What has changed in two years
April 20, 2007
All through the course of the HP 3000 Transition, we've been asked for advice on what to do. HP's extended its timeline for its 3000 presence. The market for third party services has grown up immensely. Meanwhile, the critical mass of HP 3000 users has not fallen off as sharply as HP and others predicted.
Those developments might change your decision, or your proposed schedule, about migration. We got a question from one user, reporting on what's gone on at Raytheon in the UK — and asking us what we think might happen.
We had some communications some time ago on HP 3000 MPE/iX platforms. I was trying to make a strong case for our company migrating, despite the uncertainty of what HP were going to do, and be in charge of our own destiny.
Despite this, the project was parked, and obviously HP made their decision to extend to December, 2008. However, this date is fast approaching, and I was wondering what your latest position was as regards this.
Hope you can provide me with an update,
Alex Purves
HP's changes to its timeline for leaving this market reflect the reality of the migration pace. It seems clear to me: the vendor wants to be of service to its 3000 customers as long as possible. The level of service offered has dropped. The cost has not dropped. What's more interesting is that the end of HP support doesn't seem to be as much of a motivator for migrations. Not for the customers, and probably not to HP anymore.
Make no mistake: HP will exit the market, at the end of 2008. Or perhaps later. The question to be answered is, "How much will this exit matter?"
In practical terms, HP's absence from the 3000 market shuts off development of 3000 software. That's MPE/iX, and IMAGE, a couple of keys to successful HP 3000 mission-critical use. No more HP patches in 2009 (or maybe 2010, if HP follows its "or later" clause in its December 2005 communique.)
HP support won't be available for purchase at that time, either. This is not the same kind of thing as losing the efforts of an MPE lab. Today, support from third parties is the predominant channel for keeping 3000s and the MPE/iX operating system up and running. Companies who use resources other than HP report consistent success in the support market.
In our view, migration and its driver doesn't really involve HP, for most customers. There are many sites who simply cannot allow a mission-critical application to run on a system the vendor no longer supports. These are the current HP Support customers. See the statement above; they are in the minority, by our estimates.
No, the migrating shops — and there are many — proceed because their applications need something the 3000 community cannot provide anymore. It may be application enhancements from a vendor. Or perhaps a technology that HP supports elsewhere, but never fleshed out or ever offered for the 3000. (Java comes to mind here right away. There are others.)
All along HP has told the customer that HP won't know what is the right time to migrate. (Later on, HP would add, "If at all," but that advice seemed to be aimed at smaller customers where HP sees the environment as static, with few changes.) The customer's business plan would dictate that, HP has said.
Our position, as Alex requested, is to look at the resources your company can assemble to migrate, as well as the upside of the migration. If it is simply to control your own fate, as Alex mentions, there are many places to do this. But the best may still be the platform that continues to run for you, provided you can assign support to a third party. Second on our list, if a free destiny is your main goal, is embracing open source solutions using Linux, MySQL for the database and a bounty of tools. No vendor will ever cut off those resources.
Finally, the migration decision needs to be taken with some external advice in hand, in my opinion. Kind of like a second opinion, really, if your company can agree on a first one. At Raytheon the management decided to "park" the migration plan a few years ago. That's a first opinion. Inviting a second from a consultant, migration services supplier, or even another application or systems provider, is a good start at that second opinion. (Although I'd bet that final group of vendors and app providers recommends a migration. And they could well be right.)
HP 3000 customers were always a more independent group than the IBM batch users. Now the prospect of independence appears at a time when Windows bolstering comes not from Microsoft, but third parties, with two to four letters in their names: HP, IBM, Dell. Third parties are a way of life in 2007 and beyond. Unless the embrace of HP's Unix is a key to your plan, your company will survive through the efforts of any party beyond HP. Coming to terms with that lifeline is the first step to freedom.