[illumos-Discuss] SPARC

[Matt Lewandowsky]

I'm not advocating anything like what you're talking about. Right now, 
Illumos needs to be a useful set of code. The quickest way to get there is 
to ensure that you have at least one fully-functional platform, then bring 
the other(s) back to parity.

At this point, there's not not much divergence from Oracle's tree, and 
therefore little reason to suspect lots of breakage.

Once the latest bits are fully tested and people have a good way to upgrade 
from ON 134 -> Illumos on x86, I'll personally start looking at the SPARC 
issues myself. As I said, I have an interest in Illumos on SPARC, but it's 
still early enough that it's not useful to work on both platforms 
simultaneously quite yet.

This is one of those situations where picking your battles is the best 
approach, and x86 is the battle that makes most sense first.

--Matt

-- 
Matt Lewandowsky
Big Geek
Greenviolet
matt at greenviolet.net   http://www.greenviolet.net
+1 415 578 5782  (US)  +44 844 484 8254 (UK)
+86 755 33944325 (CN)  +39 02 8715 7053 (IT)

--------------------------------------------------
From: "Joerg Schilling" <Joerg.Schilling at fokus.fraunhofer.de>
Date: Thursday, 19 August, 2010 1:28
To: <matt at greenviolet.net>; <garrett at damore.org>; <f.middleton at apogeect.com>
Cc: <discuss at lists.illumos.org>
Subject: Re: [illumos-Discuss] SPARC

> "Matt Lewandowsky" <matt at greenviolet.net> wrote:
>
>> I agree with Garrett on "x86 first, then see what breaks on SPARC", at 
>> least
>> while Illumos is young. The project is currently at a point where it 
>> needs
>> to be usable on at least one architecture. Unfortunately, the resources 
>> to
>> keep both in sync aren't yet there, I don't think. That means a decision
>> needs to be made as to which gets priority. I think everyone will agree 
>> that
>> x86 is the no-brainer answer.
>
> Starting this way may cause a similar desater as we see with the include 
> files
> with interfaces from the kernel. In 2003 the Linux kernel developers 
> started to
> destroy their include files (making them inconsistent when included from a 
> user
> level application). If they had been chided for this early enough, there 
> had
> been a chace to avoid the resulting problem.
>
> I strongly vote for fixing bugs early.
>
> This however means that someone needs to install a Sparc with the right sw 
> base
> to be able to compile and report.
>
> Jörg
>
> -- 
> EMail:joerg at schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 
> Berlin
>       js at cs.tu-berlin.de                (uni)
>       joerg.schilling at fokus.fraunhofer.de (work) Blog: 
> http://schily.blogspot.com/
> URL:  http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily
>