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

[lists at openmailbox.org]

On Fri, 13 Mar 2015 06:28:17 -0700 (PDT)
johnwallace4--- via Info-vax <info-vax at rbnsn.com> wrote:

> I've no wish to argue. I'll be happy to be better informed than I was.

I'm sorry I didn't answer your question before. I'm not in a position to
inform anybody on Intel or many other things. I just wanted to express my
distaste for the platform. I didn't answer you because you're asking what
could be added and I don't think that's the way to approach this.

> I tried to ask here "what *specifically* do you see as missing from
> today's x86-64?" (note: you say Intel. I *assume* you mean x86-64, and
> not iAPX432 or 4004 or any other non-x86-64 stuff. But that's an
> *assumption* on my part.

Quality, sanity, and simplicity are what's missing. We can't add that back
in, nobody can. That's why I don't like the platform. The ISA and the OSs
that run on it are a house of cards built on sand. I don't think any amount
of money or engineering is ever going to get over that.

The instruction set architecture is abominable. Encoding and decoding are
atrocious. It's hard to find two assemblers that create the same output for
the same input. Intel never provided software support the way they should
have and now there are a bunch of free assemblers, all mostly incompatible
and all generating different code.

The 32 bit Intel ISA didn't have enough registers and the few it had were
dedicated to various instructions and were not generally useful throughout
a large piece of code. It is annoying and limiting compared to other
available platforms.

64 bit support was and is a disaster. 32 bit and 64 bit code can't coexist.
The ABI is not upward compatible. You can't use 64 bit registers in a 32
bit piece of code running on 64 bit hardware. You can't call 32 bit code
from 64 bit code and vice versa and you can't combine code written in both
ABIs in one executable.

If you have 32 bit code you have to have a 32 bit libc and all the other
necessary libraries and for 64 bit code you have to have a 64 bit libc and
other libraries. Why all the duplication? It's a broken implementation that
nobody thought ahead 5 minutes implementing. And then they came up with X32
that still doesn't solve the problems. Intel and their designs are a
complete failure and incur a lot costs with endless compatability
breakages, expense of twice as much code to run legacy apps, and
recompiling things just to get 32 and 64 bit versions. They have never made
a smooth transition on anything.

> I will happily agree that the original 8086 and the stuff that went with
> it (in software and in hardware) was a nightmare in many ways. But then
> AMD64 made many improvements which kept the market sufficiently happy
> that Intel had to copy them and abandon their home-grown IA64 stuff.

Who is the market? Did they keep Microsoft happy? Ok, I'll give you that.

> What is it *specifically* about "other stuff" that makes "other stuff"
> high end and AMD64 (and compatibles) not high end? [Preferably without
> including IBM-style mainframes, but...]

:-)

> Are those differences really sufficient to put off potential VMS
> customers?

It depends what you're used to. I'm way too new to VMS to answer that but I
can tell you from working on a platform that has none of the issues I
mentioned about Intel that it puts me off and always has.

-- 
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