[Info-vax] OT: news from the trenches (re: Solaris)

[lists at openmailbox.org]

On Wed, 11 Mar 2015 16:03:28 -0700 (PDT)
"johnson.eric--- via Info-vax" <info-vax at rbnsn.com> wrote:

> On Wednesday, March 11, 2015 at 11:40:06 AM UTC-4, li... at openmailbox.org
> wrote:
> 
> > > What's your alternative to x86-64 here?   How much will the "special 
> > > kick booty chip" cost to produce and integrate?
> > 
> > There's a lot more wrong with Intel than can ever be fixed. I agree a
> > good processor would be nice from a programmer's POV and from a healthy
> > market POV.
> 
> The idea that x86 is "inherently broken" and "wrong" is kind of silly.

I have not ever heard an effective argument made for that point except from
people on the Intel payroll, or people who have only coded on Intel and
have no significant experience with other ISAs.

> The reality is - only a handful of people really have to deal with the
> instruction set.

Fine, but that is the subject of discussion here so it's relevant.

> Yes it has become complicated, but 30+ years of success against a forever
> changing landscape will do that.

It has nothing to do with how old it is. It has to do with how badly it was
designed and how they kept compounding that in a patchwork way instead of
fixing it. It is a good example of marketing uberalles and a strong proof
that if you throw enough money at something infeasible, broken, and horribly
overcomplicated it can still run like stink.

> It's the price of success.

No it's not. There is at least one other architecture that has been around
much longer than Intel and it evolved without breaking old code. And who
knows how many more architectures could have been successful if Windows
would have run on them an not Intel?

64 bit support on Intel and the software that runs on it was not well
though out, to say the least. The problem had already been solved but
Intel's severe NIH made them stumble into complexity and half-baked
solutions again. Now they're trying to fix it with X32. Intel never learned
you can't fix things retroactively and they never learned to design things
properly from the beginning. They rely on the fact they have enough R&D
budget to make broken designs workable.

> I don't see how you can get around that. Even a freshly crafted
> instruction set (this time done right! *pounds fist on the table*) will
> have the same kind of warts after 30 years of kicking ass and taking
> names. That's even assuming your better instruction set lasts that long. 
> 
> I'm sure even the designers of Itanium thought they had it right too.

It doesn't matter what they think. Intel has been proven over the long
run not to be able to do anything elegant. Complexity is the product of sick
minds and they have it in spades. Their stuff is ugly and a shambles,
Itanium is perhaps even worse than x86 in that regard. It only "succeeds" by
the sheer force of the bigger hammer.

-- 
Please DO NOT COPY ME on mailing list replies. I read the mailing list.
RSA 4096 fingerprint 7940 3F02 16D3 AFEE F2F8  ACAA 557C 4B36 98E4 4D49