sneak it in: the spectrum is the photo."

RISC was designed to consolidate the development of peripheral interfaces for all all three of its computer lines: HP 3000, HP 9000s, and its real-time systems the HP 1000s. About late 1986, the real time version of HP-UX on PA-RISC —  demonstrated at the 1986 Madrid Interex conference on an HP 9000/840 — was quietly dropped. "We used to have an HP publication about real time support for HP-UX, but I think it went to the Living Computer Museum in Seattle when we gave them our manuals about two years ago," Sieler said.

Posted by Ron Seybold at 01:27 PM in History, News Outta HP | Permalink | Comments (0)

February 12, 2020

Adding a naked Seagate drive to a 3000

Seagate Barracuda 31841

James Byrne reported on a way to get a Seagate disk drive to mount in a Series 918. 

We have a 918LX that we are trying to configure as a spare. The unit has three 18Gb disc drives installed, Seagate model ST31841N. We can see the drives in Mapper at 52.56.6/5/4. We can use DISCUTIL to mount 52.56.4 and 52.56.6 — but we cannot get the drive at 52.56.5 to mount.

This problem drive is a new unit just removed from its factory packaging.

Naked Seagate SCSI drives require a low level format to a sector size of 512 for the HP 3000 to mount them. We have a Windows-based tool called Seatools from Seagate that can perform this formatting from a Windows host — at least, from a host that has a suitable 50N SCSI interface card installed.

The same thing can be accomplished by doing a full install of MPE/iX from tape to such a disc. The install of MPE/iX directly to that disk which we could not mount solved the mounting problem.

Posted by Ron Seybold at 07:08 PM in Hidden Value, Homesteading | Permalink | Comments (0)

February 10, 2020

3000 market maven Charles Finley dies at 70

CharlesFinley_8_2_2013
Charles Finley, whose career in the HP 3000 community spanned eras from powerful regional user group conferences to trusted HP reseller status, then led to new success as a migration maven, has died at age 70.

Finley built a reputation with the community from his first steps in the Southern California 3000 market. Buoyed by the remarkable manufacturing community in the area, by the middle 1980s he was operating the ConAm reseller and worked to establish the Southern California Regional User Group. SCRUG hosted conferences successful enough to rival those from Interex in scope.

Finley also played an essential part in founding an invitation-only MPE developer conference, using a novel format called the un-conference. It delivered information that otherwise would not be presented if only one person were in charge of the agenda. In the early times for groundbreaking tech, the 3000 community had a forum to explore new choices. "Things that could be overlooked like NT, Linux, VMware are noticed, because one person in the group happened to notice it and think it was important," he once said. "The rest of us benefitted by it."

Once HP curtailed its 3000 futures, Finley evolved the ConAm reseller business into Transformix, owned and operated with his wife Deborah. She assumes the post of president of the firm that has created and deploys a migration suite for carrying legacy applications from MPE/iX and other environments applications onto new platforms, especially Linux.

Finley was a Vietnam-era Marine Corps veteran. His widow said the CEO of Transformix delivered his skill and innovation with a duty to the work and the customer.

"Charles was unsurpassed in his passion for the business, his drive for perfection and professionalism, and his commitment to the integrity of customer relationships," Mrs. Finley said. “I saw that every day in the way he spoke about his work."

"This is both a personal and professional loss for many of the people who have known and loved him. Everyone who knew Charles regarded him as a man devoted to his family, his employees, his customers, and his friends."

Condolences may be sent to Deborah via email. The family requests that donations in lieu of flowers can be made to the charity Charles held close: Copley-Price Family YMCA 619.280.9622. Deborah asks to please designate that your donation is in memory of Charles H. Finley, Jr.

The company he leaves in her management is an integration, reseller, and consulting organization specializing in migration of legacy systems to current hardware and software. Transformix is headquartered in San Diego.

Mrs. Finley said the passing was unexpected. Charles Finley is listed as a speaker at next month's SCALE 18X technology conference. His seminar, Transforming Legacy Applications to an Open Source Modern Technology Stack, was the latest in a line of talks at the Southern California Linux Expo.

This year's seminar would "provide attendees with an understanding of the steps involved to transform legacy applications by retargeting them to an Open Source Java Framework. The seminar shows how the CUBA-Platform framework—which was designed for development of modern web application—is also well suited to enhance and extend legacy applications."

Finley was a significant voice in the migration community. While outlining differences between legacy migration, modernization, and transformation,
his experience smoothed the way for legacy applications to use modern technology stacks, including Java.

His SCALE seminar for this year was "a hands-on workshop transforming a legacy application for those who want to know more."

"If you have a PostgreSQL database already, you can generate a working Java web application in minutes using the CUBA-Platform. Moreover, you can do this without knowing any Java! Also of interest is the fact that professional developers and 'citizen developers' can use the platform for development."

Posted by Ron Seybold at 04:41 PM in History, Homesteading, Migration, Newsmakers | Permalink | Comments (0)

February 04, 2020

Chicken, egg: First the 3000's OS, then chips

Rooster
Editor's Note: A technical paper from the DEC world asserted that VMS was the first operating system designed before the chipset that it ran upon. MPE's earliest designs were just as innovative. We asked Stan Sieler for some history.

By Stan Sieler

I'd assume that the 16-bit Classic instruction set architecture and the original MPE were designed at about the same time — probably with the architecture being mostly ready/running (real or simulated/emulated) before the software was ready. Once MPE was up and running, some years later there were arguably one to three architectures designed for it (exclusively or not).    

FOCUS

A group of about 12 of us (labs, chip people, me for the OS lab) designed a 32-bit architecture for the next generation HP 3000.        

The architecture was an evolution of an earlier FOCUS used by Ft. Collins for some HP 9000s (after the 68000 models, before the PA-RISC models), and it (the earlier) was either used by the Amigo (HP 300) and/or was inspired by the Classic 3000 architecture. The project got dropped in favor of the VISION architecture.

VISION        

This was the object-oriented architecture (with 64-bit virtual addressing) that was going to be the next-gen HP 3000, running what was going to be called HPE. We had HP 3000/4x computers with rewritten firmware emulating it, and there were a couple of hand-made real CPU boards beginning to run when I left HP in September 1983 to start Allegro. 

At that time, I had a crude command interpreter running on it under my process management code (I was in charge of process management). VISION was very very interesting.  If I had access rights to an object (say, a record from an IMAGE database with an employee name, a date-of-hire, and other information), I could send another process a "descriptor" (virtual address) that would give them access to precisely the subset I wished (e.g., read access to date-of-hire field of the record). But, that concept is gone now.  No one can do that :(        

VISION was dropped in favor of PA-RISC about a month or so after I left. I commented to Joel Birnbaum that it was dropped because I'd left HP. His reply was, "If I knew that, I'd have gotten rid of you sooner."    

About 1982-1983 I began to hear about an architecture that HP Labs was working on that would allow you to run MPE, RTE, and maybe even HP-UX simultaneously.  It was code-named "Rainbow." I think Rainbow turned out to be PA-RISC.

PA-RISC

In the 1980s, RISC (Reduced Instruction Set Computer) was all the rage. People thought it meant quicker execution due to less complex instructions. I am still dubious. I think they underestimated the types of computations and instruction/memory interactions needed — and, indeed, you can see people throwing more and more and more cache towards RISC in an effort to address the speed imbalance between the CPU and memory.

HPE, essentially an extended version of MPE, was designed to run on PA-RISC. To the extent that the virtual memory (and IO) was quite different, that part of the OS was designed for the architecture.

Most of HPE, later MPE XL, then MPE/iX, doesn't care what the architecture is, any more than  Linux/Unix/Windows cares what the architecture is. I seem to recall that a few aspects of the memory protection mechanism (including the Protection ID registers) may have been influenced by HPE's needs.  

Of course, at the same time, HP-UX was being ported from 68000 / FOCUS to PA-RISC, so there may have been interactions there, as well. Note, however, that HP-UX never fully utilized the PA-RISC architecture — particularly the memory addressability, where HPE / MPE XL / MPE/iX had it beaten by far. I don't think HPE, HP-UX, or Netware (which was on PA-RISC briefly, circa 1993) used all the capabilities, including the ability to, in controlled circumstances, let user code directly access some IO instructions.

Itanium (IPF)  

I think I heard that a basic MPE/iX kernel of sorts had been successfully ported to Itanium before the HP 3000 was killed. Obviously, HP-UX was also ported to IPF.   The primary influence MPE/iX and HP-UX probably had was the Itanium ability to run in either little-endian mode (Intel X86 style for Windows) or big-endian (Class, and PA-RISC style, for MPE/iX and HP-UX).

Other operating systems running on Itanium — which have been released in some cases, not released in others — include Windows, Netware, Solaris, OpenVMS, and Tru64 Unix. This list of systems tends to imply that the architecture was not specifically oriented towards one particular operating system.

In short, I think most operating systems exist (perhaps in an earlier form) prior to the chip architecture, but that most architectures are mostly independent of the operating system design/features. The memory addressability mechanism almost always affects major aspects of the internals of the operating system (as it would with VISION).

Photo by Ashes Sitoula on Unsplash

Posted by Ron Seybold at 04:29 PM in History, Homesteading, News Outta HP | Permalink | Comments (0)

January 30, 2020

Calendar date issues are already surfacing

Hurdles
The 2028 date hurdle for MPE/iX has been well documented and thoroughly discussed. Although the January 1, 2028 deadline — when MPE/iX CALENDAR processes will start to report dates as January 1, 1970, and so on — seems like it's years away, it's much closer. Calendar issues emerge as programs call for dates.

Programs that call for dates in the future are already facing the hurdle. Systems that use Unix, Linux, or other operating systems this month have triggered these involuntary date rollbacks already.

In one recent case, a top 100 pension fund had a nightly batch job that computed the required contributions, made from projections 20 years into the future. It crashed on January 19, 2018 — 20 years before Y2038.

HP 3000s have been key tools in many financial and resource planning operations. While dates are usually used to track transactions as a matter of history, some ERP users look forward to forecast their resource needs.

MPE/iX has a Y2028. Unix and Linux have a Y2038. This is important to know for a legacy system manager's planning and tactics. There's no good reason to tear down a legacy system if its only show-stopping flaw is date handling. A solution for the 3000 community is already at hand in several spots. 

Stromasys reports it has been working with an independent developer for a 2028 fix, something available to its Charon emulator sites. That update was shared with us in July of last year. It's not public yet, but that indie developer confirms the work is in progress. Beechglen has a 2028 solution it is selling as a service.

There are additional developers and consultants who say they're ready to repair 2028 issues with MPE/iX systems. It's important to know that the HP 3000, as one of the older legacy systems still working in businesses, is in no worse shape than systems driven by Linux or Unix. It's only a matter of when, not if, a date handling process will need to be addressed.

The legacy of an operating system is a condition defined very broadly. Legacy systems have been successful for a long time, and the vendor's focus has usually slipped away from these legacies. It can remind us of that term "proprietary" that was hurled at the 3000 for a decade before HP quit on its futures. Nearly all technology has a proprietary aspect, even if it only amounts to a support clause that makes one knowledge resource crucial to the OS health. 

Photo by Interactive Sports on Unsplash

Posted by Ron Seybold at 01:23 PM in Homesteading, Migration | Permalink | Comments (0)

January 28, 2020

What good are Nike arrays?

HP NIke Array
By John Burke

3000 users still can employ used HP Nike Model 20 disk arrays. There was once a glut of these devices on the market — meaning they were inexpensive — and they work with older models of HP 3000s. Here's a note from one company using these Nike arrays.

"We’re upgrading from a Model 10 to a Model 20 Nike array. I’m in the middle of deciding whether to keep it in hardware RAID configuration or to switch to MPE/iX mirroring, since I can now do it on the system volume set. It wasn’t in place when the system was first bought, so we stayed with the Nike hardware RAID. We’re considering the performance issue of keeping it Nike hardware RAID versus the safety of MPE Mirroring. You can use the 2nd Fast-Wide card on the array when using MPE mirroring, but you can’t when using Model 20 hardware RAID.

So, with hardware RAID, you have to consider the single point of failure of the controller card. If we ‘split the bus’ on the array mechanism into two separate groups of drives, and then connect a separate controller to the other half of the bus, you can’t have the hardware mirrored drive on the other controller. It must be on the same path as the ‘master’ drive because MPE sees them as a single device.

Using software mirroring you can do this because both drives are independently configured in MPE. Software mirroring adds overhead to the CPU, but it’s a tradeoff you have to decide to make. We are evaluating the options, looking for the best (in our situation) combination of efficiency, performance, fault tolerance and cost.

Note: Mirrored Disk/iX does not support mirroring of the System Volume Set – never did and never will. Secondly, you most certainly can use a second FWSCSI card with a Model 20 attached to an HP 3000

All of the drives are accessible from either controller but of course via different addresses. Your installer should set the DEFAULT ownership of drives to each controller. To improve throughput, each controller should share the load. Only one controller is necessary to address all of the drives, but where MPE falls short is not having a mechanism for auto-failover of a failing controller.

In other words, SYSGEN reconfiguration would be necessary to run on a single controller after SP failure in a dual SP configuration. You could have alternate configurations stored on your system to cover both cases of a single failing controller but the best solution is to get it fixed when it breaks. The best news is that SP failures are not very common.

There is a mechanism in MPE for ‘failover’ called HAFO - High Availability FailOver. It is only supported with XP and VA arrays, and not on Nike’s or AutoRAIDs (because it does not work with those).

Andrew Popay provided some personal experience.

"We have seven Nike SP20 arrays, totaling 140 discs spread across all the arrays, using a combination of RAID 1 (for performance) and RAID 5 (for capacity). We use both SP’s on all arrays, with six arrays used over three systems (two per system). One of our systems has two arrays daisy-chained. The only failures we have suffered on any of the arrays have been due to a disc mechanism failing.

"We never find any issues with the hardware raiding; in fact, as a lot of people have mentioned, hardware raiding is much more preferred to software raiding. Software raiding has several issues, system volume, performance, ease of use, etc. Hardware raiding is far more resilient.'

As for anyone concerned about single points of failure, I would not worry too much about the Nike arrays, I would say they are almost bulletproof. For those who require a 24x7 system and can’t afford any downtime whatsoever, maybe they should consider upgrading to an N-Class, with a VA or XP. Bottom line: SP20’s are sound arrays on the HP 3000s, easy to configure, set up and maintain.

Posted by Ron Seybold at 01:58 PM in Hidden Value, Homesteading | Permalink | Comments (0)

January 23, 2020

When the HP Way Led the 3000 Astray

Winding road forest

Editor's Note: Being a legacy system expert has its frustrating days. If experts of today ever wonder why they got into the lifespan of Hewlett Packard and MPE, they can look back to the start and the promise of the 3000. Bill Foster was a part of the HP team that created the system, before he went on to found Status Computing. the story below shows the an HP which had to remake MPE.

All the Foster you want, in an HP history worthy of being a book, is at his website.

By Bill Foster

If there ever was a company that always seemed to do the right thing, it was HP back in the 70’s. We had a term called The HP Way. There was no written definition — it was something you felt. When something good happened it was part of The HP Way. When you had the inclination to do something bad — cut corners on a project, treat a customer badly, turn in an inflated expense account, fire a really bad employee — these things didn’t happen. They were not The HP Way. It’s like we walked around with little halos over our heads.

Of course, if this was the only place you worked, you assumed all companies were like HP. You had to leave Hewlett Packard to become a part of the real world. So, we shipped HP 3000 serial #1 to the Lawrence Hall of Science in nearby Berkeley. A couple of weeks later they shipped it back. That 3000 could support at most two or three users on a good day -- nowhere near 16 or 32 or whatever they promised.  And MPE was crashing three or four times a day.

A few months and a couple of machines later HP punted and withdrew the 3000 from the marketplace. They gave free 2116 computers to the customers in hopes of appeasement — The HP Way. Bill and Dave were fuming -- this had been by far the most expensive project in the company’s history, and Hewlett Packard was being inundated with bad press — something that had never happened in the entire history of the company.

In fairly quick succession Paul Ely came down to save things and a few months later my boss Steve Vallender left. I don’t think Steve was fired — HP never fired anyone back then, they just promoted them into oblivion. But Steve was somewhat un-promotable — he lacked a college degree and HP was pretty snobbish about that.

Dick Hackborn asked me if I wanted Steve's old job. Are you kidding? Sure! Hurt me! I was looking to move up the ladder — this was a fantastic break. My guess was they chose me over my hardware counterpart because management finally figured it was better to put a software guy in charge of computer projects. No matter -- here I was, not even 30 years old, running all the hardware and software development for HP's computer business.

My first and most important job was to come up with a plan for the hockey pucks.  A year earlier, Dick Hackborn had hired a couple of smooth-talking marketing bozos out of IBM. Hackborn created a group called Product Marketing within his Engineering group to compete with the real Marketing group at the other end of the building.

This was very out of character for HP — to hire senior people from the outside. One of their first actions was to give mementos of a project to the engineers who had developed it — something tangible to remember their efforts. Apparently this was done all the time at IBM. The IBM marketing bozos came up with the idea of a brass paperweight about the size of a hockey puck, but about half the thickness. Stamped on each one were three overlapping circles signifying batch, realtime, and timesharing — things that the 3000 was supposed to do.  And each individual’s name was engraved on the back.

These were supposed to be handed out months earlier, but with all the problems, Vallender had hidden them away in a file cabinet. My first command move was to sneak in one weekend, lug them out to my car, and take them home to my garage.  The last thing I wanted was for anyone to get wind of them. The next step was to try to get some kind of usefulness out of the 3000 machine, and that meant fixing MPE.

Image by David Mark from Pixabay

Posted by Ron Seybold at 05:36 PM in History, News Outta HP | Permalink | Comments (0)

January 21, 2020

Using RAID5 on an HP 3000

Hard-Drive
By Gilles Schipper
Homesteading Editor

RAID storage, including a low-cost MOD20 array, can improve a 3000's performance. Here are a few things to consider if you will be acquiring a MOD20.

Although possible, I would not recommend utilizing RAID5 LUNs in an HP 3000 environment — unless your greatest priority is to maximize disk space availability at the expense of performance.

RAID5 offers fail-safe functionality over a group of disks (minimum of three) by means of one disk of the RAID5 disk being allocated as a parity disc. The benefit of RAID5 over RAID1 is that it results in a greater amount of overall usable disk space than RAID1. However, it performs poorly in an HP 3000 environment, and cannot be booted from if specified as the system disk (LDEV 1).

Although the supported maximum memory configuration of each Storage Processor (SP) unit is 64MB, 128MB works best (although not all of it can be used).

Each SP has 4 memory slots. You can maximize the performance of the MOD20 by populating each SP with four 32MB memory SIMMs, 72-pin, FPM with parity, 60ns.

The NIKE MOD20 is a very capable and useful solution to the fragile environment afforded by a JBOD environment — particularly because most 3000 JBOD disk systems tend to be very mature and consequently relatively unreliable and prone to failure.

And, although the MOD20 disk system itself is also quite long in the tooth, it’s got built-in fail-safe mechanisms. Also, the MOD20 would appeal to those with very limited budgets, since the devices are quite inexpensive in the used-equipment market.

There are other, more advanced RAID systems that also support the HP 3000 environment. These include the HP Autoraid12H system, various VA7400 systems, some of the HP XP-family members, as well as EMC systems. This list is in order of increasing cost, for the most part.

The bottom line: if you are not already utilizing RAID technology for your 3000, now would be a good time to consider it seriously.

Posted by Ron Seybold at 01:53 PM in Hidden Value, Homesteading, Newswire Classics | Permalink | Comments (0)

January 16, 2020

MPE/iX Command File Scripts Explained

Code on screenBy Ken Robertson

The MPE/iX command interpreter has a generous command set, pushing the shell into the realm of a true programming tool. Its ability to evaluate expressions and to perform I/O on files allows the end-user to perform simple data-processing functions. The CI can be used to solve complex problems. Its code, however, is interpreted, which may cause a CI solution to execute too slowly for practical purposes.

Command files are a collection of commands in flat files, of either variable or fixed length record structure, that reside in the MPE or POSIX file space. Basically, command files are what you could call MPE Macros. Anything that you can do in the CI interactively, you can do with command files, and then some. You can use command files in situations that call for repetitive functions, such as re-compiling source code, special spooler commands, etc. Command files are also great when you want to hide details from the end-user.

A command file is executed when its name is typed in the CI, or invoked from a command file or programming shell. Just as in program execution, the user’s HPPATH variable is searched to determine the location of the command file.

MPE Scripts Versus Unix Scripts

For the average task, the MPE scripting language is easier to read and understand than most Unix scripts. For example, command line parameters in MPE have names, just like in regular programming languages.

Of course, there are several script languages on Unix and only one on MPE! On Unix you can write shell scripts for any of the many shells provided (C shell, Bourne shell, ksh, bash, etc). Although there is also now a Posix shell on MPE, most scripts are written for the CI. Several third-party tools, such as Qedit and MPEX, emulate HP scripting and integrate it with their own commands.

A command file can be as simple as a single command, such as a Showjob command with the option to only show interactive sessions (and ignore batch jobs):

:qedit
/add
1      showjob [email protected]
2      //
/keep ss
/e
:

You have created a command file called SS — when you type SS you will execute showjob [email protected]

On MPE, the user needs read (r) or execute access (x) to SS. On Unix you normally must have x access, not just r access, so you do a chmod +x on the script. This is not necessary in MPE, although, if don’t want users to be see the script, you may remove read access and enable execute access.

Structure of a Command File (aka CI script)

A script is an ASCII file with maximum 511 byte records. Unlike Unix, the records may contain an ASCII sequence number in the last 8 columns of each line. The command file consists of 3 optional parts:

1. Parameter line with a maximum of 255 arguments:
parm sessionnumber
parm filename, length=”80”

2. Option lines:
option nohelp,nobreak
option list

3. The body (i.e., the actual commands)”
showjob job=!sessionnumber
build !filename;rec=-!length,,ascii
In MPE scripts, there is no inline data, unlike Unix ‘hereis’ files.

Parameters

Notice in the example above that parameters are used with an exclamation (!), as opposed to the $ in Unix. The same is true for variables. Parameters are separated by a space, comma or semicolon. All parameter values are un-typed, regardless of quoting.

In a typical Unix script, the parameters are referenced by position only ($1, $2, $3, …). In an MPE script, the parameters have names, as in the function of a regular programming language, and can also have default values. In Unix you use [email protected] for all of the parameters as a single string; in MPE you use an ANYPARM parameter to reference the remainder of the command line (it must be the last parameter).

Here is a script to translate “subsys” and “err” numbers from MPE intrinsics into error messages. The subsys and error numbers are passed in as parameters:

parm p_subsys=108,p_error=63
setvar subsys hex(!p_subsys)
setvar error hex(!p_error)
comment the hex conversion allows for negative numbers
comment the #32765 is magic according to Stan!
setvar cmd “wl errmsg(#32765,!subsys);wl errmsg(!error,!subsys);exit”
debug !cmd

As you can see above, the Setvar command assigns a value to parameter or to a new variable. But there are also system pre-defined variables. To see them all do Showvar @;hp. To get information on variables, do help variable and to get help on a specific variable, say hpcmdtrace, do help hpcmdtrace (set TRUE for some debugging help).
In most MPE commands, you must use an explicit exclam ! to identify a variable: build !filename

However, some MPE commands expect variables, and thus do not require the explicit !. For example, Setvar, If, ElseIf, Calc, While, and for all function arguments, and inside ![expressions].

Warning: variables are “session global” in MPE. This means that if a child process, or scripts, changes a variable, it remains changed when that child process terminates. In Unix you are used to the idea that the child can do whatever it likes with its copy of the variables and not worry about any external consequences.

Of course having global variables also means that it is much easier to pass back results from a script! And this is quite common in MPE scripts.

Options

Options allow you to list the commands as they are execute (option list), disable the Break key (option nobreak), enable recursion (option recursion), and disable help about the script (option nohelp).

The script body below shows active process information. This example shows many of the commands commonly used in scripts: If, While, Pause, Setvar, Input and Run. Other commands you will see are Echo, Deletevar, Showvar, Errclear.

WHILE HPCONNSECS > 0
    IF FINFO("SQMSG",0)
       PURGE SQMSG,TEMP
    ENDIF
    BUILD SQMSG;REC=-79,,F,ASCII;TEMP;MSG
    FILE SQMSG=SQMSG,OLDTEMP
    SHOWQ;ACTIVE >*SQMSG
    SETVAR PINLIST ""
    WHILE FINFO("SQMSG",19) <> 0
         INPUT SQLINE < SQMSG
         IF POS("#",SQLINE) <> 0 THEN
           SETVAR PIN RTRIM(STR(SQLINE,47,5))
           SETVAR PINLIST "!PINLIST" + "," + "!PIN"
         ENDIF
    ENDWHILE
    IF FINFO("SPMSG",0)
       PURGE SPMSG,TEMP
    ENDIF
    BUILD SPMSG;REC=-79,,F,ASCII;TEMP;MSG
    FILE SPMSG=SPMSG,OLDTEMP
    SETVAR PROC "SHOWPROC PIN="+"!PINLIST"+";SYSTEM >*SPMSG"
    !PROC
    WHILE FINFO("SPMSG",19) <> 0
         INPUT SPLINE < SPMSG
         IF POS(":",SPLINE) <> 0 THEN
           ECHO !SPLINE
         ENDIF
    ENDWHILE
    PAUSE 30
ENDWHILE

Handling Errors

In most Unix scripts, if a step fails, you check for an error with an If-conditional and then take some action, one of which is ending the script. Without an If, the script continues on, ignoring the error.

In MPE, the default action when a step fails is to abort the script and pass back an error. To override this default, you insert a Continue command before the step that may fail. You then add If logic after the step to print an error message and perhaps Return (back 1 level) or Escape (all the way back to the CI).

     continue
      build newdata
      if cierror<>100 then
         print "unable to build newdata file"
         print !hpcierrmsg
         return
      else
         comment - duplicate file, okay
      endif

You can set HPAUTOCONT to TRUE to continue automatically in case of errors, but this can be dangerous. The default behavior at least lets you know if an unexpected problem occurs.

User Defined Commands (UDC)

UDCs are like Command File scripts, except that several are combined in a single “catalog” file. They are an older feature of MPE, so you may see them in older applications even when scripts seem like a better solution. The primary reason that they are still useful is that they support Option Logon, which invokes the command when a user logs onto the system.

More Information

Tim Ericson’s collection of UDCs and Command files has recently been resurrected and re-published in the public domain at www.3kassociates.com/index_cmd.html

Image by fancycrave1 from Pixabay

Posted by Ron Seybold at 08:42 AM in Hidden Value, Homesteading, Newswire Classics | Permalink | Comments (0)

January 14, 2020

Listserv still serving advice after 26 years

Bank vault safety deposit boxes
The 3000-L Listserv repository is the HP 3000 resource that's been in the longest continuous use for the MPE/iX ecosystem. HP had a Jazz website for about 13 years, content that was carried over to Fresche Legacy's servers once HP's labs closed. 3000-L was online for about a year or so before the NewsWire entered the Web.

The content on the 3000-L was a big reason I believed we could do a monthly HP 3000 newsletter. We curated and learned, education and advice we shared with readers. Even after 26 years, 3000-L can be searched for answers that go back to the era of MPE/iX 4.0.

That repository is full of history about the people who have created the MPE ecosystem, too. with enough patience, most answers will be hiding in the hundreds of thousands of email messages. All are logged by subject line. 3000-L can be searched within date ranges, too.

3000-L was once so robust that we could publish a column about its gems once a month as part of the first 10 years of the NewsWire. 

The columns are archived in our 1996-2005 pages. We called them NetDigest, and for awhile they were written by John Burke, who helped us found the NewsWire with his knowing voice and deep technical experience.

For the source material for those columns, refer to the 3000-L search panel.

For the columns, refer to the Tech Page of the '96 - '05 issues. Once you arrive at the Tech Page, just do a search within the page for the phrase net.digest. We've got 106 columns there.

Photo by Jason Pofahl on Unsplash 

Posted by Ron Seybold at 03:29 PM in Hidden Value, Homesteading, Web Resources | Permalink | Comments (0)

January 09, 2020

How to Encrypt 3000 Log-on Passwords

Padlock
NewsWire Classic

Is there a way to encrypt MPE logon passwords to keep auditors satisfied that the HP 3000 is secure? We need to show that they cannot be easily read with the ;pass parameter (i.e. listuser xxx.yyy;pass)

The replies generated one of the longest threads of the month on the 3000-L.

Tracy Johnson offered an opinion that “the answer to your auditors is not in encrypting passwords. The answer lies in restricting AM and SM capability to only those key personnel who can use the the “;pass” parameter within established policy. AM and SM capability also presumes the same capability to change another user’s password, and therefore also the ability to look it up.”

Chris Boggs reported in a virtual testimonial that “Our auditors were not satisfied by even limiting SM and AM capabilities to only two individuals (both in our department). Since we had Vesoft's Security/3000 already, I changed our regular logon ID’s to use the Vesoft password which is encrypted.

"There are other features in Vesoft security which are handy when dealing with auditors such as password obsolescence, password “history,” minimum password standards, inactivity logouts, day/time restrictions, automatic deactivation of logonID’s after a certain number of failed logon attempts, and probably a few others.”

Bradmark’s Jerry Fochtman said some Interex Contributed Software Library routines can help. “I developed a routine to return the passwords for user/group/account (based upon caller’s capabilities) during this time. It also signaled if the password was encrypted, simply returning blanks in this case. There was another routine which given a password, would encrypt it based upon HP’s approach and tell the caller if the entered password matched the one in the system directory.”

Fochtman also took note of the Vesoft abilities and added his humble opinion on the security solution from Monterrey Software, “SAFE/3000. It also utilizes one-way encryption for its passwords. And in terms of strictly security, it is a better tool in several areas, such as network security.”

Michael Gueterman, whose company Easy Does It Technologies does pre-audits for 3000 sites, added notes on using only session-level passwords.

“That’s fine for some things, but I still recommend keeping at least MPE Account passwords in place for all but the most “open” areas. For accounts with SM or PM, I also recommend MPE User passwords as well. Also, when at all possible, explicitly define what people are ALLOWED to access, instead of using generic wildcards. Wildcards make auditors unhappy, and an unhappy auditor is dangerous!”

Image by meineresterampe from Pixabay

Posted by Ron Seybold at 03:41 PM in Hidden Value, Newswire Classics | Permalink | Comments (0)

January 07, 2020

So many owners = so much value

Office building colored floors
Editor's Note: While the MPE/iX MANMAN customers mull over their 2020 options, it's useful to look at the history of an application being orphaned by its creator. Cortlandt Wilson, a consultant on ERP systems, wrote this early-years history of how MANMAN's ecosystem evolved. The bottom line is proof that value in an application like MANMAN is baked-in — or it wouldn't have survived so much change.

Over the years MANMAN has experienced highs and lows. At one time the software's creator, ASK Computing, was a media darling — a successful high-tech company founded and run by a woman, Sandy Kurtzig. The MANMAN product has a good reputation in the mid-sized manufacturing systems market. The company, however, unsuccessfully tried to follow up its success with a next generation solution based on a new technology infrastructure.

When I was with ASK in the late 1980s, on several occasions I heard the president and co founder of ASK say that “we are an applications company, not a software tool company.” Unfortunately, the companies on top of the ERP market all developed their own technology infrastructure. The search for a new technology infrastructure led ASK to purchase Ingres for its relational DBMS and tools.

ASK finally purchased a infrastructure and the basic application software for a ERP system from a then little-known Dutch company named Baan. As part of the sales agreement ASK modified significant amounts of the functionality and called the application MANMAN/X. Strained by development costs and weak sales, the company floundered.

By 1994, ASK was facing a severe cash crisis. Looking for a financial angel or a buyer, the board of directors finally recommended a buyout offer from Computer Associates. Many ASK employees, however, responded to the takeover by resigning.

Industry analysts’ concerns about CA’s “ferocious reputation” and the loss of experienced staff highlighted the takeover of ASK. Many MANMAN customers expressed skepticism about CA’s ability to maintain the product, and the quality of support noticeably dropped. 

By 1996, CA concluded that application software and services shouldn’t be managed like software tools and utilities. CA spun off its manufacturing products into an independent business unit to be named the MK Group (MK for Manufacturing Knowledge). MANMAN/X was renamed MK to reflect its marketing role as the flagship product.

Note: Wilson reported from a user group meeting of CAMUS in the late 1990s that the MK Group began to prove stable and was responding better to customer needs. There's inherent value in MANMAN that the repeated transfers of ownership have scarcely erased. By this summer, sites using the ERP package will have right of use for a product that has endured three changes of ownership. The software went from ASK Computer Systems to Computer Associates to SSA Global to Infor. The final owner of MANMAN, Infor, kept up support for nearly 14 years.

Photo by Takafumi Yamashita on Unsplash

Posted by Ron Seybold at 04:28 PM in History, Homesteading | Permalink | Comments (0)

January 02, 2020

Even in apps retirement, 3000 data survives

Aging hands on keyboard
A notable manufacturing datacenter in the 3000 community is making changes to its application lineup over the coming year. Although the profile of the apps and their status is changing, there's no talk of removing MPE from the datacenter yet.

Al Nizzardi is part of the IT command at TE Connectivity, the company that has more MANMAN instances running than any other in that ERP ecosystem. There's been a devotion to the 3000 that's extraordinary. Terry Simpkins has been the face of using 3000s in manufacturing since the 1990s. The IT director at TE even appeared once in a magazine ad promoting the 3000.

At TE, plans for the future of ERP applications have been aimed at SAP for several years. It's a migration, but one with echoes. SAP shares a customization practice with MANMAN: both apps are better choices when they're tuned to individual business practices.

After a few decades of use, the data repository for a MANMAN site becomes an asset that deserves its own curation. Data from a 3000 goes back to the late 1970s. The final cutover to SAP is likely to take place in late 2020 or sometime in 2021, by Nizzardini's reckoning.

"Databases are slowly migrating to SAP," he said. "I believe the final cut over will be 12 to 24 months out from today. That does not mean the end of the HP 3000. Historical data will reside on a HP 3000 of some sort."

TE runs a production N-Class, a test N-Class, an N-Class disaster box, and an A-Class. The datacenter does some Netbase shadowing, Nizzardini added. "We are still formulating a plan on our options, whether it's using an emulator or the N-Class we have" for archival MPE computing. "Either one of those options will be moved to a co-lo."
 
Experts on managing MPE/iX computing never stray too far from a place of helping. "We will be ready for when the Phoenix arises," Nizzardini quipped. "I've often said they will have to yank that HP 3000 out of my cold dead hands."
 
Image by Steve Buissinne from Pixabay

Posted by Ron Seybold at 12:41 PM in Homesteading, Migration | Permalink | Comments (0)

December 24, 2019

Gift wishes heading into the future

Christmas tree
Today's the day when generous people tuck presents underneath a holiday tree. Not so long ago, the only museum devoted fully to HP's computers was looking for gifts of classic hardware to flesh out its collection.

The HP Computer Museum is based in Melbourne, Australia. Its founder Jon Johnston passed away but left behind gift requests. The museum is downsizing now, like a lot of the owners and managers of HP 3000s. It's worth noting, though, that HP's breakthrough 3000 designs were among the most desired of museum gifts.

A table provides a listing of major hardware products the museum was seeking. This matrix lists the items by rarity and product category. Near the top quadrant: HP 3000s first released 45 years ago.

HP Computer museum needs
The white boxes represent the most needed items. The museum has no samples of these items. Pink boxes are most rare.

An HP-built 3000 server is old by definition now. The freshest pieces of hardware were manufactured more than 16 years ago. The craft and design of the HP iron, of any vintage, was legendary as well as being a gift of legacy. Even if MPE/iX is the only thing in use at some sites in 2020, because Stromasys emulation has taken over there as the hardware platform, HP's hard goods made that environment a classic.

Several resellers still trade in HP's MPE/iX iron. Cypress Technology's Jesse Dougherty continues to leave reminders about his 3000 system stock. Ebay is another reliable source, a place where the systems are often being sold by a reseller like Cypress. A Series 969 220 was for sale this week at $1,450.

Happy holidays. We're taking a break until just beyond the new year, when we mark the start of the 47th calendar year of MPE and 3000 service.

Posted by Ron Seybold at 12:21 PM in History, Migration, Web Resources | Permalink | Comments (0)

December 19, 2019

Seeking forgiveness as a support plan

So sorry chalkboard
ERP software becomes wired in deeply at corporations. Now that MANMAN has seen the end date coming for its manufacturer support, customers who rely on the ERP suite are looking for a 2020 plan to keep using it.

One aspect is a ruling about whether a product or a vendor is dead, but the intentions for its product lives on. It's an aspect of law called droit moral in France. Droit moral is not recognized in the US. Intentions are preserved in droit moral.

Some HP 3000 owners considered HP a dead entity after 2008, when no more patches were being built. HP's intellectual rights to MPE and the HP 3000 remain in effect. But there are those moral rights, too. This computer would not have become the keystone at places like aircraft makers and airline ticket agencies without customers' contributions, work that started many years ago. In fact, HP once recognized this kind of help in the market with the e3000 Contributor of the Year Award.

Contributors earn rights when measured in terms of ethics. Droit moral preserves ethics.

Source code rights might belong to customers once a product goes into permanent hibernation at the manufacturer. In 2008, I wrote that I believed that in order to honor droit moral for the 3000 community, HP's increasingly restrictive statements of licensing needed to stop. The vendor's support group needed to move on to other profitable HP markets. The vendor needed to leave owners and customers to continue using their computer, without any extra licensing payments to HP.

Droit moral lived in the hearts of some of the 3000 advocates within HP. While I visited HP's 3000 group one afternoon, the former business manager Dave Wilde and I walked across the wooded HP campus to lunch. That entire campus site is now the location of Apple Park, Apple's worldwide HQ, so things have changed a lot. At the time, through, Wilde said the 3000 group wanted to give the system "the ending that it deserves." It sounded warm and genuine.

Infor, the owners of MANMAN, are not as warm and genuine, even though they have enough sense about branding to sponsor the NBA Brooklyn Nets with an Infor logo on Nets uniforms. At the moment there's no coordinated effort from the remaining MANMAN customers to establish whether MANMAN truly belongs to customers after the exit of its creator. The customers are unsure who might even respond to such ownership questions.

Read "Seeking forgiveness as a support plan" in full

Posted by Ron Seybold at 11:50 AM in History, Homesteading, News Outta HP | Permalink | Comments (0)

December 17, 2019

Does orphaned source code belong to you?

Orphan with bike
Not many application vendors still have shingles hung out for business in the MPE market. It's also been awhile since any vendors made an exit from the MPE marketplace. Now that Infor, makers of MANMAN, will depart this coming June, its ERP customers are talking about what source code is rightfully theirs.

During a conference call, about a dozen customers and another dozen independent support vendors kicked around the idea. Every customer on the call had signed a MANMAN license agreement, way back in the 1980s or 1990s. It was generally accepted that you never own a piece of software unless your organization wrote it.

To put it more plainly, the use of a vendor's product is always governed by an agreement. Everybody agrees on the rules for ownership and use.

Then conditions change.

The vendor folds up a product line, or goes out of business altogether. It happened with MPE/iX, to note one instance of the former fate. 3000 users can scarcely take a few steps before they stumble on a software vendor who's closed down all business. That's what happens over time after a vendor has built the bulk of its business around a server that's no longer sold or supported by the manufacturer.

The new condition gets managers asking about why any license should apply to an orphaned product. Permission to own the code that's only been licensed — that's a matter for the courts, or at least lawyers represening both sides.

The hard place the managers encounter is the language that keeps software in a vendor's IP locker. In cases like these, IP not only means Intellectual Property. It means, "in perpetuity." If anyone has a digital copy of their contract, searching for that phrase will certainly bring up a hit.

Eleven years ago, the 3000 community talked this through while Hewlett-Packard considered the new licensing of MPE/iX source code. Customers wanted their intention of owning a 3000 — to run a business in perpetuity — to match the intentions of HP's product licensing. We invoked French law to give voice to our wishes for that outcome.

There is an aspect of French law which does not exist in US law. It's called "droit moral," meant to protect the moral rights of ownership of a work of art. Even more than HP's support group, the 3000 community considered MPE/iX to be a work of art.

One story about using droit moral in the movie business:

Droit moral is an intellectual right of an artist to protect his work. When an artist dies, the droit moral goes to his heirs, unless he appoints someone else. For example, a John Huston movie was colorized in the US, and the movie was shown this way in the States, despite the opposition of the Huston heirs who were trying to honor their father's artistic wishes. But in France, where the Huston heirs argued their father didn't want his film to be in color, the colorized film can't be shown because of droit moral.

The argument, one which might be tested in court, is that the intention of investing hundreds of thousands of dollars in a product is to use it in perpetuity. Ownership of source hasn't been much tested in US law. The places where cases have appeared before a judge are courtrooms where things went better for the customers than the manufacturers.

Image by Isa KARAKUS from Pixabay

Posted by Ron Seybold at 10:26 AM in Homesteading, News Outta HP | Permalink | Comments (0)

December 12, 2019

Information for MPE/iX: Always Online

NClass movie

HP's movie tour of the first A-Class systems, still online

Time machines transport us through the power of timeless information. It can take us way back, into the era when legacy technology was current and popular. In the 3000 community we are connected by wires and circuits and pulses of power. We always were, from the days of black arts datacomm that pushed data off of cards of punched paper. We’ve lived through a glorious explosion of ideas and inspiration and instruction. It’s a movie that always has another story in waiting, this Internet, so ubiquitous we’ve stopped calling it by that name. In 2019, 45 years after MPE became viable and alive, the World Wide Web is named after an element common throughout the physical world: the cloud.

And through the magic of these clouds come stories that lead us forward and allow us to look back at solved challenges. My partner Abby and I sit on the sofa these days and play with paper together, crossword puzzles, especially on weekends with the New York Times and LA Times puzzles. We look up answers from that cloud, and it delivers us stories alongside the answers. Finding the Kingston Trio’s hit BMT leads us to the Smothers Brothers, who started out as a comic folksinger act. After HP 3000 strategy TV broadcasts came alive via satellite, there were webinars. Today, YouTube holds stories of the 3000’s shiniest moment, the debut of the ultimate model of that server's N-Class.

Gravity - George Clooney One night we sat on another couch and watched the splashiest celebration of stories in our connected world, the Academy Awards. Despite racking up a fistful and more of them, Gravity didn’t take the Best Picture prize that night. A thing can have many elements of success, enjoy parts of being the best, and not end up named the winner in the final balloting. The 3000 saw a similar tally, a raft of successes, but its light began to fade from HP's vision. In the movies they call the last light of the day magic time, because it casts the sweetest shades on the players and settings.

It’s magic time for many of the 3000’s stalwart members in its special academy. The 3000 is remaining a time machine in your reaches of space. It's data is like gravity, a force to unify and propel. MPE systems contain ample gravity: importance to users, plus the grounding of data. Data becomes information, then stories, and finally wisdom.

And in our magic time, we are blessed with the time machine of the Web, the cloud. Users and owners of HP 3000s will always be able to look up wisdom of this community online, written in stories, illustrated in video, told via audio. Find it here, as well as in the cloud at the following resources:

The HP Computer Museum

3K Associates

The host of the HP Jazz papers, Fresche Legacy

Then there are the fallen, resources no longer at their original addresses.

The MM II Support Group

MPE Open Source.org

Those last two are live links today, however, thanks to the Internet Wayback Machine. The Wayback is such an enterprise now that it's fundraising this week. The arrival of Wikipedia was met with skepticism at first, and it's still sneered at in some places. The popularity of Wikipedia is demonstrated in the way it appears as the first result in many a Web search.

The Wayback will save what we don't remember, even as it moves off of its legacy addresses. These very Web pages you're reading are likely to be Waybacked. We began putting the NewsWire on the Web in 1995. The dream was that our website would be like the 3000 in one way, Always Online. By now, by way of the Wayback, it seems the dream has come true.

Posted by Ron Seybold at 08:12 PM in Homesteading, Web Resources | Permalink | Comments (0)

December 10, 2019

Wayback: HP FAQ captured its OS visions

Canary close up
It's only available through the Internet Wayback Machine, but a record of HP's intent for its enterprise operating systems still exists. For reference we traveled to LegacyOS, a website devoted to the legacies of Sun and HP's enterprise products. The promised land, as HP imagined it 17 years ago, was getting its operating systems to the Itanium Processor Family.

HP's decision to keep MPE/iX away from IPF servers was the canary in the coal mine for the company's business intentions for HP 3000s. Such a canary roosted in mines while work proceeded. If the quality of the air turned poisonous, the canary died and the miners evacuated.

At the time there were only two models of Itanium processors in working servers from HP, so calling it a family was marketing optimism. Nevertheless, moving to the nascent IPF, as well as a new OS in HP-UX, was HP's vision of end-of-life. The life ending turned out to be at HP's MPE/iX labs eight years later, rather than any useful lifespan for MPE/iX.

There is a current-day lesson in any review of the HP 3000 plans of 2002. HP noted at the time the vendor created a Business Critical Systems group. That group, HP's cheerful-but-inaccurate 3000 plans, and HP itself in its classic makeup don't exist anymore. Users can count on their community, rather than the vendor, to see the conditions for any end of life canary.

Q: What is HP’s strategy moving forward with HP e3000 servers?

A: Our commitment to HP e3000 and MPE/iX operating system is to continue delivering on the roadmap we have already communicated, delivering the planned performance and functionality, with future MPE/iX releases in the 2002-2003 timeframe. Moving forward, we are focused on moving HP e3000 customers to IPF-based HP servers that deliver more benefits to the customer, using aggressive and innovative migration programs.

Q: Does HP plan to port MPE/iX to IPF-based platforms?

A: No. MPE/iX will not be ported to Itanium-based servers. The communicated HP e3000 roadmap includes PA-RISC based servers that deliver the performance and functionality customers need in the 2002-2003 timeframe. After that, HP e3000 customers benefit more by moving to HP Unix Servers using Itanium technologies and best-in-class migration programs, and taking advantage of the industry leading performance, functionality, and lasting value that Itanium and HP-UX will deliver.

Q: Should HP e3000 customers who need to stay longer on the platform than 2004 be concerned?

A: Absolutely not. HP will support the servers at least until the end of 2006. During this time, HP is committed not only to provide full support for the servers, but also to make available the aggressive and innovative migration programs, to help customers successfully move into Itanium-based HP-UX servers on their own pace.

To recap, the end of 2006 became the end of 2010, in part because HP's aggressive and innovative migration programs were undermatched to the needs of the customer. The Itanium technologies became an also-ran, lapped by Intel's modernization of x86 processors. Intel announced its departure from Itanium futures in 2015. Now commodity hardware rules the roost in today's mines.

Photo by David Clode on Unsplash

Posted by Ron Seybold at 08:12 AM in History, Homesteading, News Outta HP | Permalink | Comments (0)

December 05, 2019

Set a Watch for Jobs That Hang Others

Guard tower
Jobstreams deliver on the HP 3000's other promise. When the server was introduced in the early 1970s it promised interactive computing, well beyond the powers of batch processing. Excellent, said the market. But we want the batch power, too. Running jobs delivered on the promise that a 3000 could replace lots of mainframes.

Decades later, job management is still crucial to a 3000's success. Some jobs get hung for one reason or another, and the rest of the system processing is halted until someone discovers the problem job and aborts it. When it happens over a weekend, it's worse. You can come in Monday and see the processing waiting in queue for that hung-up job to finish.

Is there a utility that monitors job run time, so that it can auto-abort such jobs after X number of hours? Nobix sells JobRescue, a commercial product for "automatically detecting errors and exception messages; JobRescue eliminates the need for manual review of $STDLISTs, making batch processing operations more productive."

Then there's Design 3000 Plus. The vendor still has a working webpage that touts JMS/3000, a job management system that was at one time deployed at hundreds of sites. Its powers include "automatic job restart and recovery. Whenever a job fails, a recovery job can be initiated immediately."

The home-grown solutions are just waiting out there, though, considering how few 3000 sites have a budget for such superior software. Mark Ranft of Pro3K shared his job to check on jobs. The system does a self-exam and reports a problem.

Read "Set a Watch for Jobs That Hang Others" in full

Posted by Ron Seybold at 07:32 AM in Hidden Value, Homesteading | Permalink | Comments (0)

December 03, 2019

Where MANMAN support goes, post-Infor

Abandoned storefront
Sometimes established structures go vacant without even knowing they've been abandoned. That might be the case for the support service for the MANMAN ERP software. There are at least 35 companies making use of the application suite on HP 3000s.
 
Those users have been served by the Computer Aided Manufacturing User Society. It's not only an operating user group, it's got a surplus in its accounts. That's the opposite of Interex, which stopped operations while owing millions.
 
On a recent conference call, one CAMUS board member said CAMUS is the best source to contact other MANMAN customers. Nobody on the group's call reported using Infor support anymore. Many of the users have an arrangement for help from an independent company like The Support Group.
 
Doesn’t that mean the customers have already made arrangements for their MANMAN support outside of Infor?

Infor had said last year that its cutting out app support because system vendors don’t support the hardware and OS for MANMAN. That wasn't true up to July of this year for VMS MANMAN users. But it’s been true for the MPE people since 2011.
 
It looks like Infor was happy to collect support for MPE systems for years after HP left the 3000. Now it looks like the VMS support migration away from HP Enterprise and into VMS Systems Inc. is the trigger for shutting down all of the MANMAN support at Infor.
 
The systems haven't been turned off, but the vendor has departed. That's a familiar situation for MPE/iX customers.
 
Photo credit: Rafał Malinowski on Unsplash

Posted by Ron Seybold at 07:27 PM in Homesteading | Permalink | Comments (0)

November 28, 2019

HP still keeps MPE data behind a paywall

Payphone
Photo by John-Paul Henry on Unsplash

It can be surprising to see how much value remains in an operating system that's not been altered in almost a decade. Hewlett Packard Enterprise has 3000 documentation on its website that is still behind a paywall of sorts. Users access this info by validating their HP Passport credentials — the ones that indentify the user as being current on a support contract.

The HPE website has plenty of advice and instruction available without a validation. If you ask, for example, "Can the HP 3000 and GSP LAN configuration be on different subnets?" HPE reports

There are two server platforms (A-CLASS [A400/A500] and N-CLASS [N4000]) that can run MPE, which uses the GSP (Guardian Service Processor) console for offline hardware operations like startup and shutdown of the system, access hardware console or system logs, etc.

It is possible for management purposes to place the GSP operation on a different subnet from the MPE server LAN, thus isolating or protecting either environment from one another. One reason for that can be to prevent normal users from telneting or in other ways accessing the GSP console or the other way around.

Or, another morsel that's useful in the era of declining hardware know-how: A-Class IO path memory configuration guidelines. Useful for the manager who's trying to set up memory cards in one of those $5,000 replacement 3000s.

However, if you'd like to read the most current documents, a support contract stands in the way. An updated NMMAINT listing is behind the paywall. HPE created the document in August of 2019. There's no available support to be purchased from HP for MPE/iX.

The documents that survive can be extensively redacted. A HP3000 License Transfer Process document references a web address no longer in service. The address licensing.hp.com no longer answers to requests.

Some information on MPE/iX at HPE's website is among the 4,386 documents at the site. Having the confidence that it will remain in that place is the next step in learning to rely on HPE resources. Independent MPE/iX resources have been more reliable, although the web pages for MM Support went dark this year.

Posted by Ron Seybold at 03:05 PM in Homesteading, News Outta HP, Web Resources | Permalink | Comments (0)

November 26, 2019

Xerox HP fight copies 3000's exit saga

Copier user
Xerox has been trying to buy the part of HP left over after the vendor's split up in 2015. The latest $33.5 billion offer, rebuffed by the HP Inc. board, is going to get pushed out to the HP Inc. shareholders. "It's a better deal that you're getting now" is the message to the thousands of HPQ stock managers. Voting shares for or against a merger has a spot in the 3000's legacy.

This is also the outcome that helped cement HP's exit from the 3000 world. In 2002, HP's acquisition of Compaq got pushed out to a proxy battle. Xerox says HP is defying logic by refusing to be acquired. That's the kind of resistance HP loyalists — the HP blue, they were called — tried to muster around Bill Hewlett's son, who was an HP board member.

Without that successful buyout, HP would've had no Digital VMS customer base to court and invest in to feed a business-focused Itanium operating system. HP-UX was a lock for Integrity, to be sure. The 3000 and MPE/iX were there, ready, but just too small for HP's designs on being Number 1 in all of computing. The Compaq PCs were going to make that a reality.

However, about three years after HP rammed through the Compaq merger through a proxy battle, the spark of that deal Carly Fiorina was forced to resign as HP CEO and chairman. PC growth had not contributed to significant HP market dominance. At the same time, the health of its enterprise business began to slip ever so slightly.

Another CEO pumped up HP's sales, even while its ability to sell OpenVMS and HP-UX faltered. Enterprise computing with HP-built operating systems was in decline. HP became an all-Windows enterprise supplier when full business server sales were measured.

The juicy fruit that HP's board dangled in front of uncommitted shareholders was Compaq's roaring PC business. A combined company would be No. 1 in market share almost immediately. That was promised, anyway. The fortunes of OpenVMS seemed secure, heading into the portfolio of a technology giant that had enterprise legacy to match Digital's.

By 2016, OpenVMS was in the chute toward ex-product status at HP. The coup de gras took place this year when VMS support customers were told the future was in the hands of VMS Software, Inc. OpenVMS users, as well as the MPE customers who were the casualty in that 2002 merger, can look at Xerox and watch the conflict knowing it won't change their fates.

Those were set in motion by the last proxy battle. The juicy fruit that HP's board dangled in front of uncommitted shareholders was Compaq's roaring PC business. The fortunes of OpenVMS seemed secure, heading into the portfolio of a technology giant that had enterprise legacy to match Digital's.

MPE customers were sent down the path where Tru64, another Digital creation, sits today. Formal support ended for them. However, MPE/iX was more than a new edition of Unix. It built a community around vendors. There was no other choice once that proxy war was lost.

Mergers are a good way to see where the soul of a company resides — if there's an open fight. Of course, there wouldn't be a shadow of the old HP to fight over — printer-PC HP Inc. — if the Compaq acquisition had failed. HP might be in the position of seeing itself absorbed and erased. A new afterlife seems unlikely for a company founded on something as common as Windows and printers.

Posted by Ron Seybold at 11:50 AM in History, News Outta HP | Permalink | Comments (0)

November 21, 2019

Values hidden by time get revealed by vets

Brass treasure key
Photo by Michael Dziedzic on Unsplash

Twenty-four years ago we started unlocking Hidden Value for HP 3000s: Commands that only the veterans know, plus the processes that have been plumbed to bypass MPE's blind alleys.

Some of the value is specific to a 3000 process like using EDIT/3000. It's antique, that editor, but it's on every HP 3000.

I use cut and paste with EDIT/3000 to enter data to batch files.  It works well except that I am limited by the size of the scratch file. Can I change the size of this file so I can paste more at a time?

Immediately after entering Editor, enter "set size=######" to give yourself more space.

For other tasks, like finding forgotten passwords, and keeping them fresh and the 3000's data secret, more elaborate answers have surfaced.

A system manager pitched his plight.

"My operator, in his infinite wisdom, decided to change passwords on manager.sys.  Of course he forgot, or fingerchecked... I don’t know.  At any rate I need some help. Any suggestions, other than a blindfold and cigarette?"

Several versions of help involved the use of utilities from security experts VEsoft. "Do you have the GOD program on your system? If so, it has PM capability, and so it can give the user who runs it SM capability. So it will allow you to do a LISTUSER MANAGER.SYS;PASS=

(That's why GOD should be secured, by the way. A randomized lockword will do the job, visible only to users who have SM capability. When VEsoft installs MPEX, for example, it installs a randomized password to MGR.VESOFT, and to GOD.PUB.VESOFT.)

Paul Edwards, ever a source for HP 3000 training, ran through the backstop methods every system manager should practice to avoid such a dilemma.

1. You run BULDACCT prior to each full backup so you can look in BULDJOB1 for the passwords 
2. You have another user on the system with SM capability and a different password as a backup in case this happens  
3. Your operator used LISTUSER MANAGER.SYS;PASS just after changing the password to verify the accuracy as spelled out in the Operations Procedures section in your Systems Manager Notebook   
4. You have a Systems Manager Notebook

  Then Duane Percox of K-12 app vendor QSS opened up a clever back door:

If your operator can log onto operator.sys:
file xt=mytape;dev=disc
file syslist=$stdlist
store command.pub;*xt;directory;show

Using your favorite editor or other utility search for the string: "ALTUSER MANAGER  SYS" You will notice: PAS=

, <passwd> which is your clue

Posted by Ron Seybold at 12:25 PM in Hidden Value, Homesteading | Permalink | Comments (0)

November 19, 2019

Keep Passwords Fresh on 3000s: Methods

Fresh bread
Photo by Clem Onojeghuo on Unsplash

It's usually a good practice to keep passwords fresh. A 3000 development manager once posed a question about how to do this while staying inside the bounds of MPE/iX. He had the usual limited budget associated with HP 3000 ownership.

"Management wants users to be forced to change their passwords on a regular basis. Also, certain rules must be applied to the new password. I don't have budget for the good tools for this, like Security/3000, so I need to write something myself, or see if there's any contributed code to do the job."

Homegrown and bundled solutions followed. When Jeff Vance worked in the 3000 lab at HP, he offered the pseudo random password generator as a solution. It's in the HP Migrations webpage hosted on the website of the company formerly known as Speedware. These HP Jazz solutions that used to be on the HP website are still available at Fresche Solutions.  

There are UDCs on Jazz which force a password to be supplied when using NEWUSER, NEWACCT and NEWGROUP CI commands. These required passwords can be random (uses the script above) or user entered with a minimal length enforced.

Then Vance added as an afterthought, a strategy to program your own password system:

I haven’t thought about it much, but it seems you could have a password file (maybe a CIRcular file?) for each user on the system. This file would have their last N passwords, and the modified date of the file would be the date their password was most recently changed.

A logon UDC could detect if the password file for that user exists. If not create it and require a new password right then.  If the password file exists then get it’s modified date and compare that to today’s date. If greater than X days then in a loop prompt for a new password. Validate the entered password against previous N passwords and your other rules. Maybe run a dictionary checking program to make sure the password is not common, etc.

Update the user-specific password file with their new password, and then logon the user.

From the user community, Donna Hofmeister weighed in with this advice:

If you have no choice other than to develop your own software, then I’d certainly model it after what VEsoft has already done. That is:

Based on a system-wide UDC, examine all sessions (it is just sessions, yes?  By the way, a DSLOGON from inside a job is still a session....) against a ‘database’ (By the way, just how secure is this database?  A real database needs passwords... Who’s going to maintain that? A flat file could be lockworded... but that’s not a slamdunk answer.) which is looking for the ‘age’ of the password (By the way, are you going to provide an advance warning period?). 

If it is time to change the password, get the ‘new’ password from the user... but writing the rules is a pain, and keeping track of reused passwords is just annoying. Auditors in the states love when you can say the password is one-way encrypted.  Dunno what your management is saying for encrypting an MPE password.

Read "Keep Passwords Fresh on 3000s: Methods" in full

Posted by Ron Seybold at 12:27 PM in Hidden Value, Homesteading | Permalink | Comments (0)

November 14, 2019

Welcome to Year 19 of the Afterlife

Cheated Death on printer
You remember where you were, perhaps, on the day you learned Hewlett-Packard was done with MPE/iX. You might have been in a meeting, or checked your email before heading home. You might have been installing something a 3000 needed to keep serving your company, or even ready to order a new server to replace the old 9x8 box. Some unlucky vendors were holding orders for new systems.

People did all of that and more on the day HP revealed its 3000 era was on its way to a finish. By 2003 the community was calling the new era The Afterlife. The lifespan of building new HP 3000 hardware was ended when a box rolled off the line at the Roseville plant in early November that year.

Afterlife shirt

And so, on November 14 of 2001, the afterlife of Hewlett-Packard's lifetime started with dismay, anger, and then resignation. The five stages of death proceeded through discussions in a lively 3000 newsgroup. Taking a cue from the horrors of 9/11 in that season, programmers and vendors howled about the relatively unimportant death in their lives.

Doug Greenup was leading Minisoft in that November week. The CEO of a software company whose products were on thousands of systems, he became aware of the HP pullout with just one day's notice.

Alvina Nishimoto from HP called me. She was in charge of third parties with HP at the time. She asked me to sign a non-disclosure which she'd just faxed me. She said she had important news. I signed it and faxed it right back. She called to tell me HP was announcing they were discontinuing the HP e3000, and that HP-UX was their future direction.

HP might have been worried the story was going to get into the world without its influence. The news had been roiling through the 3000 community for more than a week before I learned about it. Wirt Atmar, the founder of AICS and a 3000 stalwart, threw off the lid about the pullout by posting on a developers' mailing list.

I spoke to two of our oldest, most trusted customers yesterday, one a ten-year customer and the other 15 years, about this upcoming announcement. Their first reactions were that it simply sucked their breath away. When I told them about HP's proposed plans of migrating their applications to HP-UX — which as an option has all of the practicality to them of trying to establish a penguin colony in Death Valley — their second reaction was "the hell with HP. If we move, it will be to anything but HP." I think that that's going to be the general reaction.

HP learned a great deal about ending a product line with the lessons that began 18 years ago. Earlier this year the company made a graceful transfer of responsibility for OpenVMS, sending the software as well as support opportunities to an independent firm, VMS Software Inc. HP won't sell any more OpenVMS licenses, although it continues to build Itanium servers that will run the apps created for that OS.

This was a vision that the HP 3000 community took to heart during the first of the 18 years that followed. A similar group of OS experts, organized and led by Adager, wanted HP to transfer the future of MPE/iX to new, independent stewards. HP didn't know how to do this in 2001 or in 2002. The offer took another form in the OpenMPE advocate organization, but eight more years had to slip past before HP's source code made its way into independent software labs.

A new history began 18 years ago, a record of a group of computer customers who kept their own counsel about walking away from a corporate computing asset. The next two years or so will show HP Enterprise customers what might have been possible had MPE/iX found a third party home. VSI is predicting that it will have production-grade OpenVMS ready for a late 2021 release on Intel hardware.

This is more than a shift away from the HP Enterprise resources. VSI is carrying OpenVMS to a new chipset, a commodity home. In 2001, Atmar told HP's business manager for 3000 operations, Winston Prather, such a move was what the 3000 customers deserved.

Opening MPE up and migrating it to an Intel platform offers at least some real hope for a continued and bright future for MPE. More than that, it keeps a promise that most customers believe HP has made to them, and that is very much the nub of the moral and ethical question
that faces you now, Winston.

Migrating MPE to Intel hardware would have permitted MPE to run on inexpensive but high-quality servers. Earlier this month, VSI announced a timeline for such a thing to happen with OpenVMS. A different HP paved the way, perhaps chastened by the past, than the corporation that launched the afterlife. In the beginning, the launch of the server took place in this month. The slogan in 1972 was "November is a Happening." Nothing can change what happened nearly 30 years later, but the decisions over what to do with a loyal enterprise customer base have changed in the years since 2001's happenings.

Posted by Ron Seybold at 08:03 PM in History, Homesteading, Migration, News Outta HP | Permalink | Comments (0)

November 12, 2019

Customized code care can add time to apps

Sewing spool and scissors
Image by TooMuchCoffeeMan from Pixabay

Almost all MANMAN sites have the original ASK FORTRAN source code, but a few have lost track of some of their mods. It seems hard to believe for some analysts, but source code for applications can go missing, too.

It's easier to believe while considering the age of this software. A 3000 which first booted up a manufacturing site for TE Connectivity in 1978 recently got powered down for the last time. The software had outlasted a half-dozen servers, until finally the need lapsed for a host of an application launched 40 years ago.

TE has its source code for every 3000 instance it continues to run to manage manufacturing across North America and Asia. You can imagine, though, how much care it takes to keep that corporate asset in good working order from the era before PCs until today.

Managing the modifications is the other essential piece of the lifetime-extending effort for MPE/iX ERP solutions. No app suite is as often customized as ERP; the term itself is an extension of the Manufacturing Resource Planning of the 1980s. What was once MRP is now ERP. The planning always needs custom code, tailored to the business processes.

Terry Floyd, the founder of MANMAN support resource The Support Group, said most MANMAN sites — and there are under 100 by now — own and manage their own source code for the main application set. His company deals with modifications for its clients, too. Without the mods, using such apps is a matter of being frozen in time for features. It's like trying to take out a pair of pants without enough material.

Posted by Ron Seybold at 09:54 AM in Homesteading | Permalink | Comments (0)

November 07, 2019

What MANMAN sites didn't know until now

 
Just this week I sat in on a call among MANMAN users. There were not a lot of them, but 20 people dialed in or opened a GoToMeeting for a conference call. Regional Users Group meets have disappeared except for the Computer Aided Manufacturing User Society. Unlike some 3000 owners and managers, CAMUS still has something to talk about.
 
The discussion topic is change. Here in 2019, CAMUS users manage HP 3000s that continue to manage manufacturing. That situation will change sooner or later. Manufacturing software is wired in deeper to a company's nervous system. Enterprise Resource Planning (ERP) can apply to many MPE/iX shops. ERP has the longest lifespan of any application package.
 
Not long ago, an ERP system was shut down at TE Connectivity. TE is probably the biggest customer by number of ERP 3000s. The system that was powered off for good hosted software that had been running since 1978. Other ERP instances there, not as old, continue to run.
 
Forty years. One operating environment. A legacy within a legacy community. In essence, though, just the longest lived veteran in a room of greybeard application suites. The shutdown didn't even come up during the RUG call.
 
The meeting was organized by the CAMUS president Terri Lanza, who led the discussion. One of the more interesting parts was the group’s take on Infor support. The vendor's dropping support altogether for MANMAN customers next year. Some customers on the call were just learning that Infor's support verdict has been out since 2018. 
 
Support disappearances are not exactly news in the 3000 world. This strategy by Infor, though is notable. It's perhaps the last departure from an application vendor.
 
That might be due to the number of MANMAN sites still paying Infor support fees. Nine, in all, and that was as of last year. Revenue from support is the canary in the mine shaft for vendor decisions like this.
 
Infor is closing down an operation for MPE/iX that had been available in name only, if the RUG's intel is accurate. In less than a year, even the backup to the best engineers will be working on something else. By May, the app will stop being sold or supported. Go-time for software that's free, right? Not so fast there. The 3000 community has been through this before.

Read "What MANMAN sites didn't know until now" in full

Posted by Ron Seybold at 01:34 PM in Homesteading | Permalink | Comments (1)

November 06, 2019

HP's tech lures Xerox offer to buy

HP printer tech
Photo by Dario Seretin on Unsplash

Plenty of writers and customers get confused about the HP of 2019. Back in 2014, the corporation split into two units, operations that align on devices and datacenters. Hewlett-Packard Enterprise now sells datacenter products and services. It's the arm that created HP 3000s in the 1970s. HP Inc., the part of the company that makes printing, imaging, and PC products, received an offer this week to be purchased by Xerox.

HP Inc. acquired 55 patents from Samsung for business printing not long ago. Now the corporation is being courted by a suitor whose printing legacy is wired deep into the DNA of Hewlett-Packard's spinoff. Patents for LaserJet tech, the engineering that in 1984 took HP into the realm of resellers farther removed from HP than any 3000 VAR, are part of what Xerox is bidding for.

The offer, which HP Inc. confirmed it has received, would sign up Xerox for more than $20 billion in debt, financed by Citigroup. The appetites of the HP which created the 3000 and then cut down the vendor's future in the MPE/iX market helped to spark that 2014 split. HP was striving for an overall Number One status as a technology supplier in the years just before it announced its takedown of its 3000 business.

Management at the vendor aligned on growing sectors of business. While the 3000 had enjoyed a nice revenue increase for several years leading up to Y2K, HP saw the unit as one whose growth had a limited future. "If it's not growing, it's going," one 3000 vendor said he'd heard in reports about HP's future.

This was in the era when PCs were soaring on the HP balance sheets and printer products were being sold in groceries. The corporation had just acquired Compaq and Compaq's Digital group, so while there was a future for OpenVMS and its growth, cutting back on enterprise products became essential in HP's strategy.

Xerox developed the first viable graphical interface technology in its Star systems. The Palo Alto Research Center's tour for a young Steve Jobs at Apple led to the mouse interface becoming an essential part of the Macintosh release. 

In a move that proves there's always value in technology that outlasts its creators, there's now a deal in the market to return some of HP business technology to a corporation that's been a part of business computing history since the 1980s. It can be hard to tell what's going to survive in computing and where it will land. HP Inc. recently announced it will be cutting 9,000 more jobs. Betting on good management is at least as important as betting on good technology. As 3000 owners know, technology like MPE/iX is able to outlive the interests of its creators.

Posted by Ron Seybold at 05:46 PM in History, News Outta HP | Permalink | Comments (0)

October 31, 2019

Wayback Wed: Leaving a Wake on an exit

Chris Gauthier  Jackie at Wake
Simpkins  Nizzardini  Johnson Wake redux
Above, a 2019 commemorative lunch today at Tide Mill Café in Hampton, VA with Terry Simpkins, Al Nizzardini and Tracy Johnson, all 3000 experts and veterans of MPE. 3000s are in use at their company, TE Connectivity. At top, a 2003 World Wide Wake picture with Chris Gauthier and his co-worker Jackie Mitchell, both supporting 3000 customers as contractors to Terix.

Today we're marking the 16th anniversary of the World Wide Wake. The event was a marker of the end of HP’s 3000 manufacturing on Oct. 31, 2003. Alan Yeo, who passed away recently, organized the Wake and posted photos contributed from attendees onto what we were still calling the World Wide Web. A Web gallery for 3000 people was groundbreaking at the time.

Yeo said back in 2004, a few months after the event that drew more than 400 devotees to meetings in 15 countries, “We have created a simple single Web page that by country just lists the venues and who attended, and also has a link to any pictures for that venue," Yeo said. “The information will be condensed into a single Web page, linked to a directory of about 75 images. We have had several offers to host the information, so rather than try and pick a single host, we thought that allowing any interested attendee to host it would be best.”

Thanks to good Web hunting from Keven Miller, the Wayback Machine link to the original Web page tells the tale of who attended, and where, along with some of the photos.

Our own archive of the photos, sans captions, is here on the blog.

The photos from that day look like party pictures, even though nobody in them was celebrating anything except Halloween. The memories were on the minds of everyone in the frame, though. The future without any more new 3000s didn't seem to scare anyone on that day, at least not for the cameras. It was a coincidence that the building of new computers, as well as the licenses for the MPE/iX that made the boxes genuine 3000s, stopped on the spooky holiday. HP's fiscal year ends every year on October 31.

The Wake gatherings were all across the globe. New Zealand was the furthest away from the Epicenter of Grief, as the 3000 faithful had dubbed Lori's Little Shack in Roseville, the town where HP's 3000 factory was ending its birth of the servers. 

Loree's Epicenter Grief

Read "Wayback Wed: Leaving a Wake on an exit" in full

Posted by Ron Seybold at 03:56 PM in History, Homesteading, Migration, News Outta HP | Permalink | Comments (0)

October 24, 2019

Crave sounds of connection? Call in Nov. 7

HP 3000 owners and managers, as well as developers working in manufacturing, can hear each other's voices next week. Those who crave the connection of conversation can call in on Nov. 7 at 10:30 CST (8:30 Pacific, 15:30 UK) for this year's CAMUS user group meeting. Email organizer Terri Lanza to register (it's free) and get the dial-in number.

For many years, the members of the Computer Aided Manufacturing Users Society gathered in person. The meetings were small in number compared to the attendance and exhibits at Interex events. But here in 2019, some 14-plus years after Interex died, CAMUS still hosts gatherings including this call. Sometimes the group, led by the cheerful and redoubtable Lanza, has met in person. This summer she set aside a meeting room at the local Dave & Busters in the Chicago suburbs for a Sunday afternoon gathering.

Everyone at this year's meeting was 20 years older than in this 1999 photo at the same Dave & Busters.

CAMUS RUG 1999 Larry_Vicky_Steven_Holly

At that same 1999 meeting, robotics toys were a part of the agenda. Because even in that year, it was easy to see that robotics was going to be a big part of manufacturing IT in the years to come.

VickyFalk 1999 CAMUS MWmeet

A conference call still has the word "conference" within, so the November 7 call is a gathering without the toys and games. There are important things on the agenda. In the wake of Infor announcing they're ending MANMAN support in 2020, the meeting will give attendees time to share strategies on hardware, software, hosting, application modifications, education options "and just plain answering questions forever," Lanza says, "or as long as you need it."

That's a better offer now than the one Interex was able to maintain. There's an advantage to running lean to stay in the game of gathering. Register with Lanza at [email protected]

Posted by Ron Seybold at 03:43 PM in Homesteading | Permalink | Comments (0)

October 22, 2019

HP's kids: Children who can't say yes, or no

Merry-go-round amusement parkPhoto by Marjorie Bertrand on Unsplash

Editor's note: Developer, vendor, and advocate Alan Yeo has passed away at age 65 after a lifetime of work for the 3000 community. His essay below was written in 2005 amid the early years of the computer's Transition Era. He wrote about the damage done after migrations were first triggered by HP's 2001 pullout, then postponed on a fuzzy timeline from the entity the community was calling the virtual HP division for the server, vCSY.

Vendors like Yeo who weathered HP's stormy strategy took on a lot of water because of HP's revision of its end of service deadline. Yeo's use of metaphor and allegory here are a fine tribute to his wit and intelligence that the world has lost. ScreenJet, his company, followed his insight to survive the turmoil.

By Alan Yeo

That's it, children, just give the merry-go-round another shove, just when passengers thought it was stopping and they could get off it and get on with planning the rest of their lives. Oh yes, some of the children will be happy; the period before they have to decide which ride to take next has been extended. But for the adults either already behind schedule, or struggling to get attention-deficit children to concentrate on important decisions, it's just another frustrating delay.

Now it wouldn't be too bad if the very Careless Stupid Youngsters ("vCSY") nudging the merry-go-round on weren't the same vCSY who had planned its retirement, and had then encouraged its customers and partners to seek out new more exciting future-proof rides. But no, to compound the disappointment they caused their passengers when they announced the ride was ending, they now have to say, “We lied, we didn't mean it, the ride's not ending yet!”

Is this because they think their passengers are still having the best ride in the fair? Perhaps they think they can just keep it spinning under their control for a while longer, that there are another 3000 pieces of silver to be extracted for their parents, the only Happy Party ("HP") in this.

And what of vCSY partners, and the encouragement they received to help transport the passengers to other rides when the Merry-go-round stopped. Or even those they encouraged to build an organisation to help those passengers that wanted to stay on the Merry-go-round and even maintain it after the ride had stopped.

For yes, there was an organisation of such Open Minded Passengers Established ("OpenMPE") that hoped to provide counseling and support for those who chose to stay, and even to build a work shop to repair the Merry-go-round Physical Environment ("MPE") for them. What of OpenMPE's chances now? Why would anyone invest in them when they need it, if the HP and vCSY are going to keep the ride spinning and the MPE supported?

And what of those who vCSY encouraged to build the transport for the passengers to other rides — their parents (the HP) had no transport of their own. Those vendors built the busses the planes and the trains, and even migrated some of the passengers to new rides. What are they to do now, just sit there with the engines running for a couple more years whilst the merry-go-round spins on?

And what of those partners vCSY encouraged to build infrastructures to keep old merry-go-round's functioning and provide support for the MPE? For them, the ride has been delayed for two more years, and it has reduced the number of potential passengers to the point where it may not be economic to hang around and wait.

Read "HP's kids: Children who can't say yes, or no" in full

Posted by Ron Seybold at 08:29 PM in History, Homesteading, Migration | Permalink | Comments (0)

October 17, 2019

Alan Yeo, 1954-2019

Yeo at Reunion

Alan Yeo, a software vendor and developer whose business ultimately led to success as a nexus for the 3000 community in its Transition Era, has died at age 65 after a battle with a small cell cancer. He is survived by his wife Helen, a lifetime of creations he designed with partners, and a gripping voice that gathered and rallied MPE customers after HP quit on their marketplace.

Yeo’s company Affirm, Ltd. rose up in the 1980s as a resource for manufacturers who used the HP 3000 to manage their enterprises. He served a group of customers across the UK and began to move in wider circles with the advent of ScreenJet, his software to modernize the 3000’s bedrock VPlus application interfaces.

ScreenJet arose in the years just before Hewlett-Packard scrapped its business developing 3000s and MPE. While the HP decision left Yeo undaunted in his business aspirations, it also led him to a new role as a leader for a 3000 community that was dissolving after the implosion of the Interex user group in 2005.

His first effort surrounded the final date of HP’s manufacture of the system. On Oct. 31, 2003, he organized and led the HP 3000 World Wide Wake, a collective of gatherings to celebrate the server and the people who’d made it their life’s work. Across North America and Europe, customers and managers held parties and met at pubs, bars, and restaurants. Photos from the events poured into a web server that Yeo hosted. Earlier in the year, Yeo asked out loud where else the HP 3000 community might gather in a user conference — a question he posed in a meeting at the Atlanta HP World, where few 3000 customers had appeared.

In the year that followed, he shared his strategy of being a master of one. It was built around the nugat of collaboration that led to his ability to connect.

"We’re starting to see more collaboration between migration tools providers and migration service partners," he said in a NewsWire Q&A. "To get some of this stuff right, you really, really need to know it. I think it’s too big for any one person to do anything right. If you want good fish you go to a fishmonger. If you want good meat, go to a butcher. If you just want food, go to Wal-Mart, and if you just want to eat, you go to McDonalds."

Community meets and reunions

Many of the stranded customers using the HP 3000 got an introduction to Yeo’s voice in those first years of the 3000’s Transition Era. He commissioned an editorial cartoon during 2002 that became a mainstay in his company’s ads, one built around the HP move to end its MPE plans and sever relations with the thousands of companies that grew up using the 3000’s extraordinary solution. The CEO of the company at the time, as well as the 3000 division’s GM Winston Prather, caught the brunt of the brilliance in a cartoon that compared killing off HP's 3000 futures to the evil in a Disney movie.

WinstonDalmations
A few years later, after the user group Interex folded its operations overnight and stranded users’ plans to meet at the now-canceled annual conference, the first of a series of Community Meets sprang up for 3000 owners. After an impromptu gathering in the Bay Area for community members already stuck with nonrefundable hotel reservations and air tickets, a single-afternoon lunch gathered several dozen managers, developers, and owners.

The first Bay Area meet was replicated and expanded twice more with single-day meetings in 2007 and 2009, organized and underwritten by Yeo and his business partner Michael Marxmeier of the database and language vendor Marxmeier Software. Other companies contributed to cover expenses, but the largest share of the organizing always went to Yeo.

In 2011, he and Marxmeier teamed up with some help from the NewsWire to mount the HP3000 Reunion, a multiple-day event with a meeting at the Computer History Museum. In addition to seminars and a group tour of the exhibits, a catered dinner, a briefing on the upcoming 3000 emulator, and a meeting of enterprise resource planning software users made for a busy weekend with dozens of community members.

Alan_Yeo_at_Reunion
Yeo was pragmatic while keeping his lights on for every software customer who’d invested in his products. Marxmeier Software has taken over support and services for ScreenJet Ltd. in the wake of Yeo’s death. ScreenJet and Marxmeier Software have had close ties for a long time. Yeo was a valued board member for Marxmeier Software and Michael Marxmeier is a director at ScreenJet.

To ensure the continuation of ScreenJet products and services, as of June 2019 support, license renewals and upgrades have been administered by Marxmeier Software. "This will not affect any ScreenJet customer product licenses or agreements which will remain with ScreenJet Ltd," said Marxmeier. "The teams at ScreenJet and Marxmeier will combine their long time experience and resources to guarantee efficient and reliable ongoing support and services."

Alan Yeo with dogsWith his beloved dogs at his Gloucester home

Ever-prepared, Yeo worked out the details of a smooth transfer over the months when his cancer recovery had failed. He'd rallied after treatments and recovered enough to race vintage cars on rural road rallies in 2018. In his last months the disease progressed to cut off motor functions of one arm. He resolutely typed long messages one-handed.

Failures were always a topic he could approach with candor as well as compassion. “Most software on the HP 3000 was too expensive, compared with other platforms,” he said in a 2004 interview examining the collapse of HP’s ecosystem. “However, because people could reliably write applications for the system, many of these were developed far too cheaply. Many customers got far too much for the money they actually spent.”

A reach for personal connections

The ScreenJet product was a recovery from a valiant effort to make the 3000 a vital part of the dot-com PC world. Millware was to deliver software that gave 3000 customers a way to make their VPlus interfaces behave like modern graphical interfaces. The software was to be free in exchange for giving over some of the screen real estate to messages from vendors. Before that user base could be established, dot-com computing staggered, a blow to the vendor element of the formula.

Yeo also picked up the pieces from the effort to market ScreenJet, developed as a connectivity product and sold by Millware.com until that marketing company went bust during the dot-com implosion. ScreenJet earned an award for migration solutions from Acucorp. But for all of his effort toward helping migration customers, Yeo was clear-eyed about 3000 transitions. ScreenJet achieved its best technical release just one month before HP announced its withdrawal from the 3000 market — and the product’s development up to that point was not driven by any need to move companies away from the platform.

Yeo also took a role as producer in a new feature for 3000 customers long abandoned by HP: Transact users. The advanced development language was kicked to HP’s curb in the middle 1990s, but sites continued to run extensive Transact applications, long after the “strategic” badge fell off the language. The TransAction software from his team give Transact sites service and tools to move programs to COBOL, a way to prepare for the journey away from the HP 3000.

Marxmeier, who reached out to break the news about Yeo's death, said he would miss his ally's organizational gifts, but even more so, Yeo's ability to write and speak with, well, eloquence. After drafting a heartfelt letter to inform the ScreenJet customers about the founders' demise, Marxmeier said he already felt a gap in the story. "It's something I would have liked Alan to read, before I released it," he said.

Yeo said he wanted no florid speeches of eulogy at his passing. Months before he died, he said if there was any afterlife at all, "I could be a little sprite, one who could occassionally make it rain on somebody who was being pompous, that would do me quite nicely." It's fair to say his narrative for the 3000's transition era was rich with the words that rained on misfortune and miscalculation.

Carly_cartoon_dalmations

Posted by Ron Seybold at 07:15 PM in History, Homesteading, Migration, Newsmakers | Permalink | Comments (0)

October 14, 2019

How to make databases live past shutdown

Index card file drawer
Photo by Maksym Kaharlytskyi on Unsplash

In 2011, a systems manager for the power utility at the City of Anchorage was looking toward a shutdown of the municipal HP 3000. It's a situation that surfaces from time to time even now. Back in 2011, the manager could see another 10 years of useful service for the 3000. His management had other ideas. This might sound familiar.

Wayne Johnson said, "We have an HP 3000 that we are going to decommission, sadly. I have powers that be who want it turned off sometime next year, although I think it will be longer than that. Is there a service that will read DLT IV tapes or convert them to some other usable format on a Windows platform or some Unix server?"

He went on to say that most of his data files were TurboImage database files. They were for archival purposes only. "Of course, the simple solution is to run the HP 3000 N-Class, probably for the next 10 years. It never goes down. But that call is not mine to make. They want to unplug it."

Alan Yeo of ScreenJet supplied a database tape migration solution that still works today.

"One very simple but elegant solution is to get a copy of Marxmeier's Eloquence database which is very inexpensive for your choice of Windows, Linux, or HP-UX and just load the databases in. Then either with the Query3K tool or with ODBC, you can just access the data as and when required.

"You could copy the volume sets to Network Attached Storage. I'll make a bet that the smallest NAS device you can buy for less than $400 will comfortably store more data than you managed to create on the HP 3000 in its lifetime.

"Allegro has a product, Rosetta Store, that will directly load Eloquence from databases in STORE format on tape, if you want to skip the step of restoring from the 3000 tapes and then unloading for import into Eloquence. I think the Allegro product will also do flat file conversion."

Beyond the Marxmeier and Allegro software, there was another suggestion offered in 2011 about a product that has come to change the way MPE databases live on beyond hardware shutdown. HP's iron, after all, isn't the final resting place for 3000 applications and data.

Read "How to make databases live past shutdown" in full

Posted by Ron Seybold at 11:41 AM in Homesteading, Migration | Permalink | Comments (0)

October 10, 2019

Who's to blame when the lights go out?

Power-lines-towers
Photo by Peddi Sai hrithik on Unsplash

Yesterday the lights didn't come on in Northern California. Everywhere, it seems, because the Pacific Gas & Electric corporation didn't want to be sued for windstorm damages to its power lines. They cut the juice to prevent lawsuits. Tesla owners got a dashboard warning.

The surprise about the outage was as complete as the shock over Interex dowsing its lights overnight in 2005. Except the cynics could see the PG&E blackout coming.

Solar panel-owning residents of California and electric car owners were most surprised. I went to a 3000 tech mailing list to look for people worried about topping up their Teslas, because some people who picked 3000s are pioneers, so Teslas are well represented among MPE veterans. Like the usual chaff on a mailing list, there were turds of political opinion floating there about who's to blame for California's darkness.

So I wasn't surprised to see more attacks on the state of California. "A third world country" is the shorthand smear, although you can say lots of the US isn't a first world country any longer. In the exchange on the mailing list, it was apparently too much trouble to keep a state’s government separate from talk about Pacific Gas & Electric’s corporate moves. Once PG&E goes bankrupt, then the private corporation’s demise will be blamed on California voters, using that logic. It’s easier than keeping commerce and government separate, I suppose. 

Blaming the tough regulations about state rate hikes for the disaster that is PG&E business is having it both ways: Government is crucial, and government is ridiculous. On and on it goes, until we are supposed to trust a government that lets PG&E do whatever it wants, so long as profits stay high. 
 
Because every corporation with ample profits has always taken care of its customers in every need. 
 
Some people on that mailing list sure have a short memory about such nonsense. We are all survivors of a meltdown of a business model where corporate profits were ensured — because revenue growth was the only thing that mattered — while legacy technology got scrapped. Millions of dollars of investments, the fate of hundreds of vendors, and thousands of careers were lost.
 
The mailing list name still has the numerals 3000 in it. You’d think people would remember what brought us into each others' lives, along with the lesson we learned the hard way together. Oversight is important. The problem which hit the Hewlett-Packard 3000 customers was a lack of oversight from top-level management and the board of directors. It's sometimes hard to know what to do while things are changing (the computer business) and ambitions are high (make HP bigger than anybody, so it will win every deal).
 
A good rule to follow, though, is like a physician. First, do no harm. The 3000 community got treated by HP like a limb that had gone gangrenous. Old history that'll never be changed, yes. Also, a lesson for managers on how to treat older bodies like an operating system and software that's not new but is still performing well.  
 
Complaining about oversight, when you'd rather have none at all, is what got HP into the state it's in today. Two corporations, neither growing, both unable to honor the promises of forever-computing that drove companies to buy its products. HP's cut itself loose from the future of OpenVMS, and the thousands of companies that rely on that legacy OS need to trust VMS Software Inc., new owners of the OS's future.
 
It's a better deal than the one HP gave its 3000 customers. Private money would've taken over MPE futures in 2002. HP wouldn't sell or license it, but again, that's just history. Now that the lights aren't going on for the 3000 at HP anymore — so many of HP's 3000 web pages are dead or buried alive — it evokes the powerless situation in California.

Read "Who's to blame when the lights go out?" in full

Posted by Ron Seybold at 11:23 AM in History, Homesteading, News Outta HP | Permalink | Comments (0)

October 08, 2019

Debugging the diagnostics

Fire-ant
Photo by Mikhail Vasilyev on Unsplash

The Command Support Tools Manager (CSTM) replaced SYSDIAG as of MPE/iX 6.5. Managers who are keeping MPE/iX working here in 2019 rely on CSTM, just as they did SYSDIAG before it.

There's evidence out there that CSTM has problems while running on 6.5 MPE/iX systems. One well-schooled developer recently noted while trying to run CSTM on his MPE/iX system that the diagnostic told him on startup, "an error dialog could not be built to display an error."

The developer community suggested a few fixes for this problem with the diagnostic software. CSTM was ported onto the HP 3000 from HP-UX, so the repairs that CSTM itself suggested regarding memory (increasing it, removing processes, reconfiguring kernel memory limits) probably don't fit.  CSTM has a special page in the Hewlett-Packard Enterprise website devoted to the problem.

The developer at least had another 3000 running the same version of MPE/iX, a system where CSTM was starting up without a problem. One bit of advice suggests that while using console debug, "check out what a your working system looks like at the CSTM prompt when idle. Use psuedomap “XL” to get symbols from the libraries and program. Attempt to set some breakpoints near initial program launch."

Using DEBUG, the open heart surgery of HP 3000 management, is sometimes a required diagnosis. When your diagnostics software requires diagnosis, nothing but DEBUG will get the job done.

Much more detail followed on using DEBUG to discover what's failing in CSTM.

Read "Debugging the diagnostics" in full

Posted by Ron Seybold at 04:50 PM in Hidden Value, Homesteading | Permalink | Comments (0)

October 01, 2019

Shrink ray 3000 services: what you'd pay

Lens-shrink effect

Photo by Stephen Kraakmo on Unsplash

The number of MPE and HP 3000 experts is declining. It could hardly go in any other direction but downward, given the age of the expertise. There's still a number of companies — no one is sure how many — using the servers and wondering how they'll get along when something goes wrong.

One solution that's been successful up to now is shrinking the footprint of resources needed for using MPE/iX. Rather than each customer using up environmental conditioning and physical space for a server, owners of 3000s can have their systems hosted in a centralized location. It's co-location, but offered by companies which have MPE/iX and 3000 experience. The latter is most important because the components in an HP server are specialized. 

Good answers for hardware issues are the prize in a shrink ray hosted offering. Browsing the postings on the 3000-L newsgroup this month, I'm struck by the number of questions that are not only specific to MPE, but focused on component problems. Sending a 3000 off to a co-located datacenter has been offered for many years by now. The Support Group, an Austin-area firm for helping MANMAN owners, built a disaster-proof datacenter on its site that houses 3000s from customers.

There are others in this market who do the same service for 3000 owners. Beechglen Development has services that will harbor a 3000 and take the computer out of the everyday management stream for participating companies. Solutions for reducing the 3000's footprint to zero, while keeping MPE apps at work, use the shrink ray effect.

Less easier to measure: what such shrink ray services should cost, or what the remaining 3000 owners would be willing to pay. It's far better to imagine the cost of that fading HP iron becoming unresponsive, as they like to say when you're holding the line on life at an advancing age. Good resuscitation can be priceless; that's why people move into continuum of care facilities in their most golden of years.

A good friend has moved into one of the best independent living facilities in Chicago. When she had a heart scare this summer, though, she was able to get to a hospital through the help of her friend. Returning to her apartment the next day, she checked in with the facility director to see if the building's staff might look in on her that night. That's assisted-living, she was told, not independent living.

Some of the 3000 hardware still in production is too old for independent living. Shrinking it before sending it to assisted living is a good first step. Reducing a footprint, by shipping it away to a support company with a disaster-proof datacenter, is the shrink ray magic that can keep MPE alive for years to come.

Posted by Ron Seybold at 08:36 PM in Homesteading | Permalink | Comments (0)

September 19, 2019

Making a place for retired 3000s

Owners of HP 3000s are facing the end for their HP hardware. The MPE/iX software has a longer lifespan than the components that have carried it. Even in places where the apps will live on, the hardware is deteriorating. What to do with the aging iron is a question coming up more often.

The HP hardware isn't disposed of easily. It's got the same kind of environmental hazards as every other computer: rare minerals are the prize in there, but there's lots of the weight of a 3000 system that's just going to be classified as scrap.

In any conversation with an owner of a 3000, the solutions to this issue revolve around a reseller-broker. These third party companies have made a business of moving 3000s in and out of datacenters. Lately the movement has been almost exclusively outward.

In the reports from the field we've heard, used hardware often has little value unless it's from the latest generation of 3000s. There are individual items that still will return some dollars to the sellers. K-Class MFIO boards have become rare, and since those components prop up the older 9x9 servers, the boards can carry value that might be equal to a complete system of the A and N-Class generation. 

Used hardware has always been a marketplace with great malleability in its value. It's been a lot like being a coin collector for 3000 owners. The valuations might say your 969 should be worth $500, but you'll only get that from a buyer who can sell your coin for more — or one who needs the hardware enough to deliver that price.

Posted by Ron Seybold at 07:22 PM in Homesteading | Permalink | Comments (0)

September 12, 2019

What was done to HP 3000s for good in 1990

Frozen-waterfallPhoto by Vincent Guth on Unsplash

One week when two of the 3000 community's greatest icons connected with me, it drew my attention back to the start of the 1990s. To say that decade was a very different time for the HP 3000 simplifies a much richer story. What's more, there are parts of that decade's accomplishments that continue to serve the community to this day, for those customers who rely on the frozen nature of MPE/iX.

The year 1990 was galvanizing for the 3000 community. I was reminded about the year when Adager's Alfredo Rego asked on the HP 3000 newsgroup, "What were you doing in 1990?" In a brief message, he noted that 1990 was the launch date for the world's first Internet browser, created by Tim Berners-Lee on a NeXT workstation. Rego pointed at a history page from 1990 about the start of the browser era. Then Rego noted

Enjoy it (typos and all).  Be sure to click on the links to the screen shots. Ah... Memories. Fortunately, the NeXT ideas have survived (and thrived).  Just as MPE ideas have (not). Sigh.

But 1990 was a high-water mark in HP 3000 advocacy, a habit which works today to survive those three decades. The HP 3000 users formed a community in way no other computer can claim, led by Wirt Atmar, founder of report solution provider AICS Research, creators of QueryCalc as well as QCReports and the free QCTerm.

Atmar knew better than most about advocacy, for in the fall of 1990 he helped spark a charge that changed HP's business practices about the 3000 — changes which you might argue lasted until the vendor stepped away for good. Especially for the MPE users who have changed little about their HP 3000 stable environment.

Read "What was done to HP 3000s for good in 1990" in full

Posted by Ron Seybold at 02:36 PM in History | Permalink | Comments (0)

September 10, 2019

Relative performance online as 3000 history

Snapshot of partial HP relative performance
As the HP hardware to run MPE/iX ages, it's on the recycling and scrapping block for companies that still have an HP 3000 box on-premise. Now hardware is so cheap you can throw 3000 gear away.

The slow, old, and heavy boxes go first, of course. I remember taking a trip with Stan Sieler in the Bay Area where he took me to a scrap facility. There, shrink wrapped on the outside of a pallette, were HP 2645 terminals, right alongside Compaq boxes.

Relative performance charts can be our friend as we triage our older HP gear. There's an adequate one available online at bitsavers.org as part of a breezy page covering the history of the 3000. 

We've got The One Chart to Rule Them All you can download to use while you have HP's gear on the chopping block. There's a section for A-Class comparisons, and another compares HP's boxes in the N-Class line to older system performance.

Such numbers are relative in more ways than just the comparison between servers. HP actually massaged the numbers themselves back in the late 1990s. Our story in 1998 reported that 

HP is “restating” the performance rankings for much of its hardware, starting with this month’s rollout of the Series 989 systems. The new rating is an HP 3000 Performance Unit, not based on Series 918 performance. And the new numbers are between 29 and 52 percent higher for all systems except HP’s largest ones, the Series 996 and 997 units.

As I observed, while looking askance at the new figures, "HP wants you to think of HP 3000s as faster than ever, but its new rating measurements don’t really make existing systems any faster. They just sport higher numbers than they did last month."

There was some technical logic to the HP adjustment. The 3000 hardware from HP had just acquired some newer and faster cousins.

Dave Snow, product manager for the 3000, said "the measuring techniques for our midrange and high-end platforms were producing results that were not consistent with each other. You had a 918 performance for the midrange and a different relative performance for the high end, but the two relative performance numbers weren’t the same.”

The discrepancy was a big deal, he added, “but it was a big deal we could sort of live with, so long as the 9x9 and 99x performances were dramatically different from each other,” Snow said. “As we added performance to the 9x9 platform, it is approaching the 99x. That’s caused us to have this quandary. In some sense we’ve had two different sets of 918 numbers. We had to bite the bullet and reconcile the numbers."

Posted by Ron Seybold at 01:47 PM in Homesteading, News Outta HP | Permalink | Comments (0)

September 05, 2019

TBT: The Flying HP 3000

3000 Crash Test
Twenty-two years ago this month, HP thought enough of the 3000 to send it flying off a three-story rooftop. It was called the HP 3000 Crash Test. The demonstration was more like the tests conducted with safety dummies than anything from a software lab.

HP spent some of that year celebrating the 25th Birthday of the 3000 with fun stunts like this. The rooftop trip was called a skydiving event. Alas, no parachute.

A dazzling disco evening played out in Stuttgart during the same month as the Crash Test. The Europarty was held not far from the Hewlett-Packard manufacturing facility in Boeblingen. That soiree featured a saxophone player riding on a zip-line. Different times then — but maybe the 3000 was ahead of its time with a zip-line at a party.

The Crash Test was similar in its mission to make us smile. It also proved a point about the hardware that people can't seem to get rid of by now -- the boxes were built to withstand remarkable abuse. For example, Joe Dolliver told us about another Lazurus-like performance of HP's gear.

Back at Amisys in a previous life, Bud Williams sent an HP3000/957 to the Amisys Dubuque programmer office back in September of 1999. The system was there for Y2K issues testing for the staff in Dubuque. It was sent via North American Van Lines.

As the story goes, the system got crushed by another heavy skid of material and the 3000 looked like Gumby with broken sides and smashed connectors. Another 3000 expert, John Schick, got the box in place and the system ran fine. Yet another story of the HP 3000 taking a licking and still ticking.

The last line is a reference to a TV ad for Timex watches, a reference too obscure for anyone who's in charge of a datacenter today. The Crash Test lives on as a movie on the Newswire's YouTube page. When we started all of this, just about 24 years ago, YouTube was just a magic act in the mind of some wizard working for what would become Google. Instead, HP distributed the movie via VHS cassettes: perhaps another reference too old for the junior programmers on staff now, working on their virtual servers in the Amazon AWS cloud.

Read "TBT: The Flying HP 3000" in full

Posted by Ron Seybold at 01:00 PM in History, Homesteading | Permalink | Comments (1)

September 03, 2019

ERP Tips: Using work orders to backflush

Pipe-and-plumbingPhoto by Samuel Sianipar on Unsplash

MANMAN still runs operations at companies around the world. Not a lot of companies, of course. It's 2019 and everything is smaller in size, not just your hearing aids. The MANMAN managers are still looking for tips. Here's one generated from a question out of the Altra Industrial Motion Corp. from senior systems analyst James English.

We are on MANMAN version 9.1 on an HP 3000. We have all MANMAN modules, including MANMAN/Repetitive. Is it possible to backflush work orders without using Repetitive? Our one manufacturing location is looking at simplifying work order transactions. They are manually transacting each operation on their work orders, even though they don’t collect actual hours.

Short question: How can they use work orders instead of using Repetitive?

When a work order has been received into stock, it comes to the scheduler-planner to push the times through each sequence, since the operation no longer does time cards. Once that time-pushing is done, the work orders are closed for material and labor. Once a work order is received into FG, instead of pushing the time through each operation, could we just back flush?

Alice West of Aware Consulting says

You can set all the components on your bill as “consumable” and then when you complete the WO the system will consume all the materials.  We always called this feature “poor man’s Repetitive.” 

However, it sounds like you are trying to simplify the labor portion of the transaction.  For that, you can look at your COMIN variable settings. Here is a chart I put together to show how 3 different variables work together.

Read "ERP Tips: Using work orders to backflush" in full

Posted by Ron Seybold at 10:52 AM in Hidden Value, Homesteading | Permalink | Comments (0)

August 29, 2019

Hurricane season was a hit with a 3000 show

Andrew_23_aug_1992

It's late August and the hurricanes have begun to march toward Florida and the Gulf of Mexico. Many of the annual 3000 national conferences -- okay, we had to call them Interex, but those shows were about the 3000's heartbeat, volunteers and vendors -- took place during this month. Several tempted the weather gods, being scheduled in places like Orlando. Only one, though, found itself astride the path of a Category 4 hurricane.

The luck of Interex luck ran aground when Hurricane Andrew made its landfall during the week that HP planned to celebrate the 3000's 20th anniversary. The storm came ashore near New Orleans, where that 3000's birthday party was scheduled. I was reporting from Interex for the last time as editor of the HP Chronicle.

It was a week when the company's that getting a new CEO, Lew Platt, who was on the cusp of making his debut at a keynote in front of 3,500 customers at Interex '92. Platt was only the second man ever to be elected to the top job at HP. Up to that point, its founders both took turns as CEO. The next executive to hold the job after Hewlett and Packard was John Young, who didn't have an engineer's roots like his predecessors. Platt's arrival was touted as a return to HP's technical leadership. He was an HP insider who was a technologist, proud of his roots — and humble enough to have a habit of eating his meals in the HP cafeteria.

The outgoing Young had been scheduled to deliver a keynote to the Interex conference, but Hurricane Andrew changed those plans. The storm had just ravaged the Florida coastline with Cat 4 winds the day before Young was supposed to appear. His assignment was transferred to Platt, although the leadership of HP wasn't going to pass on to Platt until November.

The severity of Andrew set even the CEO-designate into flight from the show.

In the plaza in front of the Hilton Riverside Towers, Platt was trying to make his way to a running limo that would get him to the airport before all flights were grounded. But one customer after another wanted just a moment of his time on the way. After a handful of delays, his wife Joan insisted on his safety. "Lew, get in here," she shouted from the limo. One of the company's most grassroots leaders had to depart before his debut in in storm-lashed show week.

The second generation of the PA-RISC chipset for 3000s did remain at the show. The Series 987 servers were also making their debut that week. HP pushed the message that MPE/iX was an easy porting destination for applications on the move away from Unix, pointing out that General Mills had moved a third-party warehouse app from Unix to the 3000.

"It had been generally accepted that it was much easier just to buy a new platform for the application," HP's Warren Weston wrote in the HP Chronicle. "However, after further investigation, the decision at General Mills was made to port to MPE/iX." It might have been the last time the vendor promoted the 3000 over Unix in a public message.

Read "Hurricane season was a hit with a 3000 show" in full

Posted by Ron Seybold at 06:50 PM in History | Permalink | Comments (1)

Search