Mailing List Archive

What to do for better support?
Hi,
I have been trying Gentoo Linux on sh(SuperH) target board.
I starting at stage3-sh4-2007.0 binary, and synced to newer portage
snapshot. Then successfully built stage1,2 and 3 using catalyst.
I'm impressed with the system running just like as PC (but speed).

I want sh to be supported more, and to be released officially in future.
Currently, it doesn't have stable profile, and no 2008.0 release
(not even as an experimental), though.

I wonder what should I do. After reading docs, I've found most of all I (as a
non-gentoo-developer) can do is testing and sending bug to Bugzilla or so.

I've got nothing to BTS so far, because my trial was so successful.
So, maybe I should...
- Test more packages of "sh" to find and report bugs?
- or Test "~sh"s to change them to "sh" ?, or "" to "~sh" ?
- or Test using (or providing?) binary at place like tinderbox?
- Any others? (perhaps, seeking arch maintainer...)

Please give me hints.
/yoshii

--
gentoo-dev@lists.gentoo.org mailing list
Re: What to do for better support? [ In reply to ]
Hi,

Takashi Yoshii <yoshii.takashi@renesas.com>:
> I've got nothing to BTS so far, because my trial was so successful.
> So, maybe I should...
> - Test more packages of "sh" to find and report bugs?

Yes! Always. Maybe test stabilisation requests for packages. See
the archtester programs of amd64 or x86 how that works.

> - or Test "~sh"s to change them to "sh" ?, or "" to "~sh" ?
> - or Test using (or providing?) binary at place like tinderbox?
> - Any others? (perhaps, seeking arch maintainer...)

You could become a developer to support the SuperH platform, but you
should prove yourself before.

V-Li

--
Christian Faulhammer, Gentoo Lisp project
<URL:http://www.gentoo.org/proj/en/lisp/>, #gentoo-lisp on FreeNode

<URL:http://www.faulhammer.org/>
Re: What to do for better support? [ In reply to ]
On Thu, Jun 12, 2008 at 05:20:47PM +0900, Takashi Yoshii wrote:
> I want sh to be supported more, and to be released officially in future.
> Currently, it doesn't have stable profile, and no 2008.0 release
> (not even as an experimental), though.
Seeing your employer's domain, can you offer pointers to
reasonably-priced and useful SH boards/systems?

The ARM world has recently seen the TinCan 'Nail' board, which is pretty
damn good (both solar and I have one). Embedded PPC has the Efika.
Embedded MIPS I'm less certain of, I know I have a MyCable XXS1500
(Au1500 core), but I don't know beyond that.

There's a really low-memory SH box available for developer access:
http://www.gentoo.org/proj/en/infrastructure/dev-machines.xml
However, I don't really know if it's sufficient for building releases or
doing more testing (I suspect not).

> I wonder what should I do. After reading docs, I've found most of all
> I (as a non-gentoo-developer) can do is testing and sending bug to
> Bugzilla or so.
Pretty much all of your suggestions are good, plus seeing about more
hardware maybe.

--
Robin Hugh Johnson
Gentoo Linux Developer & Infra Guy
E-Mail : robbat2@gentoo.org
GnuPG FP : 11AC BA4F 4778 E3F6 E4ED F38E B27B 944E 3488 4E85
Re: What to do for better support? [ In reply to ]
Takashi Yoshii <yoshii.takashi@renesas.com> posted
4850DC5F.7090800@renesas.com, excerpted below, on Thu, 12 Jun 2008
17:20:47 +0900:

> I've got nothing to BTS so far, because my trial was so successful. So,
> maybe I should...
> - Test more packages of "sh" to find and report bugs?
> - or Test "~sh"s to change them to "sh" ?, or "" to "~sh" ?
> - or Test using (or providing?) binary at place like tinderbox?
> - Any others? (perhaps, seeking arch maintainer...)

Hi. I know little about SuperH personally, but perhaps the following
will be helpful general orientation and pointers toward more.

AFAIK, the Gentoo sh arch is severely understaffed ATM, and that's an
understatement. (I /believe/ most of the work has been done by one
person (possibly two), who has other Gentoo responsibilities as well, not
to mention the fact that Gentoo devs are all volunteers, so he has a day
job too.) That is why as you mentioned it has no current release or
stable profile.

Thus, if you have reasonable skills and a decent amount of time to
dedicate in the medium term, preferably two years or so minimum so the
mentoring time investment has some return, what Gentoo on that arch could
really use would be a dedicated arch developer. Yes, the personal
investment isn't trivial by a long shot, and it's entirely possible you
don't have that sort of time available to dedicate, but if you do and
want to, read over the dev handbook and seriously consider starting the
mentoring process.

If you don't have that sort of time or skills available, even a shorter
term lower time investment could be very beneficial for the arch. A
number of archs have what's called an arch tester position. This is a
fairly well defined position where you aren't a Gentoo dev but you work
very closely with one or more Gentoo devs, with them doing the final
commits based on your testing and results. Many such closely cooperating
users have ultimately become devs themselves, while others have remained
at the AT position by choice, often due to lack of time and resources to
be a full dev, but wanting to give what they can. While becoming a dev
has a purposefully longer testing and mentoring process, becoming an AT
is much simpler, altho there's usually (archs differ in AT procedures a
bit) a general ebuild quiz involved.

Then there's folks like me, who spend a lot of time on the various
channels/groups/lists/forums/etc (often being more comfortable in some
than others, I'm not an IRC person for instance and rarely do forums, so
spend most of my time on the lists/groups) trying to help out as best we
can, but who haven't taken a formal position of any sort. Of course,
that may lead to more later, or not. Personally, I don't consider myself
really skilled enough at this point to be a dev (tho I've been asked),
tho it's also a matter of not yet having really decided to make the
commitment and buckle down and do it.

So whatever level you are comfortable and chose to participate at,
welcome! Gentoo can put you to work, particularly with your alternative
arch experience! =8^)

Here's the developer handbook, which can be a very good place to start
even if you aren't directly targeting that initially, and should at least
answer some questions you no doubt already have, as well as provide you
some idea of what's involved if you do wish to head toward devhood.

http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml

If you're looking more at AT, that's to some degree defined by the arch
and arch-team in question, and certainly the dynamic on a small arch such
as sh will be rather different than on x86 or amd64, but here's the x86
arch-tester's FAQ, to give you an idea, at least.

http://www.gentoo.org/proj/en/base/x86/arch-testers-faq.xml

Other than that, I'd suggest you contact the dev(s) for your arch
directly. The one listed for Sh is Mike Sterrett, aka mr_bones_ . You
can of course make yourself immediately useful by contributing bugs and
patches, etc. Many people find the bug day project very helpful at
introducing themselves to the larger community, for instance.

--
Duncan - List replies preferred. No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master." Richard Stallman

--
gentoo-dev@lists.gentoo.org mailing list
Re: Re: What to do for better support? [ In reply to ]
Thank you for your reply.

Christian Faulhammer wrote:
> Yes! Always. Maybe test stabilisation requests for packages. See
> the archtester programs of amd64 or x86 how that works.
Hmm, archtester seems to be good for my first aim as a role.
I will be on IRC as the faq says( Though, No one in #gentoo-sh ;).

> You could become a developer to support the SuperH platform, but you
Yes. Studying gentoo development system, I really wish I were a developer.

> should prove yourself before.
Sure. So, I am not going to ask you how to become a developer.

BTW, I'm happy since gtk+ build has failed just now.
It is strange to say failure makes me happy, though..
/yoshii
--
gentoo-dev@lists.gentoo.org mailing list
Re: What to do for better support? [ In reply to ]
Hi,

Takashi Yoshii <yoshii.takashi@renesas.com>:
> Christian Faulhammer wrote:
> > Yes! Always. Maybe test stabilisation requests for packages. See
> > the archtester programs of amd64 or x86 how that works.
> Hmm, archtester seems to be good for my first aim as a role.
> I will be on IRC as the faq says( Though, No one in #gentoo-sh ;).

Be aware that the sh platform people may not be too talkative, but
nonetheless go on, test packages and announce you did on the
corresponding bugs.

> > should prove yourself before.
> Sure. So, I am not going to ask you how to become a developer.

It is really simple. Many people (including myself) started off as
archtesters.

V-Li

--
Christian Faulhammer, Gentoo Lisp project
<URL:http://www.gentoo.org/proj/en/lisp/>, #gentoo-lisp on FreeNode

<URL:http://www.faulhammer.org/>
Re: What to do for better support? [ In reply to ]
Robin H. Johnson wrote:

> On Thu, Jun 12, 2008 at 05:20:47PM +0900, Takashi Yoshii wrote:
>> I want sh to be supported more, and to be released officially in future.
>> Currently, it doesn't have stable profile, and no 2008.0 release
>> (not even as an experimental), though.
> Seeing your employer's domain, can you offer pointers to
> reasonably-priced and useful SH boards/systems?
>
> The ARM world has recently seen the TinCan 'Nail' board, which is pretty
> damn good (both solar and I have one). Embedded PPC has the Efika.
> Embedded MIPS I'm less certain of, I know I have a MyCable XXS1500
> (Au1500 core), but I don't know beyond that.
>
> There's a really low-memory SH box available for developer access:
> http://www.gentoo.org/proj/en/infrastructure/dev-machines.xml
> However, I don't really know if it's sufficient for building releases or
> doing more testing (I suspect not).
I did build and test a couple of my packages there. It is definitely too
slow and hasn't enough RAM.


--
gentoo-dev@lists.gentoo.org mailing list