Re: Alpha remembrance day

Andrew wrote:
Bill Todd wrote:
etmsreec@xxxxxxxxxxx wrote:
Well, that's one view.

Can
you say "lack of applications"? Can you say "lack of operating systems
to run on it"?
Can you say "incompetent blowhard?" I think David addressed that latter
chimera adequately, and given that one of said OSs was Windows
(including support of x86 application binaries) I'd say that puts the
former to rest as well (not to say that VMS and Tru64 didn't have
adequate application support in their own right, of course).


I read this post with some amusement for a number of reasons.

Unfortunately, overlooking probably the best one - for which you would have had to have been looking in a mirror.


1. One of the key reasons for the decline in the Alpha business was
as the previous poster quite rightly stated the lack of software.

Ah, well - taking a week off from debating with uninformed morons can be very relaxing, but they're usually right where you left them when you come back.

For the edification of the educable (a group which history suggests does not number you as a member), the main reasons for the decline of Alpha were:

1. The Great Palmer Contraction - the transformation of DEC from a forward-looking aggressive (though not always ideally-focused) competitor to a multiple-amputee just trying to remain afloat. No one on the inside mistook this for anything but what it was, so it seems unlikely to have been missed by a lot of customers.

2. The infamous mid-'90s 'affinity' program (a Wes Melling Production IIRC), which encouraged the VMS community to switch to NT on Alpha - a great way to turn much of a loyal and robust customer base into a combination of skittish stalwarts and disenchanted ex-users, and to stall the growth of a leading OS without much compensating return (NT on Alpha never having come close to taking up the resulting slack).

3. The drastic loss of VMS development momentum at the end of the '90s (more handwriting on the wall for Alpha's primary OS for those paying attention). When 'supporting new hardware' becomes the main attraction of new releases rather than vigorous evolution of OS features in their own right to keep pace with the evolving OS competition, the new trend line is clear.

4. The Slough of Despond which Alpha entered immediately after Curly's ascension to the throne, when both performance increases and new releases slowed to a crawl and allowed other platforms to challenge what had until then been the unquestioned industry performance leader. What better way to get people wondering about the real level of commitment of a corporation to an architecture? Well, one could have been

5. The butchering of Win64 (and Windows in general) on Alpha in August, 1999 - another Curly brain-fart which managed to call Alpha's future into such question that the infamous "Commitment to Alpha" letter had to be written in an attempt to quell customer fears.

6. The Unix vacillations which first embraced MIPS as the corporate Unix platform and then haltingly switched to Alpha after that damage had been done. It took Digital Unix sales years to recover, but recover they eventually did: while Tru64 had only about 1/3 the market share of AIX, or HP-UX, or Solaris as of Y2K, it was growing far faster than any of them and threatening to overtake VMS revenues until

7. The Alphacide, of course - a fine way to kill such a resurgence.

The fact that DECpaq's Alpha system business remained one of its most profitable hardware franchises despite those first six major handicaps, and that its Tru64 business eventually attained robust growth, is a testament to Alpha's (and its OSs') continuing perceived strength and potential in the eyes of customers and their willingness to try to leverage them. Perhaps a 'lack of applications' was some part (along with the other problems noted above) of what kept Tru64 from growing at *more* than the 30% annual rate it was achieving shortly prior to the Alphacide, but it sure as hell wasn't contributing to anything resembling a 'decline' in either absolute or market-share terms.

- bill
.

Relevant Pages
Quantcast