How could a vendor suggest that a widely-installed and mission-critical product be turned off? Have a look at what Microsoft is doing this year. The advice has been to turn Windows XP off, replace what's working. HP 3000 users got the same advisory in 2001.
That was a momentous year for MPE users, but the year that followed contained the same confusion from the vendor that Microsoft is facing now. I noticed this as I dug into Jon Diercks' MPE/iX System Administration Handbook. It carries fine information, an opinion I expressed in our recent mini-lesson about BULDACCT and some automatic security that it provides. As I did my digging I found a stale message inside the book, but it wasn't one that Diercks created.
You might believe that nobody could apparently see what was about to happen to HP's 3000 business, considering what appears on pages xxi through xxiii. It's a foreword from the General Manager of HP's Commercial Systems Division, Winston Prather. A book that was released in 2002 -- yeah, months beyond that 2001 exit notice -- includes this advice about ownership.
Today, with technologies like Samba, Java, GUIs, our WebWise products and our partners, the HP e3000 still provides a great environment for the creation and support of new object-oriented, web-based applications, as well as e-service and e-commerce environments.
The book's readers absorbed that message for years after HP insisted that Prather was wrong. Or to be accurate, when Prather took pains to tell his customers the 3000 was not a great environment for any of the above tasks. It was probably as confusing as what Microsoft's done this month by releasing an XP security patch after it insisted it would not. Some writers believe that patch should not have been released. That's the kind of sentiment I continue to hear about HP twice-delaying its 3000 exit.
IT admins, faced with the harsh reality of finally having to upgrade to a modern operating system, will sleep well knowing that Microsoft is a pushover and will continue to support XP while it has a significant number of users. The status quo is preserved.
Except it's not preserved, not any more than MPE/iX status was preserved during a pair of HP's two-year extensions. It's just that many companies -- perhaps the same percentage as the 3000 owners of 2002 -- find it beyond their budgets and resources to dump XP. And so even today, Diercks' book has value a-plenty to any company that still finds MPE to be the best tool for their circumstance. The book's not perfect, and Diercks always knew that was so, citing the rule all authors live by: omissions and errors will be in every creation. You must let your work go, however, learning from the creation and promising yourself to Do Better Next Time.
Put another way, perfect is the enemy of good. MPE was never perfect. If it were, than a mighty fine product like MPEX, an eXtension of MPE, would never have gained its candidacy as one of the elephants for the 3000 owner, developer, or administrator. Elephants, supported by a turtle.
Turtle? Elephants? And this has what exactly to do with MPE?
In a Hindu legend, the world is supported by four elephants, and those elephants ride on the back of a turtle. (It's a legend, so you'll have to take my word for this. Look at the picture above and you'll see four, with the fourth one tucked just behind the first.) But in this model of the world, MPE is that turtle. A few key independent software vendors are those elephants. You can decide for yourself who they might be, those elephants holding up the world -- which is the 3000 community.
But none of the elephants are mentioned in that fine MPE book. Diercks took care to note that he'd mention nobody's software except HP's in that book, no matter how fine the vendor's software performed. We've been in this situation, where including anyone just ensures that someone who's overlooked would be upset. You handle this situation a lot like Diercks did in his prologue: We know there are good products out there, but he wanted "to avoid being accused of unfairly representing (or failing to represent) any vendor or product."
One way of looking at the 3000's legend is to consider that MPE (and its database IMAGE), comprise the turtle in the Hindu story. Without MPE, there would be no Suprtool, no Adager, no MPEX. These might be considered elephants, and I'll leave you to fill in that fourth pachyderm. (If you're still with me, nominate a fourth elephant; that's going to be fun to share.)
While you're pondering all this, don't forget that Diercks' book is also something of a time machine. I mean that it contains a snapshot of the full faith of Hewlett-Packard in the 3000's MPE, right down to the hp.com/hpbooks webpage (now defunct) and HP logo on the title page Then there's Prather's best guess about the 3000's future, although it really was not written as a guess. See the language in the excerpt above; click for details.
Like Prather said in that foreword, things in the IT industry can change fast. Much faster than a printed book can be manufactured and released, with essential edits and reviews and distribution. If the publisher Prentice Hall had finished that book any later (because a publisher controls the birthdate of a book, not the writer) the MPE/iX System Administration Handbook might not even exist. That would be a great loss. At the least, this fine book wouldn't take us all back in time to HP's confusion about the 3000, confusion that mirrors Microsoft's of today.