I’m finding that Jonathan Schwartz’s journal provides an interesting perspective on Sun and this industry. Regarding HP, he writes:
But that said, I think HP faces an enormous challenge. And it’s not related to the cancellation of PA-RISC, or weakness in their Itanium transition. Or even Dell’s printer onslaught.
To me, HP’s problems spawn from the death of… their operating system, HP/UX. Like IBM, they’ve elected to ask their customers and ISV’s to move to Red Hat Linux or Microsoft Windows on x86 systems. And if you’re an ISV, how does that differentiate HP? – they’re a box vendor. If you’re a customer, where does that leave you with your HP/UX investments? Facing untimely change – with a vendor no longer in charge of their OS.
Contrariwise, Ian Murdoch points out that HP picked Debian in order to exercise more than a little influence.
In the same piece, Schwartz remarks
I continue to hear customers disappointed in the realization that ISVs don’t qualify to “linux” (or specifically, Fedora) – so they have to pay big bucks for RHEL if they want commercial support.
By ISVs he means Oracle, etc. — the part of the solution stack on which customers depend but over which they exercise little control. I think the problem there is with the qualification process. This is a problem which several Linux distributions attempted to address with the United Linux specification, before SCOX decided that they’d be more successful in court than in code, but I think that branding effort was going awry before SCOX v. IBM.
The problem is the lack of discrete test cases leading to a synergetic whole. Instead of qualifying components, the whole blob is tested, so failures in discrete parts may not be immediately apparent.