The truth will out. But will it be out in time? - Mike Magee
WELL, THERE'S an eye-catching headline, even for The Inquirer! But it brings to mind several questions that ought to be addressed. Since it's usually easier to ask a question than to give an answer, here's a few queries about the hypothetical sale of OpenVMS, an rock-solid bet-your-business OS that I have used on three processor architectures during the past 21 years.
How Not To Get Out the Word
For longer than I care to remember, the Number One gripe about VMS was why don't you advertise it? Good
question. I'd start by asking IBM why they don't advertise MVS. (The likely answer: we sell big iron and the stuff that
makes a mainframe a mainframe. not just one of the OSes that'll run on a mainframe!)
Having queried IBM, I'd ask HP Marketing--or Carly Fiorina or Peter Blackmore if the chance arose-- exactly to whom do you advertise OpenVMS, and how do you go about doing it? I doubt that either executive could or would answer the questions. Advertising a specific OS isn't part of their job descriptions, and truth be told, I couldn't answer the questions, either.
Since most people don't buy OSes as stand-alone products (Linux being an exception) determining your target audience and message delivery system ought to be an interesting--and time consuming--exercise. I don't think a VMS Blimp (which didn't do AltaVista a hell of a lot of good!) or a tidal wave of TV ads is in the offing. A VMS balloon at the annual Albuquerque Balloon festival would be pretty cool. (I'll pledge the first ten bucks--fifty if I get to ride in the balloon--but such an undertaking must go for at least $100K USD, and that's a conservative number). (If you get the money, you're welcome to sleep on my floor in Albuquerque, everything else will be sold out!)
Getting the Word Out the Right Way
Still, getting the word out is important. That's what symposia, Technical Update Days, OpenVMS Ambassador
Meetings, Sue Skonetski's VMS Pearls, www.openvms.org, www.hp.com, and of course,
Shannon Knows HPC and its companion site are all about. Obviously, to attempt to restore OpenVMS to modest
growth mode, more low-cost, high return, carefully targeted marketing is needed. One good example of this is the series
of OpenVMS Platinum Forums for which Mary Ellen Fortier is responsible. These one-day events are held globally to bring
customers, key influencers, and prospects up to speed on where OpenVMS is today and where it is going. Platinum Forum
speakers include representatives from VMS Engineering and Marketing, influential partners, and once in a while an
industry analyst or special guest. Attendee feedback is generally excellent, so if there's no Platinum Forum scheduled
for your area, let HP know what level of interest exists for such an event.
The HP Services Gordian Knot
Unbeknownst to some is the fact that as many as half of OpenVMS sales are services-led. Sure, one could sell off
every OpenVMS worker in ZK03, but the tight connection between services and OpenVMS puts a damper on that deal. Peter
Mercury and I go way, way back, but this is an Ann Livermore decision. Ditto for ISVs who sell specialized VMS apps and
have well-established business relationships with HP OpenVMS folks. Companies like Oracle (Rdb) and Cerner and Tutor
come to mind. Assuming Terry Shannon buys VMS, what happens to these relationships? (And does Terry have to make a
modest real estate investment in ZK03?)
Do The Math
OpenVMS is on track and actually ahead of schedule for a port to Intel's Itanium processor. Much of the heavy
lifting has been done already, and, to borrow a phrase from one highly-venerated Tier One analyst firm, the round peg
sure seems to fit right into that square hole. Like we Yanks in New England say, if it ain't broke, don't fix it.
OpenVMS may be a lot of things, but broken is not one of them. It just keeps getting better. (If I haven't managed to
break it in 21 years, one can safely assume that OpenVMS is an unbreakable commercial OS!)
HP wouldn't be broke without OpenVMS, but any product that brings in $800M USD in profit each year would be sorely missed by the beancounters. Put another way, VMS profits are responsible for 1/5 of the $4B USD annual HP Labs budget.
Customer Base Erosion--or Evaporation
Dropping $800M USD per annum in profit is one thing, but how do you quantify good will? Like their NSK brethren,
VMS users demand the best and are willing to pay for it. Does Michael Dell trust his online ordering system to Wintel?
Hardly. He uses NSK. There are reasons that numerous classified government applications run on VMS: they run, they are
stable, and they are secure. The same holds true in the commercial sector: security, stability, scalability,
availability, and business continuity (a more innocuous name for disaster tolerance) have become requirements, not
options. Stay tuned for developments in this space!
Take away OpenVMS, and what do you have: the Mother of All Marquee Account Defections and a deluge of absolutely dreadful press. I'll leave it as an exercise to the reader to name some Tier One VMS accounts, but it's an easy exercise. Hint: visit www.hp.com and go rummaging around for material of a VMS flavor. Unlike Heinz Soup, you'll find more than 57 varieties in this little hunt. Now, imagine yourself as the CEO of a firm that has cast its lot with OpenVMS, and has just gotten the word that OpenVMS roadmap hits a brick wall in 2 years.
Chances are, after you've properly medicated yourself with your fortitude-booster of choice, you'll call an Emergency Staff Meeting and do the obvious: decide to change IT vendors effective at once. As a VMS Named Account, your options are limited, but fear not: IBM will be ready to help you through this bolt-from-the-blue decision on the part of a rival vendor. IBM is no stranger to VMS migrations; in fact they have over a decade of experience transitioning customers from VAX and Alpha VMS to the IBM OS of their choice. (In fact, IBM several years ago won a bid against Compaq. Compaq bid a Himalaya NSK system, IBM countered with a two-node Wildfire OpenVMScluster and clinched the deal.) More importantly, IBM Services nailed a long-term contract to support the Wildfires, and has a host of experienced OpenVMS customer and field engineers at its beck and call.
Analyze THIS!
Continuing the
HP Drops VMS scenario--a scenario to which I certainly do not ascribe and absolutely cannot justify--BCS
revenues will plunge, Services revenues will suffer significantly, and whatever degree of trust HP has managed to
instill in its installed base will plunge. Transitioning from Alpha to IPF--which, based on the information I have seen
and analyzed--was a logical decision from a business and economic standpoint, but left a lot to be desired on the
messaging front! Nevertheless, the so-called Alphacide decision created plenty of angst, some of which is still
evident two years later. (And indeed, a few of the nay-sayers continue to articulate valid points to this very
day.).
Promises Kept
Thus far, Compaq and HP have delivered on their architectural roadmap and consolidation promises, which is no
mean feat given that the roadmaps were very hastily drawn two years ago by less than a dozen people who were not able
to factor in an eight-month merger hiatus that relegated many senior engineers, technologists, and strategists to the
Clean Room. In spite of that, the promises they made two years ago--an eon in this industry--have been honored, and
there is no reason to doubt that the longer-term (2005-2006) roadmaps remain valid as well.
Paging Doctor Kevorkian
HP has indeed kept its promises, exceeded the financial expectations of many, and has continued to develop,
patent, and introduce new products. I'd give Carly and her team an A for the effort they've put forth in Year One.
Year Two may well bring more downsizing, a high-level reorg, and a slew of new products, including Madison-based
Itanium systems scaling up to 64 CPUs and perhaps even 1M tpmC, new low and midrange products, enhanced storage, new
service offerings, further improvements to OpenView, and the first VMS-on-Itanium software developers kit. (A second
kit will be available before the end of the year.)
And in the End
Netting it out, nobody at HP is paging Doctor Kevorkian, nor are there plans to sell the OS. In fact, under
OpenVMS Group VP Mark Gorham, we can expect a slew of OS enhancements for both Itanium and Alpha systems. Many of the
improvements remain under NDA, but keep your eye on VMS for some good news. Sign up for a Platinum Forum or request
one. If you've been to one and liked it, by all means let Mary Ellen know! And of course, for the latest and greatest
VMiScellany, drop by
www.openvms.org and
www.shannonknowspc.com early and often!
© 2003 by Terry C. Shannon, Director, T.C. Shannon and Associates, LLP