[Info-vax] Current VMS Usage Survey

[Bill Gunshannon]

In article <l7pu1l$3eg$1 at lnx107.hrz.tu-darmstadt.de>,
	m.kraemer at gsi.de (Michael Kraemer) writes:
> In article <bgb9gjFb5thU2 at mid.individual.net>, bill at server3.cs.scranton.edu
> (Bill Gunshannon) writes:
>  
>> Would you want to try to compete when you were the only one not getting
>> a discount?
>  
> Well, firstly I don't buy that story with x86 discounts as the main
> reason to adopt the Itanic.
> And secondly, even if true, if everybody gets a discount, what's
> the difference?

Why is this so hard to see?  PC margins are razor thin.  If I am the guy
not getting a discount then my PC's are more expensive and I will be the
first man out of the game.

>  
>> I doubt there was ever hope. I think it was all the corporate "must save
>> face" attitude.  What I am suprised about is that the stockholders didn't
>> revolt.  The fiction being handed to them must have been a real piece of
>> work.
>  
> At some point there surely was. Of course VLIW/EPIC is a crazy idea,
> but at some point, one has to give crazy ideas an opportunity to materialize.
> Of course, if they don't materialize after three, not even five years,
> one could at least ask some questions.
> 
> Another flaw inherent in the Itanic adventure was that intel's and
> HP's interests were rather different.
> intel didn't desperately need a new chip, they could have continued
> forever with x86. Add another 32bits to its address registers
> to cope with RAM beyond 4GB, once it becomes affordable.
> HP otoh needed a replacement for PA before it would become too
> expensive to maintain it. 

Intel had others that have had even more success than Itanium.  Like
Arm.  :-)

> 
>> > 
>> > Totally absurd.
>> > Back in 2001, Itanic had a bright future, 100% industry support,
>> 
>> Maybe in the marketing world.  I never read anything technical that
>> would have made me want to use it. 
> 
> well, you are not "the market".

No, I'm not.  I'm a techie. (And, no, I am not insulted by the epithet!)

> 
>> Even the "smart compilers" that 
>> were going to be able to modify code based on profiling information
>> (did anyone ever actually do this?) struck me as a cute but improbable
>> idea.
>  
> Yep, yet another flaw in the Itanic concept.
> But it was OK to give it at least a try.

Anyone dealing with commercial software would have known from the start
that no company was going to be recompiling custome versions of their
software for every customer.  Without that, the feature is meaningless.

>  
>> > including M$. Alpha had nothing of that kind, M$ support was
>> > cancelled in 1999 or 2000, iirc.
>> 
>> I never really understood that.  have to wonder if it wasn't at the
>> request of people outside MS as I saw Alpha machines running Windows
>> and it worked quite well.  I would be very happy if I could get the
>> OS (it was NT when I saw it, I don't think it went to 2000) to run
>> on my one remaining Alpha just for fun.
>   
> The problem is not whether it runs well or not,
> but rather whether if enough people see any sense
> in buying a Windoof/Alpha box when a Windoof/intel box
> is cheaper, has way more hardware and software support
> and doesn't run much slower for most practical purposes.
> The sales numbers of those boxes where marginal
> within the already marginal sales numbers of Alphas altogether.

Yeah, butt he marginal sales may well have been due to the fact that it
got killed before third parties even had a chance to get products out
the door.

> 
>> > Going with Alpha would have been entirely irrational.
>> > Just as if they had chosen the 68K.
>> 
>> Funny you should mention that as the original plan for the IBM PC was
>> the M68K, not the 8088.
>  
> The world would be a better place if IBM had made a wiser decision.

Wisdom had nothing to do with it.  Ethics (or the lack thereof) did.

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999 at cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>