Re: AMD64 Status Update -- And Future Directions
On Fri, Feb 13, 2004 at 12:48:06AM +0100, Roland Fehrenbacher wrote:
> John> After all, someone that needs a lot of 32-bit apps could just install
> John> Debian i386 today and be done with it.
> John, your arguments might make more sense in a Debian centric world, but a lot
> of, if not most users (rather than maybe some developers) don't live in such a
> world, and still want to use and love Debian. So maybe it boils down to the
I am not some cloistered away monk of a Debian developer. I know about
the "real world". I work for a manufacturing company, where I
personally maintain the Debian and AIX servers that our Windows clients
connect to. I maintain somewhere around a dozen Debian boxen at work,
ranging from an old-but-solid P100 firewall to a multi-processor Xeon
machine that, thanks to Linux vserver, runs almost a dozen Debian
installations on a single machine (some stable, some unstable) doing
everything from mail serving to SGML generation.
*Not one* of these two dozen or so physical or virtual Debian machines
runs any software (except for our custom stuff, which is Python) not in
Debian main. Ergo, in this real, existing, non-hypothetical example,
binary compatibility is a complete non-issue.
While I absolutely know that there are those for whom it matters in
Debian, the significant portion of Debian users that are happy with what
is in Debian know that it's often completely irrelevant.
I am a Debian developer because I was first a Debian user and wanted to
improve Debian. I am still a Debian user.
> question: is Debian for their developers only or does it also listen to their
> users. I think the latter is the case. As I said, it is your time you invest,
> and there are certainly worse ways of where to put it. But please don't try to
> influence people with in my opinion simplistic and (sorry for the repetition)
> Debian centric arguments.
Yes, I offer my apologies right now for making Debian-centric arguments
on a list dedicated to porting Debian :-)
If I didn't care about Debian, I'd just install NetBSD on my AMD64
when it arrives. They can port their entire operating system in about
the time it took me to port less than a fifth of our userland. :-)
> John> Fortunately, Debian is an operating system build from source. A pure
> John> 64-bit AMD64 distribution will be just as useful as PowerPC, Alpha,
> John> Sparc, or any of the many other distributions we have that work
> John> without being binary-compatible with anything else.
> Comparing niche architectures like that to x86 is not valid. Backwards
I think few would agree that Sparc is a "niche architecture", and likely
even fewer would agree that PowerPC deserves that characterization.
PowerPC, I think, gets used on an even greater range of machines than
chips like the Xeon, thanks to its use in high-end IBM minis.
> compatibilty is a major issue (just in terms of number of users) for x86 but
> not for the platforms you listed.
Binary compatibility is in the eye of the beholder :-)
I have moved from x86 to PowerPC before, and from x86 to Alpha. I did
not care about binary compatibility then. Why should I when I move from
x86 to AMD64?
*nix is cross-platform and has been for literally decades. That is a
It means I ca ngo to the store and choose any laptop I want, asking only
"Can I install Debian on it?" I used to use a PowerBook (still do,
really.) Now I use a Thinkpad. Same software on each. They look,
behave, and act identically once booted.
> John> Look at PowerPC as an example. You can not run binaries from any
> John> other platform or operating system on a Linux PowerPC machine. Yet
> John> Linux on PowerPC is very successful, enjoying not just good support
> John> in Debian, but also fostering whole distributions dedicated to Linux
> John> on PowerPC.
You didn't really seem to like this particular point :-)
> John> That is completely false.
> Please reread my e-mail from Monday. A little cleanup (which should be done
> soon if possible) will fix those problems.
That will be good to see. Until then, what I have said corresponds to
what we have now.
> >> For most people, the opposite will be true. Concerning performance: A
> >> lot of performance is actually compiler dependent. And many benchmarks
> >> have shown that 32 bit code using the intel compiler is still faster
> >> than the corresponding 64 bit versions compiled with gcc or other
> >> compilers (Portland, ...).
> John> Debian does not use the Intel compiler, so I fail to grasp what
> John> relevance this has to a *Debian* distribution.
> Well, you mentioned performance right. Some people actually really need it,
> and they require the Intel compiler whether it is part of Debian or not.
Well then they are free to install it on their Debian i386 distro, if
they feel that is the fastest option anyway.
> >> This is really a flawed comparison. x86 emulation on the Alpha was just
> >> a marketing gag, and Alpha certainly wasn't out to replace i386. Anybody
> >> who
> John> No it was not. DEC never supported em86 in any way.
> Well, they certainly touted it's existence.
I checked and have been unable to find even one single statement from
DEC about it.
> John> Heck, they never even promoted *Linux* until after em86 was
> John> virtually dead anyway. It was solely an effort from Linux hackers.
> There was a relevance to this from the Microsoft port as well.
You are thinking of FX!32, which they did heavily promote. I used NT on
Alpha only very briefly (it sucked), and can't speak to it at all, and
completely fail to grasp its relevance, or the relevance of DEC's
marketing operation 5 years ago, to this discussion.
> Emulation in the kernel, and actually having the instruction set in the
> processor is still a big difference performancewise.
It is, but the situation in the userland is exactly the same. My point
is that once an architecture matures and gains acceptance, need for
binary compatibility with other archs dries up.
Again, this is not just some head-in-the-sand developer spouting
theories. I experienced this.
> John> Again, I repeat: if, because of pure64, enough people leave multiarch
> John> to make it wither and die, that means that multiarch was never
> John> popular enough to stand on its own, and the only reason those people
> John> supported it was because it was the only option to get to what they
> John> wanted.
> John> And that is NOT a Bad Thing.
> Short term popularity is not always the best indicator.
In this case it is, because I believe that multiarch on this platform
has no real long-term future :-)