[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gnewsense-dev] Yeeloong arrived, let's hack
From: |
avr |
Subject: |
Re: [Gnewsense-dev] Yeeloong arrived, let's hack |
Date: |
Fri, 6 Feb 2009 01:58:19 +0100 |
User-agent: |
Mutt/1.5.13 (2006-08-11) |
On Thu, Feb 05, 2009 at 11:14:23PM +0100, Robert Millan wrote:
> On Thu, Feb 05, 2009 at 10:57:49PM +0100, avr wrote:
> > > My Yeeloong arrived today, and I'll get to hacking ASAP. Please can
> > > someone
> > > give me info in what remaining tasks are there (specially those not
> > > currently
> > > being worked on) so I can pick one?
> >
> > For Orchid you actually sum them up nicely below.
> >
> > [...]
> >
> > I think the ubuntu installer/live-image needs a lot of graphical bits and
> > large parts of
> > the gnome desktop, which still need to be compiled. This goal shouldn't be
> > far off
> > apart from missing a kernel.
>
> You're actually porting Ubuntu to MIPS? I assume you're aware that porting a
> huge codebase to an architecture it didn't support before requires a
> significant amount of QA.
I'm using glibc, gcc and binutils from debian, which makes most of the pain go
away. And
most of the software did actually support the architecture when it was still in
debian. So
the only problematic packages will be those where careless ubuntu patches break
stuff
specifically for mipsel. So far I've compiled 100's of them with very few
problems.
The Q&A is a very valid concern though, but if gNewSense intends to grow there
will
be a time where upstream bugfiling won't be enough anymore, and then we'll need
to figure
out some way to maintain it anyway.
> It's fine if you want to work on that, but I doubt any of us has the resources
> to even find all those new MIPS-related runtime issues, let alone fix them.
These issues are far and between, mostly I guess due to Ubuntu's debian
heritage. If not
for the kernel, I'd feel secure in claiming to be running gnewsense-mipsel
right now
already on the yeelong.
friendly,
Andreas