November 30, 2006
HP responds to insider trading lawsuit
The Associated Press is reporting this morning that HP is being accused of insider trading. The claim was added to a lawsuit filed in Santa Clara County, California when the company was accused of spying on reporters, board members and others.
In an amendment to that lawsuit, investors allege that the company CEO Mark Hurd CEO Mark and seven other executives got rich unfairly by selling HP stock worth $41.3 million during the two weeks prior to the spying scandal becoming public in early September. Those two weeks saw more insider HP stock traded than at any other time in five years, according to the suit. Most of the sales involved cashing out stock options.
HP responded to the suit by saying that it had no merit. It's a typical early-stage response from any company that's been sued. The trouble in the lawsuit seems to be the financial motive behind the insider sales. HP's stock price is now higher than it was in mid-August, two weeks before HP admitted to violating privacy.
If those executive options got sold in August instead of November, the top HP insiders didn't get the best price they could've had during 2006. HP shares were above $36 before the spying news broke. Today they're above $39 a share, and even topped $40 since the spying accusations.
The new lawsuit action also consolidated other suits in California over the media-leak probe at HP. The probes, conducted over the past two years, led to criminal charges against former Chairwoman Patricia Dunn, former ethics chief Kevin Hunsaker and three private detectives - Ronald DeLia, Matthew DePante and Bryan Wagner.
None of this has dragged down HP's shares, however. With a Dow roaring above 12,000 and the company's Q4 and year-end reports looking rosy, HP stock has been climbing since mid-summer. The lawsuit asks for unspecified damages, as well as alleging that HP wasted corporate assets and initiated a $6 billion stock buyback in August “to prevent a free-fall in the Company’s stock price as the news of the defendants’ misconduct reached the market.”
HP's response said the suit “represents a transparent effort to exploit issues related to HP’s recent investigation for personal gain at the expense of HP, its shareholders and its employees. HP will defend itself vigorously.”
November 29, 2006
Who, not how many, matters
Owners and managers of HP 3000 might be prone to measure the vitality of their world by numbers. What percentage of the customer base is still running the 3000 in a production environment, for example. Or how many people made the trip to the Gulf Coast this month to take in all the learning at the Greater Houston RUG's HP 3000 Conference.
The latter was the first of its kind in three years, perhaps even six You have to go back to the Interex e3000 Solutions Symposium to find the last 3000-dominant conference. Yes, there might have been only 50 attendees at this year's conference. We hear early reports that the conference will be mounted next year, too, at the same University of Houston Clear Lake venue, on the weekend of Sept. 14. You can bet on the RUG's board having more expertise and energy behind marketing the next show.
Nobody can fault the RUG for its speaker lineup, though: Birket Foster, Alfredo Rego, Bill Hassell, Gilles Schipper, Paul Edwards, even Jeff Vance of HP, doing an impromptu and excellent review of an updated paper on the 3000's Command Interpreter.
There were more, like Charles Finley showing how the transformation of a COBOL app like a Time Entry program can be moved to another platform, it interface improved and its flexibility extended. Finley, whose Transformix company works closely with Sector 7, showed us a solution that had J2EE applets, servlets, Java Beans and the Jakarta Tomcat Web application Server.
Finley's solution relies on the interface of Java Swing client, although he also showed Web client for the application. The major revelation: Oracle's JDeveloper Swing Application and JSF Application elements, priced within reason and relevant to moving a 3000 app.
Presenters like these did not attract the massive numbers of conferences of the 1990s. But even with a marketing machine such as Interex had at its disposal, the Solutions Symposiums never drew more than 200 attendees. What mattered then still matters now: who is presenting.
So long as the top-echelon gurus from the 3000 community remain willing to teach and lead at such conferences and up on the Web, a customer can feel confident about their transition clock. There's been talk in the community of graph that describes the rate of decline in the number of 3000 suppliers and customers.
There were a lot fewer T-Rex dinosaurs on the planet than any other species in the days when dinosaurs roamed the earth. Numbers didn't matter then, either. A T-Rex dominated the food chain then, as much as these experts offer the dominant training and insight on using, maintaining and transitioning from the 3000.
The number of players at a poker table in the latter stages of a game doesn't matter, either. These are the best of the rest. You could say the same thing of the customers still relying on their 3000s, those taking needed time to make a transition, and the vendors and service providers helping out where needed.
November 28, 2006
Houston meeting spreads OS training
Organizers of the recent Greater Houston HP 3000 conference want users to know: It's not just a 3000 conference. Found teaching Friday through Sunday: Bill Hassell, HP Unix guru extraordinaire.
''GUIs are for wimps," he told the collected audience of about 20 in one of his four sessions. He's a command-line man himself.
Which, if you consider it, makes Unix a comfy spot for the old-time 3000 veterans to land. After all, prowess with the command line was pretty much the bulk of the old 3000 experience. Even the new 3000 experience, aside from things like GUI/3000.
At the Houston show, Adager's Alfredo Rego gave an agnostic talk in the 40 minutes carved out for his keynote. He drew on an large inflated ball to show the audience where you could go one unit south, then one unit east, then one unit north, and end up at the same place. The North Pole, of course, but Alfredo's lesson — which he demonstrated with the help of a Mac which was also running Windows — "don't necessarily accept the obvious answer."
"The obvious way is not so good. the non-obvious way is better," he said. Using his read-between-the-lines method of study, I read that to mean that manipulating the schema in IMAGE/SQL is the obvious way of trying to deliver Large File datasets. Work at the root level, less obvious, would have been much better, I have heard.
You might even say that sticking with an HP 3000 is the non-obvious way to run an IT operation. Alfredo didn't say that. And he was generous in his praise of the hard work HP has done to make Large Files operate correctly, without corruptions. And just as pleased that HP chose to bury its work, in the interests of not forcing change in programs like QueryCalc, Adager, Suprtool and DBGeneral.
Then there was Birket Foster, recounting for everybody all the things you'd better plan for whether you're staying on your 3000 (it's called a Sustainability Study, and you ought to have one) or making the jump off to another platform. After one night when Foster led attendees to the suite hotel near the Clear Lake UH campus for some impromptu BBQ, he was part of a party that headed off to the Johnson Space Center for a tour of rocket chassis, training rooms and Mission Control.
Not just a 3000 show? You betcha. At lunch on Day One, IBM bought time to show off the merits of its latest Series i servers, the integrated solution that resembles an HP 3000's value proposition. It's not the first time IBM has sent a missionary to Houston to talk to HP customers. Each time the presenter has been polished and funny.
Gilles Schipper's favorite MPE/iX utilities? Well, there's DIRK, which lets an IT manager check out open trap doors where intruders could access the 3000 without a login and password.
Attendees and exhibitors and consultants numbered about 50 in total at the conference. But the caliber of who was on hand — Paul Edwards giving advice on homesteading, Michael Marxmeier and Alan Yeo telling of migration bulwarks like the Eloquence database and ScreenJet, used live to show a migration happening — the speaker roster says a lot about the future of this conference. Next year it will have even more, leaving not fewer 3000 slots like Interex did in its later HP Worlds, but just more to learn alongside the essentials of the HP 3000 experience.
As a reunion and re-education of 3000 wizards, it was a great weekend. Next year's version, in September of 2007, will have more on hand of everything, including solution diversity and attendees.
November 27, 2006
What a certification is worth
A computer with a dodgy future shouldn't have this much customer interest in certification for its skills. But the HP 3000 is unlike a lot of systems abandoned by their makers. People still want to prove their 3000 skills are up-to-date. The difference? The HP community, still providing training and cert service long after HP's lost interest. When people will pay for this training isn't as interesting as the fact that it's offered at all.
Inside the folders passed out at this month's GHRUG HP 3000 conference (there were no bags, another cost-saving measure other conferences ought to observe), I saw a color flyer from Jon Diercks, who said that he was "sorry I couldn't be there to participate, but wanted to send my regards." Under Diercks' smiling face was an e-mail and Web link to garner contacts from your community. Like a lot of 3000 pros from the modern era, Diercks is able to take work from either a homesteading or migrating customer.
But he's important to the certification process of the 3000, the way professionals will try to demonstrate they're able enough to administer, program for or size up a 3000 for either homesteading or migration. Diercks wrote The HP MPE/iX System Administration Handbook a few years ago. You literally could use the book to study for a certification test. As a 3000 resource, the book remains the only title close to up to date with the operating system.
Why in the world would you want to pass a certification exam for a computer being abandoned by its vendor? The question sparked a lively debate not so long ago on the 3000 newsgroup. Good points were being made by both sides, even though a peristent gravedigger for the system said long ago all is lost for the 3000 customer.
"If I had to pick the biggest story of 2005 it would be HP’s killing OpenMPE, the HP 3000 and its support “ecosystem” of vendors that were looking to take over support at the end of this year," said an advocate who's now embracing Windows. "By extending HP’s support of the HP 3000 for two more years, they’ve not only burnt the HP 3000 and its community to the ground, they’ve eaten the ashes as well."
"If you didn’t think that the HP3000 was dead before, you’ve got to believe it now."
Like a lot of predictions delivered in advance of the data, this doesn't sound like the 2006 that followed those comments. Third parties like Paul Edwards and Associates, Pivital Solutions and independents like Diercks still want to help you homestead, even it's only for awhile until you move away. Dead? Depends on your definition of alive. How's the health of OpenVMS, currently trailing in all of the HP virtualization offerings behind Windows, Linux and HP-UX.
Then there's Jim Chance's "how can anybody see any value in this, without exception" point of view:
I a 22-year IT guy with 16 years HP 3000 and personally I don’t know why anyone would want to pursue or care about this particular cert. Just my opinion. As much as I hate it, the HP 3000 market, shops, places with it fully running in production, conferences/workshops, new sales, new ERP on that platform, and technical knowledge is DEAD or quickly DYING. When will folks realize this?
Turnover to other platforms and skills is at an all time high. Universities and corporations are de-commissioning there boxes at an astounding rate. How do I know this, by being a contractor since 1997. I could go on and on about this trend.
I asked......who was it, I forget, but some vendor who puts on training to become MPE cert isn’t getting any paying students, none. So why does the board or anyone else lobby for something that frankly — and realistically, just plain out isn’t in demand? I appreciate efforts, but really? I can think of a dozen former and/or dwindling HP3k guys who are now pursuing MS cert’s; there is no way they’d want to spend $ or time on certifying as MPE.
Donna Garverick-Hofmeister, an OpenMPE board director, made a case :
There is value in MPE certification. For migrations are taking longer than expected, for companies who plan to stay on MPE long term -- they’re going to need people who know MPE They’re going to need people who know how to straighten out problems left behind or situations that have cropped up due to infrastructure changes.
MB Foster founder Birket Foster calls this the "flight attendants flying the plane" situation, commonplace in 3000 shops. The trained folks have been downsized or moved on of their own volition. Who ya gonna call? Maybe someone certified.
Tracy Pierce wants to know what any certificate is really worth, MPE or Unix or whatever:
What’s the certificate’s real value?
As to an assurance of qualification, I bet someone holding said certificate would probably get consideration equal to that of a non-holder if they claimed competence in all the requisites mentioned.
While the certificate is ostensibly an attempt to certify competence, its existence can be quite misleading. I bet (I’d say I know but then you’d want examples) that there are people who hold the certificate who don’t know half as much about the machine, much less the project at hand, as do some other people who’ve never set foot inside an HP training class much less taken an HP certification test.
The certificate indicates that its holder passed a test, not that they can program their way out of a paper bag or can apply common sense to a business problem, much less unravel code written by possibly competent programmers who didn’t have time for documentation. Does the certificate imply knowledge of every gotcha to be found in VPlus?
Said tests include questions about a lot of pretty arcane stuff, just the sort of knowledge that leaves a person pretty quickly if not exercised. How long has it been since the certificate became unavailable? Are all those certified actively working with the platform and using all the skills for which they’re certified? Even if so, does that make them More competent than a non-holder? Not in my book.
If I had one, I’d probably mention “HP Certified” on my resume. But I don’t think I’d really want to work on the project if the employer’s more impressed by a certificate than by a demonstration of the real skills actually needed.
On the other hand, a certificate will make the bearer’s suit and shoes look much shinier to an HP 3000-ignorant entity offering a conversion contract; it’s also a probable indicator that the holder knows enough to hire competent staff for a project involving the HP 3000.
Hofmeister said that a more up-to-date test would prove more — a project Edwards and his colleague Frank Alden Smith are ready to take up.
Many of us have heard the complaints about the existing MPE test. Personally, I'd love for the test to be revamped. Rewrite it for 7.5, make it good evaluation of a person’s MPE knowledge, make it so the test means something. If wishes were horses, I’d like for there to be a programmer’s test in addition to an administrators.
Having said that — nothing is going to replace an interview for hiring. Piece of paper or not, the only way to know if someone is going to fit into your organization is to talk to them.
But getting on as a 3000 expert, after you've stepped away from the system for awhile, is going to take some retraining.
It’s hard to tell what is worthless and what is not. I was a system operator for an HP 3000 Series 70 with MPE V back in 1990 when major restructuring meant job losses to my entire department and beyond. It’s rather eerie going back to that building because the office/warehouse is gone and now is subdivided between several different companies.
I came here because I once again am seeking employment and am wondering — is my past experience obsolete? Can I build on my knowledge or would today’s MPE/iX 7.5 systems ( be too far removed, and I’d have to start over?
How could it hurt to have a copy of Diercks' book? At the moment, Amazon has an $11 used copy.
November 24, 2006
We're making history here
I keep hearing this in postings about the HP 3000's future:
"The growth or decline of a specific computer platform is directly akin to any other biological population."
By now, I'm wondering. What studies have been completed to confirm this theory? Come to think of it, what are the "akin" elements between computer platforms and biological populations?
See, I am a realist. Someday all the HP 3000s will be museum pieces. But there's the meantime to consider, a meantime many of you live and work in today, and for the forseeable future.
I got a call this morning from Vladimir Volokh, who'd read my November editorial about redemption. He invited me to point out that 2027 was the end of the HP 3000's 7-bit date-keeping technology; but he added that many of the customers he visits don't even know about 2027 as a CALENDAR intrinsic deadline. Too many of them don't even know HP extended its basic support to 2008. I encountered one of the latter customers myself at the HP 3000 conference in Houston two weeks ago.
That moment in Houston was an eye-opener for both me and the customer. How could you measure the conference value to a attendee who learns that his end date for his company's mission-critical platform is not the end of this year, but the end of 2008? Think of the way an IT budget could shift (like, now there's time to engage experts for planning, and maybe move our home-grown app to another platform.). Think about how the shift in IT plans might affect his company's profitability.
Sometimes it feels that this computer community is getting a bum's rush to the door Nothing lasts forever. But need we remind everyone that HP predicted the bitter end for the majority of the communit to be about six weeks from now?
I think a computer platform, and the humans who use it, doesn't match up so well with other "biological populations." What biological populations exist whose behavior is so deeply affected by a technical tool whose merits are measured in relation to other technical tools?
I don't believe using hammers or arrowheads count, being not complex enough to compare to a computer, which reflects the state of the human mind. No, you are making history with your decisions on when to leave the 3000. This is the biggest population of computer customers using enterprise-grade systems who've ever been told by their vendor, "you don't have to go home, but we don't think you can stay here."
We are taling about symbiosis here, influenced by the relative value proposition of other computer platforms. Frankly, in 2002 we found it foolish to be making any move off an HP 3000. Four years later, it looks like HP is putting together a good alternative, with hardware value and software virtualization, to offer a winner.
So I'm waiting to hear. What does the demise of the passenger pigeon or the buffalo or the native populace of the Great Plains in the 19th Century have to do with the HP 3000's useful lifespan? Studies? Papers? Testing?
I don't plan to be writing about the 3000 in 2027, when I'll be 70. But if there's something interesting to write about then, and anybody still alive to read it who cares, then I'll be tempted.
In the meantime, the Transition you're going through is producing some real data, research, on the demise of a population of computer users. You're making some history in showing the rest of the market how a shift from a technology takes place, and how long the process takes to complete.
November 23, 2006
Top 10 Reasons to Give Thanks
Thanks flow freely here in the US today, on a day when most of us here celebrate the arrival of pies, sleep-inducing turkey dinners and a holiday for nearly everyone after 2PM. In much of the world, though, this is a working day; Thanksgiving in November is a strictly-American holiday.
We give thanks for being a part of your lives, in the aid of IT managers making their Transitions; working alongside the HP partners who help you stay productive on your 3000s or marshal migration forces; carrying the message from HP and its competition about alternatives, in the spirit of looking into everything.
(Okay, we know: Apple has not succeeded in making its Unix solutions attractive to the enterprise customer, up to now. As an alternative to a 3000 solution, Xserve has its best chance in the education sector. Yes, Virginia, there is a 3000 presence in education, both K-12 and higher ed.)
But thanks deserve a list today, so here's ours. Have a second helping tomorrow, with Top 10 Things to be Thankful for in 2006 and 2007.
Top 10 Reasons to Be Thankful for Your 3000 Experience, circa 2006:
10. Developed a stronger sense of faith in inertia, five years and counting into migration era
9. Earned migration expertise to use now as well as later, no matter which platform you choose
8. Entered the marketplace of the $5.00 enterprise-grade HP 3000 server
7. Found vendors who draw their 3000 business plans well beyond HP's 3000 lifespan
6. Discovered new use for comfort in using the command line
5. Evaluated a replacement to outperform PA-RISC CPUs, after 11 years of waiting for Intel-HP project to mature
4. Opened the door to a vast warehouse of applications on target platforms
3. Cut support costs by two thirds while making a Transition by using third parties
2. Watched your vendor offer and then polish a multiple-operating-system technology with virtualization
1. Felt smart about your investment in a server with the greatest value in market's history — and more satisfaction about being part of the market writing a historic tale of migration and transition.
November 22, 2006
Learning about your 3000's database
I'd guess that fewer than 1,000 IT professionals are reading the 3000 mailing list and newsgroup these days. But just like at the end of a sporting event, the people left in these stands are the most knowledgeable and ardent in the game. They know what it has taken 20 years and more to learn. Today, like many days in the 3000 community, these biggest fans pass along what they know for free.
Today's subject on "the L," as it's called by its longest residents, is IMAGE, the database that made your HP 3000 famous. (And the best in the business, at least once, when Datamation ran a contest to determine which database won among contenders like Oracle, Informix, Ingres, Sybase, DB2, and a raft of others. Datamation is published no more, and some of those databases not survived. either. IMAGE, in its newer generations of TurboIMAGE and then IMAGE/SQL, rolls on.)
An IT pro posting to the mailing list was misunderstanding how Adager, the IMAGE/SQL all-purpose tool, was keeping him from corrupting his database. He asked, "When I try to change a dataset capacity in Adager, it tells me the minimum acceptable capacity is higher than I’m trying to set it at. Why?"
The answer gave the community a chance to teach IMAGE/SQL, instruction that's not only hard to purchase these days, but a great resource for a developer or somebody who maintains HP 3000 apps. A free resource, from the Web.
Wes, our confused IT pro, reported on 3000-L that Adager's dialogue was:
Present : 9763397
The minimum acceptable capacity is 9763396
The answers rolled in on the 3000-L, right away. First an overview:
I would venture to say that you are dealing with a detail dataset and you can’t reduce it without performing a reorganization of the set. There are detail records in entries above 5,500,000 and without updating all the pointers in the detail entries and the corresponding master entries, you simply can’t do that and hope to have a working database.
Adager prevented you from mangling your database.
Is it a detail dataset that you are trying to change? If so, you cannot make the capacity less than the high water mark unless you first detpack the set.
You need to pack the detail dataset to remove entries in the dataset that are marked as deleted logically, but not physically deleted, then you can reduce the capacity.
When a DBDELETE call is made to a detail dataset, in order to save processing time, the entry is logically flagged as being deleted, and any pointers to that entry are changed to bypass the entry in the chain. The entry is NOT physically deleted from the PRIV file that represents that detail dataset. By performing a detail pack, either with Adager, DBGENERAL, or DBUNLOAD/DBLOAD, the logically deleted (flagged as being deleted) entries are physically removed, leaving only the active, live entries.
When adding entries to a detail dataset, a highwater mark is kept that indicates the maximum number of entries in the dataset. For example, if I add 1,000 entries and delete 100 afterwards, my highwater mark remains at 1,000 because that is the most number of entries ever in the detail dataset. Adager, in such a case, is telling the user that you cannot change the capacity below 1,000 (the highwater), because there are physically 1,000 entries in the file (100 are logically deleted, but are physically still there).
Read can the entire thread for yourself to get a fine education on IMAGE, the database that makes your 3000 sparkle in the enterprise. Subscribe to get the messages by e-mail, if you prefer. It sure makes it easier to participate and learn.
November 21, 2006
Rest in Peace, Large File work
HP made a decision to help its HP 3000 customers recently, a choice that involved walking away from many man-months of HP 3000 engineering. Such HP 3000 engineering resource, at least from HP, is becoming more rare all the time. That rarity made HP's decision to help stamp out Large File Datasets, eliminating a concept that never worked quite as flawlessly as promised, as disappointing as it was necessary.
I recall the SIG-IMAGE meeting in 1999 where the database vendor community — those third parties who rely on IMAGE structures inside of the HP 3000 — asked for HP to stay the course. "Don't bother with Large Files in the IMAGE/SQL structure," they said; just build out the Jumbo Datasets so they can be as useful and reliable as the rest of the HP 3000.
However, HP took its counsel from large applications vendors back in those days. In modern times, the company knows it needs to be more aware of how it can help these third parties survive, carry the 3000 customer a little further. Introducing deep-seated database change at this point of the 3000's transition would be foolish. Besides the database tool providers such as Adager, Robelle and Bradmark, other companies and software providers would stall in their Unix-bound plans.
Adager's Alfredo Rego drew all these pieces and choices together in a message when HP announced it will be scrapping LFDS. "It is really sad to realize that the result of all of that effort by HP will never see the light of day," he said, "but the alternatives would have been much worse."
Rego makes an excellent point. Any tool or application vendor with an ounce of gumption has already accomplished the latest magic using an IMAGE structure well-known to developers. Fixing Large File databases — so they don't corrupt in "a rare incidence" of HP 3000 shops, as HP's Ross McDonald said — would've broken a lot of software. Rego said:
In fact, not only database tools would be affected. For instance, fast-speed report writers would be affected, as well as many other applications that take advantage of TurboIMAGE’s underlying privileged file system to bypass DBGET, DBPUT, etc.
It is noteworthy that underlying privileged data structures are NOT guaranteed (by any vendor) to remain stable. HP could have easily changed these privileged data structures to suit HP’s own purposes.
It is the responsibility (and duty) of privileged-mode programmers (whether in-house or third-party) to do whatever it takes to keep abreast whenever the vendor changes privileged material. In this case, HP chose to sacrifice the results of its own development process to avoid inflicting a painful blow to the worldwide HP 3000 community.
May LFDS rest in peace
November 20, 2006
Considering Unix as a destination
IT directors in the 3000 marketplace assign directions these days, their guidance systems working like Google Maps or Mapquest. The destination they plug into their search:
Safe and reliable business computer platform to replace our HP 3000; must meet the requirements of a mission-critical-grade asset.
But most of these IT directors are adding a more direct address in their search.
Find the best Unix
The application drives the decision, but environment must pass muster, too. They tell their staffs or analysts or migration partners doing the search to look harder at Unix. Do not include the many Windows destinations, even when they decide to replace instead of migrate their corporate applications. Don't linger in the land of Linux. The former locales still don't feel safe enough for many HP 3000 IT managers, while the latter doesn't sparkle with the corporate patina you've come to expect from being a Hewlett-Packard customer.
In our ongoing review of Unix destinations that 3000 sites are taking, we'll keep our scope broad as well as innovative. Our path includes one solution not obvious to many IT directors at 3000 shops: The road to Apple's offering. It's notable because of how much the solution has changed in the past 12 months.
Simply put, Apple now offers the lowest capital-cost corporate Unix platform. And unlike its offering of the last five years — the time the 3000 community has been in search of a new corporate destination — Apple now is dishing up Windows-capable desktops to communicate with Unix-driven servers. You can get an unlimited-user license, Xeon-CPU-based Xserve that starts at $2,999. The desktops that use Intel's Core Duo chips can cost as little as $499, plus keyboard and monitor.
HP 3000 shops who have made the move to Apple's solutions — well, they have been more rare than a quarter with no major Windows service pack. The largest supplier of wrestling uniforms in the US, Fergo Athletics, has run their company for years with Macs after HP flagged on its march to keep the 3000 up to date technically.
You don't have to go very far to find a fan of Apple solutions here in The 3000 NewsWire offices. We have run our publishing company on these desktops since our first issue in 1995. Our main database has been a modest Filemaker design; plenty of small businesses use Filemaker.
But Apple has remained a desktop solution in the minds of many IT managers, at least up to now. By blending the Xserve hardware with the new Leopard 10.5 OS X Server environment, Apple means to tilt the Unix choice back toward what HP 3000 customers know: Owning a server. Adding applications. Patching infrequently. Not having a computer which needs to needle you about how much faster you might be running, if only you just turned on that extra capacity, the CPUs you already put a 20 percent down payment on.
InfoWorld reviewer Tom Yeager made a strong case this month for why Apple's Unix solution is different in a better way:
Xserve strikes the perfect chord with everyone, from the server neophytes and Windows refugees who want plug and play, to the Unix graybeards allergic to proprietary system software, equipment or development tools. Freed from the never-ending spending of Windows and the do-it-yourself shipbuilding of Linux, every single buyer of Xserve will end up doing more with Apple’s server than they had in mind when they bought it.
When you toss in what's possible now with Apple's desktops — installing Parallels software for $80 per Intel-based Mac, so the same machine can run either Windows or Apple's OS X applications — well, some 3000-migrating IT managers considering Unix will get back what they always wanted: Industry standards in Intel hardware and Unix, plus desktops that run the two best client environments in the world. All from the same vendor. That sounds like the Hewlett-Packard of the good old days, when a single vendor could provide everything your IT enterprise needed.
November 17, 2006
Rocks tossed from the Blue
Since HP has now started to ship its speed-busting Montecito-based Integrity servers, IBM has fired up a loose cannon to hurl FUD rocks across the bow of the HMS Itanium. Big Blue, now trailing HP in overall revenues in the computer business, hired analyst Joe Clabby to explain why he's turned his coat against Itanium.
Clabby is careful to mention he was among the Itanium boosters as recently as last year. His admission and comments come in a 25-minute, Flash-based videocast which IBM has paid for, a screed in which he flogs a list of 10 reasons why migration to Itanium is a bad idea.
None of the items on Clabby's list are really news; it's just that now he's on the side of the skeptics about whether the migration is worth the effort, and whether Itanium is a destination worthy of the expense. Just have a look at his 10 reasons:
1. Forced migration
2. Changing market conditions
3. Roadmap slippage
4. The scale-up/cascade down dilemma
5. Lack of innovation
6. Ecosystem weaknesses
7. Workload emphasis
8. The propaganda machine
9. Intel re-evaluation
10. Future promises
What is Number 8 even doing on the list? Propaganda is what the opposition calls your truth they disbelieve. As for forced migration, the topic is absolutely no news whatsoever to the HP 3000 customer. How much forcing is really going on for anyone but the Alpha customers ? Depends on what else you're selling against Itanium?
Let's be clear here. Analysts like Clabby, or those who work for Gartner, are all for hire. In one way or another, a fee is exchanged — and nothing as value-priced as what we earn at The NewsWire — for what Joe and analysts like him think. He first turned away from Itanium boosterism early this year, when the market was chewing over defections by IBM and a blown shipment date for Itanium's Montecito models.
In February of this year, I wrote an independent white paper that openly criticized Itanium and the ISA. That white paper was funded by no one. The opinions expressed here are my own.
We will only note that everybody's opinions are their own, on any subject. That's not what is purchased. What's bought and placed on the Internet is the slick camera work, the transcript captioning in a polished Flash 9 Web presentation, the podcast and video podcast of Clabby's opinions. Instances where he was surely paid to sit by the camera, microphone or deliver his slides for use by IBM.
But they are his opinions, for sure, and that February white paper which was his audition for this paid mudslinging, yeah, it was independent. Back then, anyway.
The opinions include questions that few buyers will understand, like "Where are the new technologies such as silicon-on-insulator or strained silicon (dual stress)?"
The valid questions are issues of business that nobody can see clearly or guarantee they know the answers to; questions like "How important will Itanium remain to Intel?" or "How many companies will offer applications written natively to Itanium's architecture?"
Intel's answer to the first is that Itanium represents the company's only real shot at new business. It already owns 80 percent of the desktop CPU space with its Xeon/Pentium chips, while it's got less than 25 percent of the headcount in the beefy-server space. Oh, and you can just imagine how much more Intel can charge for an Itanium 2 than a Pentium 4.
The answer to the second question, about how many apps, is "enough for some customers, and not enough for others," but the roll call does include apps with big profiles like Oracle and PeopleSoft and lots of apps in the Windows ecosystem.
We have argued this matter in the other direction. But at least we didn't advise people to sink investment money into Itanium, then watch the company go bust. Clabby tells a story about his advice iin 2004 to Migratech, no longer in business. By Clabby's account, the reason the company went under was HP and Intel's fault, not the reliance on faulty advice. Sitting on the company's advisory board, he
Recommended that Migratec focus on helping customers and ISVs move their applications from PA- RISC to Itanium — [and Migratec is] now defunct mostly because the Itanium market did not materialize.
Well, it didn't materialize by 2004, something nearly everybody could see back then. Being contrarian works better when you're advising caution rather than selling risk. It's clear we disagree about this subject, and Clabby and I have done so all along. Have a look at the dog and pony show out at the IBM-funded Web site, if you like. Its content says more than any analyst should trumpet about how this paid-advisor business works. Vendors are always after an indepdendent to toss over their independence. You have to wonder who is really convinced by something as showy as this.
November 16, 2006
HP's fourth quarter blasts profit records
HP announced fourth quarter and fiscal 2006 results this afternoon that knocked company share prices beyond $40, as investors crowded around a company whose earnings quadrupled from the same quarter in 2005.
HP weathered a big charge on its earnings in that 2005 quarter, but the results from the last 90 days stand on their own as a significant step upward. The company finished with $91.7 billion in revenues for fiscal 2006, another milestone broken. By many accounts, HP's revenues now lead the computer industry, so long as you include the $26 billion cranked out by the company's printer, ink, paper and camera business.
The quarterly profits came in at $1.7 billion for the 90 days when HP sells the hardest and discounts the deepest. HP came just shy of selling $25 billion in the period, a record for Hewlett-Packard. Revenue grew across the globe compared to the 2005 quarter, with revenue in the Americas up 7%, revenue in Europe, the Middle East and Africa (EMEA) up 3%, and Asia Pacific up 7%.
Analysts and commentators on the market were noting that the scandals HP endured in September didn't keep customers from choosing the company's products. In a Wall Street Journal report, however, analyst Brent Bracelin said HP is going to have to increase its profits organically, rather than by cutting costs. At a conference call this afternoon, HP CEO Mark Hurd said the company is "substantially completed" with its restructuring, corporate code language for "the layoffs are about over, for now."
As for the HP group developing and selling the HP 3000 replacement models and operating systems, Enterprise Storage and Servers saw its sales rise 4 percent over the same quarter of 2005. But the growth came from "industry-standard" (read: Windows) blade servers, not the HP-UX driven Integrity servers.
The ISS server business rose up 9 percent from 2005, and blades grew 38 percent. Meanwhile, the Business Critical Servers' sales declined 4 percent. Although the Integrity server business — the future of the HP-UX operating environment — increased 77 percent for the period, HP took a serious hit in PA-RISC and Alpha server sales. Nothing like telling the customers their architecture is being sunset to get them to stop purchasing it.
By the looks of the numbers, HP is not drawing in new business with its Integrity offerings as fast as it is turning PA-RISC customers into Integrity shops. On the bright side, the BCS profits for the quarter showed a $98 million increase.
HP expects more sunny days in the quarter to come. It believes its profits for Q1 of 2007 will be in the range of about $20 million. Profits will flow from the PC business where HP battles Dell. HP's notebook revenue climbed 24 percent for Q4 of 2006, while desktop sales were flat. More and more customers are choosing notebooks as their desktop systems. Meanwhile, Dell delayed its quarterly results today because HP's chief PC competitor is being investigated by the SEC.
For readers who eat this stuff up, a replay of the conference call from the this afternoon is up on the HP corporate Web site. If you have a Windows system, you can listen to the call at a page driven by Windows Media Player or Real Player. The call is one hour, but it's indexed with 26 minutes of HP speaking and 31 minutes of questions from the business press.
November 15, 2006
On Watch at the Watershed
It's taken five long years, but it looks like Hewlett-Packard has finally assembled a solution for turning away from HP 3000 systems. There are always going to be companies and customers who cannot afford much change in their environments, if any at all. But for a portion of the market which can spend to improve their IT services, the time appears to have arrived to make a move.
Five years ago, and for nearly every year afterward, we would pose a question to our readers and the HP partners: Why would you want to migrate, or move away from HP 3000s? Now that we've seen the presentations and the proposals for the first chipsets to truly trump PA-RISC, and a version of Unix that delivers economy and flexbility along with all the change it demands, we are prepared to turn our question of the past five years on its head. Now we want to know why a customer would not want to move on to something newer, faster, designed as well as the HP 3000, albeit very differently.
We can anticipate the answers to our turnabout (some might say turncoat) question. Many of the community's customers who cannot move off a system with a great history cannot afford to go anywhere. The value of the 3000, set for a 5-10 year lifecycle and with an integrated design to keep things running smooth — well, no company is offering such a computer, abreast of the latest standards and omnipresent applications. No company but IBM, anyway. So unless your company is ready to switch systems suppliers, or add more spending to your IBM budget, or enter the churning waters of the Windows experience, it looks like HP-UX and HP's new entry-level and midrange 3600 and 6600 Integrity servers are your best future bet.
I came away from the half-day of HP's Integrity Solutions road show impressed. The hardware is now inexpensive as a capital cost, though the price tags for iron and OS are now among the smallest parts of an IT investment. Power consumption and software licenses, as well as support, have far outstripped the old way of adding up what an environment costs.
(No solution is cheaper than remaining in a non-spending situation, of course. The longer you can delay the cost of change — keeping a close watch at the watershed for when the waters run against your path — the more prudent your spending appears when you must spend at last to survive.)
It's hard to express the admiration and confidence I have in the homesteading community. Although it's now being outnumbered nearly two to one by the HP shops moving, scheduling and staffing a move or already operating an HP-UX, Linux or Windows environment, the Homestead — as I coined the term five years ago this month — will never become a ghost town. As long as change remains an expense item which a company needs to debate, then the homesteader will always have company enough to maintain a working HP 3000. Some very big companies will still be pumping billions of dollars though their 3000s five years further down the way from today's watershed.
Montecito, the overdue generaton of Itanium, opened the gates of conversion for me. Itanium is complex and so very, very different from the rest of the 64-bit world, and that will keep it from becoming the household name that HP wished for its Intel partnership project. However, HP is the second biggest computer company in the world — we don't count the billions of dollars in ink and paper as computer business, sorry. And if Number Two puts its faith in Integrity, like I have now seen many HP engineers, managers and their customers do, well, who are we to say that we know for certain that ship will founder, to fall short of HP and Intel's needs.
We all know why HP needs HP-UX to succeed, and why Intel needs Itanium processors to persist. Profit, put plainly. Each represents a product the vendor considers unique, something which might command a higher margin than the rice-paper earnings from billions of pieces of photo paper and inks HP never discounts, and the millions of "Intel Inside" processors. Every Itanium 2 which Intel sells can claim to be faster than an Opteron, without dispute. You cannot settle such an argument among anyone except people who would understand how to compare these orange groves with the apple orchards. Perhaps to a little dismay from this fellow who learned about IT in the Reagan era, those people are not in charge of very many IT budgets anymore.
Over the next several weeks we will talk, in what might seem like a lot of words, about how the future proposal of change has changed since HP dealt its surprise hand in 2001. Homestead skills are always going to have an audience and a need, as well as advocates as long as those of us around and beyond 50 years of age work in this field. Transition, however, is a path which very few HP 3000 customers can avoid. "It's a matter of when, not if," we've heard the gravediggers say about the 3000 community's upcoming change.
We have a corollary to that statement, having seen HP finally make sense of its Adaptive Enterprise, deliver the virtualization promised a dozen years ago to 3000 users, and at last power up a computer that can blow the doors off the biggest 3000 at a serious discount in capital cost.
Our corollary to the diggers' when-not-if is, "It was only a matter of time before HP got the migration options right. So long as they didn't give up, or get assimilated, of course." Those options were far from right in 2001, or for several years after that. Now we don't hear only from the 3000 giants about their moves, but from little companies who see value in difference that delivers options and flexibility. We'll tell you what we've heard in rooms of less than 50 people in two HP conferences in the Houston area during the past two months, but we want to hear your story as well. The spin zone rarely sets up along a watershed — a word defined thusly at Dictionary.com:
(noun) An important point of division or transition between two phases or conditions
Bring an able hardware successor to market. Offer its operating environment with new power to offset the differences your customers must scale. Then then do all you can to swell up the value to draw customers across the divide. HP now shows all the right signs of doing this well enough to keep your career safe if you must move.
Here at The NewsWire office we will bring in our first hardware systems capable of running Windows this month, as well as our first install of a Microsoft Windows environment. Not because we don't believe in our version of the 3000 group, Apple; our steps into a the world of the common (computer) word will be under the shield of an Intel Core Duo-based MacBook Pro, a half-inch thick laptop three times as powerful as our two tower Macs combined. We're adding that element of what is common to the rest of the world to our capability toolbelt, with its differences to be used only in cases where our proven tools are not being served by the rest of the world. We think that's a sensible mantra for our time, an era in the world's lifespan when working together, across differences, is a crucial effort for the common good.
November 14, 2006
Firing shots, five years later
It was a Wednesday, mid-day or the end of the day, when the world learned the HP 3000 was going to pass into HP's history. But not soon, HP assured its customers five years ago today. In time the vendor came to understand that not even the five years it clocked out would be time enough for some of its customers to leave the 3000, power down their power tool in their IT center.
Five years onward now, many have left their systems behind. Lots of those had a foot on the transition path already on November 14, 2001. But for many others, the journey continues. The destination of a better business computer remains on the horizon. For some, that horizon will be defined as Vesoft's Vladimir Volokh defines the term: "Someplace which, the farther you travel toward it, the more it recedes into the distance."
I am travelling between two options this week, studying opportunities. Last weekend the Greater Houston Regional User Group hosted a very successful HP 3000 conference, so useful that the board of directors has already set a date for the 2007 edition: September of next year, at $175 per registration, or free from your favorite transition, homesteading or migration vendor. Make a space in your training budgets of 2007, no matter how far along you are. That conference is a shot across the bow of a ship, one that HP said was supposed to be in port already.
Others are fired up about the newest HP Integrity servers, powered by the Itanium chips that have been evolving since before we began the NewsWire's 11-plus years of storytelling and service. I leave now for the Gunther Hotel in San Antonio for HP's Integrity Solutions road show. This hardware to replace the 3000 is fast and well-evolved, too. Perhaps more important, and where I will spend most of my morning, is in the virtualization briefings, a feature of flexiblity most 3000 shops have never seen.
Take a look at our editorial warning shot of five years ago, predicting that you can never tell what death really means, or how an afterlife will look. That was a scary Wednesday, but you have survived it with the ability to take a shot in the dark at a new solution, be it going your own way to homestead, or leave HP to adopt another platform more open and less vendor-dependent, or follow HP's torch into new territory.
November 13, 2006
HP pays OpenMPE for engineering advice
Hewlett-Packard has paid advocacy group OpenMPE for engineering services surrounding the vendor's review of the HP MPE/iX build process. Birket Foster, chairman of the advocacy group's board of directors, confirmed in a briefing on Saturday that an OpenMPE "virtual lab" engineer — a 3000 veteran who is an independent contractor — was paid to review HP's process.
The payment represents the first revenue which OpenMPE has generated since the group formed in 2002. All monies which OpenMPE received up to now have come as contributions or donations — including the $5,000 which HP pledged in 2004 to the group.
The contract was not a one-off engagement, either. Foster, who was briefing the HP 3000 Conference crowd with HP's Jennie Hou in attendance, said the engineering project will include two additional phases. In addition to firming up the relationship between OpenMPE and HP, the project also demonstrates the kind of value OpenMPE supplies through its lab services. Apparently of a high enough calibre to earn a check from HP.
Foster said the payment for the engagement was a "pass-through" transaction which delivered the money directly to the engineer working alongside HP. (The group is still reluctant to name the Virtual Lab engineer who is doing the work alongside HP. Speculation includes Mark Klein, a former OpenMPE board member who did not run for a second term — and whose MPE/iX internals experience and prior contracts with HP's 3000 ab make him a leading candidate to head up any OpenMPE Virtual Lab organization.)
The pass-through means that OpenMPE itself didn't earn income from the contract. The organization is still on the hunt for customer POs for add-on support of their HP 3000s, one-year contracts starting at $3,750 per server to receive HP 3000 patches for repairs and some enhancements. HP has not yet decided if it will allow any third party deep enough access to the 3000's source code to provide such patches.
Foster said that future engagements with members of its virtual lab might include some fee paid to the OpenMPE group, money to help pay for capital costs, communications and marginal administrative overhead. For almost five years, the group has operated on volunteer time, minimal contributions of cash and extensive in-kind donations from its board members' companies.
November 10, 2006
Webster, we've got a solution
Plenty of solutions, in fact, are being offered up today at the HP 3000 conference in Webster, Texas. At a supper last night with a few HP 3000 consultants and partners, we talked about the location of this conference, a city called Webster that hasn't received its share of fame or kudos for hosting the Johnson Space Center.
The Space Center, home to solutions for problems like the Apollo 13 near-disaster or a successful moon landing, is usually called Houston, as in "Houston we've got a problem," or "Houston, Tranquility Base here; the Eagle has landed." But that lack of recognition hasn't kept Webster from serving well — not any more than the reknown which the 3000 lacked in its HP product career squelched your computer's calibre of service.
Today the meeting begins with Alfredo Rego's keynote speech, "A Bit at Home, A Bit at the Edge." There are rumors afoot that simultaneous wireless connections to multiple servers are key to Rego's talk. He told us a few weeks ago he planned to speak about MPE, Unix, Macintosh and Windows operating environments. Then HP decided to release its solution to the potential corruption-causing problems of Large File datasets in C.10 IMAGE/SQL.
It's hard to imagine Rego passing up a chance to enlighten the crowds here about the impact of HP's decision to dump LFDS, a choice which Adager supports along with every other third party database tool supplier. Rego speaks in about an hour at the keynote. In a clear bit of definition, the HP 3000 conference has only one keynote speaker, not the multiples of other computer conferences. Here in Webster, keynote means keynote.
HP said Mondaythat downloading a DBSCHEMA will turn off the LFDS capability, a feature that has caused data corruption during the past three years in IMAGE databases. Knowing what we know about how Adager eschews relying on the IMAGE schema — manipulating the database's root files are Adager's forte — we bet some of Rego's bits will be addressed, so to speak, at the solution from HP. For the record, Adager says that "Everyone should apply the HP patch." When the solution arrives, after beta testing.
Webster, after all, has been the root of solutions for a long time. More than 30 years, in fact, just like the HP 3000.
HP is part of this historic HP 3000 conference, by the way. Historic because we can't find an instance of a vendor who has announced the obsolescence of a computer platform and then five years later helped support a conference all about that platform. Maybe HP, like Webster and the 3000, deserves some kudos for changing that bit of history.
HP will also use this forum to restate and clarify its licensing policies for using the 3000 and MPE/iX. From what we've heard, there is little chance of a return to the FBI-threatening, slap a vendor into house arrest days of 2000, the last time HP protected its property rights. But those rights are still in effect, for any company who's still doing business by the book with Hewlett-Packard.
Rego's keynote is especially appropriate because it was Adager CEO Rene Woc who discovered the LFDS corruption years ago. As for the march of history, HP heard from its community members six years ago that Jumbo datasets, not LFDS, were the way to deliver dynamic expansion of both master and detail datasets. Jumbo was already well along by then, and the customers just asked the vendor to build out the project to include master capabilities.
But mistakes in judgment cannot be changed, only offered as lessons. Just as the political tides have turned in the US after its Tuesday elections, the path of cooperation looks like the most effective trail to tread for HP and its community of users. In less than an hour, we will be well-met.
November 09, 2006
HP to update migrated app performance
HP will offer its migrating customers another update on the performance of migrated applications in a few weeks, but this time the advice will be available via a Webcast. The last time HP presented "Performance of Migrated HP e3000 Applications" the engineering advisory came as part of a paid registration to the HP Technology Forum. On December 6, the HP user group Encompass will offer the talk, usually presented by HP's Kevin Cooper, in a Webcast from the Encompass Web site.
HP has given this talk at least once a year since 2001. It's updated with customer migration experience which HP gathers during its migration engagements with HP 3000 sites.
The hour-long Webcast at 2 PM EST does not promise migrating users a definitive forumula for their figuring. Many migrating 3000 sites want to know how much hardware to purchase to get the same level of performance as what they now get from HP 3000 systems. A lot of what has been reported on this subject can be shortened to "It depends."
The Encompass/HP announcement sets expectations accurately. "From this talk, hopefully you will understand why your questions about migration performance can’t be answered with a single number!"
We're not sure if HP's got new data to offer on this subject, or simply wants to restate what's been discussed in the past. Cooper has been giving this talk for years; the presentation's bottom line in the past has been "databases are the most important element in estimating how fast your migrated app will run. A close second is your app's dominant language." These results have given applications moved to Marxmeier Software's Eloquence database and AcuCOBOL's COBOL the edge in performance.
To compare stories from HP on the subject, you can have a look at our 2001-2005 reports from Cooper's very sound presentations in stories from The 3000 NewsWire (do a search on "kevin cooper" + application in our Search Engine) as well as from our NewsBlog in December of last year
n discussions about migrating applications off the HP e3000, a common question is, “How will these applications perform on a new platform?” Users moving off the HP e3000 need to know what size system will be required to run their migrated applications. This webcast will explore what is involved in figuring out how an application will perform when it is moved from MPE/iX to another operating system like HP-UX, Linux, or Windows.
This presentation will look at the different components of a migrated application, and investigate the performance implications of some of the available alternatives for each component. HP takes a look at some of your choices in hardware, including platform, CPU speed, number of processors, memory, and disk. Cooper also examines some of the performance differences in software, such as operating system, language, database, the I/O and networking layers, and the user interface.
To register, visit the Encompass Webcast page and click on the link at the bottom of the page.
November 08, 2006
Regional results from a radio report
In a podcast report (5 MB MP3 file) that's an homage to the great broadcast voices of CBS News, we hear that customers backing an HP 3000 party were sweeping toward the Gulf Coast tonight. The staunch resistance to obsolescence was carried on a wave of strong turnout for this weekend’s first HP 3000 conference. The meeting begins at the University of Houston Clear Lake campus just southeast of Houston Friday morning. As if the 3000 experts were not enough, there's the Gizmo Guys on campus Saturday night, at just $10 a seat at the door.
November 07, 2006
Change means you don't know what good comes next
3000 NewsWire Editorial
Life can sometimes change, as the song in Cabaret says, “due to one little word: married.”
My little boy has just become a groom now, someone’s husband. The man who gets up to see what that noise was in the middle of the night. The fella who buys flowers when the fight ended last night and they haven’t made up yet. The guy who tells his woman he’s sure she’s beautiful, even when she isn’t sure he’s telling the truth.
That’s my son, Nick. He was married in a celebration last month. A handful of hours over a couple of days when everyone smiled, when old hurts healed over, when hugs worked their song-filled magic, embraced new love, soothed the sick, sparked happiness.
I drove to The Plantation, purpose-built in the 90s with an ante-bellum design, driving with my mom in the seat next to me, with my brother Bob behind me, with my wife and partner Abby riding along on the wave of joy and expectation. And the tears poured down my face when “Married” and “Sunrise, Sunset” rolled off the iPod and into our ears. The day was clear and cool, blue skies an omen for the life of Nick and Elisha together.
It’s a wonderful thing, this kind of redemption, the outcome you work toward but dare not expect. This HP 3000 community which has cared for me, brought me up in the industry for almost as long as I have been a father to this new groom, sustained Abby and I — you are reaching for resurrection, maybe hoping for redemption of a work-life-long partner, HP.
Few things stung more in my career than enduring the news of HP’s boardroom miscues. Few things hurt more than your child’s mistakes they must make on their own, to learn lessons that endure.
But when that moment of redemption arrives, whether it’s believing in the sustainability of your homestead plan, or finally marrying up with the companies that can help you move away from the home of your 3000 youth — it’s a powerful revelation.
At mine with Nick, I wore a tuxedo with a white vest and white tie for the first time. White tie, at last, the better part of a quarter century after I learned in a little newspaper office in Burnet, Texas that I was going to be a father. The woman who had told me that in 1982, waiting for my joyful response and not hearing it, finally saw my joy for our boy. We hugged one another at wedding and rehearsal in an act of contrition and submission, and told each other, “We did good.”
The cars kept arriving and emptying like a fantastic circus act, unloading those we knew and those we met for the first time, Elisha’s vast cast of Wallers and Mezettis and those they love. One fellow so beloved that even after he divorced their sister, the Mezettis loved him and the woman who became his next wife. He’s a regular at Thanksgiving and weep-fests like that That’s the power of love, back to a bright future.
I sat on the front row next to my wife and wept even as the bride and groom walked the path to the gazebo. Why wait? So many tears needed to flow that day.
And why not? It was a day of redemption. Out of struggle and failure comes success. I believe if you haven’t moved on from HP’s decision of five years ago, redeemed yourself from that scary moment when everything you’d built changed in value, now might be the time to set that rock down. Surely, HP’s humiliation of the past two months can qualify as retribution enough, a story of how big a mistake can noticed. The general press never so trumpeted the error of dropping the 3000 off HP’s future. For some customers, that was a much bigger story than the one that squatted on Page One for much of September.
But like any story, it ends sometime. You decide when.
So like my son, who saw so much from me of the hard, the lesson-through-mistake that marriage can bring, you can take a leap of faith like he did on that wedding day.
He held on, snatched up the reins of that spitfire steed we all aim to steady in our saddles. We had all hung on for that day. His grandmother, my mom, through all her years, so that at 80 she was seeing her first grandchild married before those old eyes filling with tears.
Abby and I had held on, my wife, my partner, the first person in the world to believe in the NewsWire, now in its 12th year, with more to come. She’s become my special consultant as of this issue, retiring away to teach 14 classes of yoga a week, but never further away than the room next to mine here in the southern part of our house in Austin. This morning the low autumn sun pours through my window where I write and yes, cry again.
His bride hung on too, clinging to hope while she weathered a young man’s reserve, wariness, a sensibile his dedication to the impossibility of the prospects of a lifelong pact. Two of every three marriages end in failure today. But Elisha kept believing in Nick’s power of faith, waiting for it to grow up.
Finally, Nick hung on to himself, not the doppelgangers of his parents and their partners and the shadows of disappointment that marriage had showed him.
His grandfather did not hang on for this day. But at the end of more than 100 photos, I snapped the last one, capturing Nick’s joy in a picture that would have made my father, cameraman in his life, smile with approval. Two Seybold boys, Ron and brother Bob, worked that wedding with joy and aplomb and tiny digital wonders in their hands. Bob took the best stuff, an honor I always struggled to give him when we were boys just 15 months apart in age.
Things change. Nick took on work that he enjoyed, something none of us saw coming, convincing people in the artistry of sales, handling the most charged of elements, money at Compass Bank. He had to learn to show faith despite his own doubts, to see that half-full glass and offer it to others. Earning money, earning praise, earning his own self-love, he began to grow old enough to grasp some of life’s richness.
What can a day of delight change? Perhaps very little. Perhaps everything. You never know. Set down your rock, lift up your glass. Join Abby and I in celebration of the NewsWire’s first 11 years, of a day in joyful tears. Still married to what she created, still by my side in her office next door and then in a yoga studio down the hall or across Austin, she joins me in wishing you and your company a career full of happy surprise, and redemption.
November 06, 2006
HP to help wipe out LFDS on 3000 databases
HP has implemented its plan to resolve its two-year-old Large File DataSet (LFDS) issue with the IMAGE/SQL database on HP 3000s. But the vendor’s tested and approved change which fixes the potential LFDS database corruption issue is not going to be released, according to MPE/iX lab director Ross McDonald. Instead, HP will release software to get LFDS ability out of IMAGE databases.
HP is shelving the months of work on fixing LFDS to stay in step with the HP 3000’s database community. Experts at Adager, Robelle, Bradmark, Allegro Consultants and other vendors have all said that LFDS needs to be withdrawn from the C.10 versions of IMAGE/SQL. Simply put, the fixes to the software don’t justify the benefits that LFDS promised: dynamic expansion of detail datasets.
LFDS had the potential to corrupt databases which had datasets bigger than 4GB, by all accounts both within HP and in the database utility vendor community. But few customers used the feature knowingly, especially after Adager and then other third parties warned of its corruption capability.
McDonald said HP doesn’t believe there are any customer sites experiencing corruption issues due to LFDS, a feature first released in 2004 and fraught with multiple attempts to fix its problem.
Instead of releasing the tested patch to fix LFDS, HP will be releasing a patch to disable the capability. This patch, TIXMXW7, is now in beta test, and will run on versions 6.5, 7.0 and 7.5 of MPE/iX.
“This could take some time,” McDonald said of the beta process for the TIXMXW7 patch. HP will post notices to the community through online resources, including the 3000 NewsWire blog, when TIXMXW7 goes to General Release (GR).
On a more immediate schedule, HP is using the HP Jazz Web site to offer a new tool today that lets customers discover if they’ve created any Large File Datasets in their IMAGE/SQL databases over the past two years. This LFDSDET tool, available at jazz.external.hp.com/lfds, scans a 3000’s databases — though it doesn’t need to read a full database — to locate LFDS on a 3000. The tool, like all on the Jazz Web site, is free, along with the DBSCHEMA file.
The Web page jazz.external.hp.com/lfds has full details and links to the available resources. The TIXMXW7 patch, once it reaches GR status, will be available on the HP ITRC Web site. Like any other GR patch, TIXMXW7 can then be downloaded and used by any HP 3000 customer, including those who don’t purchase 3000 or MPE/iX support from HP.
LFDSDET is HP’s first tool to examine the contents of a 3000 since HP’s Jeff Vance created SIUPDATE to do an inventory to help in migrations.
McDonald said HP decided not to release its fix for LFDS “in part because it would require all the database tool vendors to change their products — and also, some Privileged Mode customer applications,” he said.
LFDS is embedded deep in the IMAGE database code, so changing it would add risk to using IMAGE/SQL, he added. “This isn’t the right time to be adding risk,” he said. McDonald went on to indicate that “we have concluded that this is the best solution from both a customer and partner perspective.”
Jumbo datasets have been in use much longer than LFDS, and HP said jumbos meet customer needs as well as LFDS with one exception. Dynamic expansion of detail datasets can’t be handled if those datasets are Jumbos. But third-party products from database providers are good workarounds to meet that feature set, HP says.
After the TIXMXW7 patch is installed, the LFDS Detection tool will automatically be executed. If the system is found to be free of LFDS databases, a file will be created that prevents write access to LFDS databases.
However, if the detection tool finds LFDS databases on the HP 3000 system, these offending databases will be identified, and instructions will be displayed on the system console directing the customer to the jazz.external.hp.com/LFDS Web site. In this case, write access to the LFDS databases will continue to be granted so as not to impact ongoing customer operations.
HP says that after the LFDS databases have been converted to Jumbo databases, customers should re-run the LFDS Detection Tool in installation mode to verify a clean system.
November 03, 2006
Where is Integrity on its takeover curve?
Although we asked this question of HP's Nigel Ball, he could not answer until HP's year-end results are announced in a little less than two weeks' time. HP set a goal a few years ago — increase the revenues from sales of Integrity servers, HP's recommended replacement for HP 3000s, until Integrity's and their Itanium 2 processors made up 70 percent of HP's revenues.
In 2005, HP said its goal for the future of its enterprise server business was to have Business Critical Server unit revenue will grow to 50 percent by the end of 2005, and 70 percent by the end of 2006.
Attaining this goal is important to HP 3000 customers. Those who are staying with HP and investing in HP-UX in particular would benefit from HP hitting its targets, and on schedule, too. HP's Unix runs on PA-RISC processors in HP 9000s, sure. But the hardware with the biggest boost for the buck uses Itanium. If nearly three of each four systems sold wear the Integrity badge, that's a lot of critical mass for the HP Unix platform.
Last year at this time, HP was trailing its 2005 estimates of Integrity ascendence. Ball promised us numbers once HP's quarterly figures are released in the week of Nov. 13. HP, by analysts estimates, will have revenue of $91.2 billion in fiscal 2006.
On Nov. 14 the I'll be riding down to San Antonio — just a few days after the HP 3000 conference on the Gulf Coast — to see HP's Integrity Solutions road show. Raw hardware speed and feeds will make up one track of the morning-long event. Virtualization, which has extra features on the Integrity servers, makes up the other. HP is tailing off its production of the PA-RISC servers. HP 3000 sites are ordering little else but Integrity systems, according to reports from the migration vendor community.
Orders and sales of Integrity will be a part of the HP report and analyst briefing two days after the San Antonio Integrity show. On Nov. 16 customers can get a glimpse at how much ground Integrity has gained since those HP forecasts. The delay of the Montecito generation of Itanium 2 last fall — an extra half-year of waiting — could not have helped.
Montecito's speed and cost advantage can help deliver the critical mass for HP-UX to give the operating environment a safe lifespan. It's true, applications do drive the marketplace. But Itanium is very different than other architectures. It needs an ample installed base in order to maintain the interest from the application community. HP has more at stake than just meeting numbers by the end of this year.
November 02, 2006
Making a time shift on the 3000
Last weekend, HP 3000 customers saw the last of the switch-to-standard time during a month of October. Citing a desire to conserve energy by keeping the lights off as long as possible, US government officials have changed the switchover times for both "fall back" and "spring forward." HP has sprung into action to accomodate the changes on the HP 3000.
As always, customers can fall back on the community's keen eye for detail to ensure HP's time zone engineering gets onto the systems in plenty of time. After HP's Bill Cadier announced HP has created the file needed to embrace the new mid-March and early-November dates for Daylight Saving, 3000 community members pointed to related information both supplemental and incidental.
Donna Hofmeister of the OpenMPE board of directors says another director — apparently wanting to remain anonymous — thinks somebody in HP doesn't understand how the magic TZTAB.LIB file does its work.
Donna reported that the OpenMPE director said, "The author of the HP Web page seems not to understand the TZTAB function. In the section “Maintaining TZTAB” HP states:
Under Section 110 of the Energy Policy Act of 2005, the Secretary of the U.S. Department of Energy must make a report to the Congress on the effectiveness of the daylight saving time changes on energy consumption in the US no later than 9 months after its implementation. The Congress of the United States reserves the right to revert the daylight saving time rules to those implemented in 1986 and in use through 2006 based on that report. Should Congress opt to revert U.S. daylight saving time to the pre-2007 rules, this change will probably take effect after the end of MPE/iX support on 31 December 2008. This is why we recommend retaining a copy of the current (pre-2007) TZTAB.
The director continued, "In the event that Congress decides to undo the DST rules, it won’t undo the time relationships between UCT and Local for timestamps applied during the period in which the new rules were in effect, so we can NEVER return to the current TZTAB; so retaining a copy of the current TZTAB for the purpose of someday restoring it is nonsensical. If the DST rules are in fact reverted to the old rules, then the TZTAB will simply grow longer, not shorter.
"Once you have tested the new TZTAB installation, you could keep the current TZTAB for historical or sentimental reasons, but make sure it is in a safe place where it will not be confused with something valid."
Okay, somebody asked, does that mean we should apply this patch, or not? After HP's Jeff Vance said sure, install it sooner than later, Tom Emerson offered this advice:
If for some obscure reason this change gets recalled prior to the first change in 2007, then you could actually “roll back” to the prior version because the “new” version would never have taken effect.
I would say if you really wanted to hedge your bets, you could hold off on updating until the day before the next actual DST change in 2007 — provided, of course, you don’t have any software that attempts to calculate (and make use of) “future” dates (and times) that might be past that point. If, as I said, this gets recalled, then you’re good and you don’t have to actually install it after all.
However, once the first change really does take place, if any further changes occur (such as if it gets repealed), then a new file would be necessary because, for a brief period of time, this did take effect, and any software that “looks back in history” would need to take this into account when calculating dates and times “in the past."
And Gilles Schipper, who will be appearing at next weekend's HP 3000 conference on the Gulf Coast outside Houston, added this all-purpose instruction about the changing times of the 3000:
Don’t forget to also modify your SETCLOCK criteria when traversing to and from DST according to the new DST rules to come into effect next year.
For most applications, correct setting of the TIMEZONE at the appropriate date is arguably more important than a correct TZTAB file - although correct instances of both are best.
I’ve taken the liberty to include a sample job stream file from Dave Powell - from the HP 3000 newsgroup of Aug. 8, 2005 — that is suitable for the DST algorithm. (We point to our article on the NewsWire blog from that week for a little background.)
I leave it to the reader to simplify this job stream to eliminate checking the year, since this check is now superflous. The example is suitable for the Pacific timezone.
Adjust accordingly for your own timezone.
November 01, 2006
HP 3000's extended time gets posted
Hewlett-Packard — and especially the friends of MPE/iX inside the company's lab — are readying the HP 3000 for a future which the vendor never planned to support.
This week HP posted a notice of a replacement copy of a file HP 3000 customers will need next spring, when Daylight Saving Time expands to four more weeks of the year, a change in law beginning in 2007. Yes, 2007, the year that HP planned to be finished with the 3000 support tasks.
It doesn't take a lot of imagination to see tasks like this file replacement popping up on HP's 2007 and 2008 to-do lists. Major work? Not really. Crucial to the smooth ride you've come to expect from 3000 ownership? Yes indeed.
HP has already made the new file available from its Jazz Web site, months before it's needed. Now the customer will be able to manage time changes in early March, and early November, of a year when the majority of customers were supposed to be migrated. Perhaps by next November the time of a majority of 3000 sites migrated will have arrived.
HP engineer Bill Cadier reports HP's going to fix the problem for all three supported versions of the 3000's operating system:
There will also be an official patch called LBCMXX4 versions A, B and C for 7.5, 7.0 and 6.5 respectively. I anticipate that these three versions will be available on the ITRC by around November 15. Once they are available, the Web page on Jazz will be updated to reflect that.
There is absolutely no difference between the TZTAB you get from Jazz and the one delivered by the patch and all the patch will do is replace the TZTAB file. The patches were created for those folks who prefer that delivery mechanism.