HP calls first Webcast at C-level

CIO briefing treads old ground, pitches immediate migration

First of two parts

HP kicked the Transition issues onto the boardroom carpet last month, hoping to get HP 3000 customers moving along its recommended migration path by targeting corporate officers. But HP’s PowerPoint message to these C-level managers was only being watched in about 70 offices, an audience that read and heard crib notes attempting to sum up the volumes of detail which HP had Webcast during 2002.

HP’s George Stachnik — who made a record seven appearances during 2002 pitching HP’s migration mantra in his Webcast series — led the first salvo of 2003, “What Your CIO Needs to Know about Your HP e3000 Transition Project.” HP perhaps hoped that CIOs would actually attend the broadcast. A list of user names revealed many attendees without the word corporate in their job titles.

At least the content was tuned for a C-level attendee, starting with a sweeping summary of market decisions that led to HP’s exit from the 3000 community. HP’s history said in the 1990s software vendors “could very easily port software from one platform to a variety of platforms from different vendors.” The C-level customers didn’t hear that the HP 3000 had gained also Java capability during that same decade, a language that has come closest to delivering the Holy Grail of “write once, compile anywhere.”

HP’s market summary swept straight into packaged applications availability, the battleground where the HP 3000 kept shedding soldiers throughout that decade. The script noted that HP’s interest in open systems dealt a shortage of resources for “proprietary platforms that had dominated the industry to that point.” HP identified the 3000 as one of those proprietary platforms — but didn’t mention the 3000’s technology remained more closed because of HP’s lack of investment in the platform in the first half of the 1990s.

HP also tried to assert that its investment in PA-RISC came at the same time that application providers started leaving behind their MPE plans. Any CIOs in the Webcast missed the point that these RISC investments came years before the 3000’s application diaspora. No mention was made of HP’s 1998 plan, aborted in 2001, to move the 3000 onto its new Itanium RISC architecture. Instead, “we were able to keep the 3000 alive” with the PA-RISC investments, “when other proprietary systems were being discontinued.”

Exploring investments

Such curious new-speak history aside, the Webcast made a case that continuing investment in owning a 3000 represents a growing risk. Stachnik’s script tried to assert that third party support providers, who haven’t pulled out of the market like HP, “are subject to the same pressures that we are.” The comments attempted to tar these support alternatives with a risk factor — ignoring the huge discount of third party support and lower operating overhead of focused providers.

“The support costs do continue to rise,” HP told its C-level targets. This led to a summary about how much it might cost to leave the platform. “The answer to that question is, ‘It depends,’” Stachnik said. Packaged applications provide the most cost-effective migration path, HP said, by following existing app suppliers onto new platforms. HP didn’t report that few companies have done this yet in the 3000 community — and those which have traced this path haven’t reported their budgets yet.

HP said that abandoning apps for an off-the-shelf app is another level of cost, and migrating home-grown applications represents the steepest, “worst-case” price level. Stachnik characterized the tool selection for migrating such home-grown apps as developed “long before HP announced it was leaving the platform.”

HP wanted the C-level attendees to believe the cost for moving off the platform is dropping at the same time the cost for remaining on the platform is rising. No evidence of either a reduction in migrating costs or a homesteading price rise was offered during the Webcast.

Attendees did get advised to compare the cost of staying to the cost of going. Completely unsupported: Stachnik’s statement that “many HP 3000 owners have discovered those two curves have already crossed, or will be crossing very shortly.”

HP hoped to introduce some urgency to move by saying it believes the cost of migrating or moving “will begin to move up again” as consulting resources become engaged. “These resources are finite,” HP said — again without any report of how fully engaged the market’s migrating companies are today.

One of those migration Platinum partners, MB Foster, had its founder Birket Foster report that senior managers want to see a return on investments in migration. Places to find this ROI will involve comparing costs of ownership, as well as proposing some of the return be realized as improvements to the applications being migrated. However, HP’s Transition Tour seminar (see story, page 20) has advised that such improvements should only proceed once a migration has been finished safely.

HP’s examples of migration covered the now-legendary Ceridian Tax Services’ port of its multi-million line application in 1997-99, as well as a testimonial from Summit Information Technology’s president Kevin Sparks. The Summit chief said the company has been able to assemble analysis that shows a lower cost of ownership using its credit union software on HP-UX.

Sparks said one of the notable savings is in training Unix-based staff, since “Unix-side education is very inexpensive to achieve.” Summit is well-known for being a turnkey provider to its credit union customers, meaning it prefers to sell all services and software to them directly, rather than provide open access between its customers and third parties. In such a model, Summit might be able to offer Unix training at whatever cost it could subsidize. The Webcast, being a C-level briefing, contained no specifics on training costs, either.

Next time: Capital cycles, replacing with packaged software, and how big migration projects are being portrayed.