[Sugar-devel] SoaS on XO issues for discussion

[Tomeu Vizoso]

On Fri, Apr 24, 2009 at 10:39, Martin Dengler <martin at martindengler.com> wrote:
> On Fri, Apr 24, 2009 at 10:35:05AM +0200, Martin Langhoff wrote:
>> On Fri, Apr 24, 2009 at 10:23 AM, Martin Dengler
>> <martin at martindengler.com> wrote:
>> > Of course.  That's why I'm asking these questions, since we know SL
>> > doesn't want to become a distro generator, but we're getting questions
>> > that push us in those directions.
>>
>> I didn't know SL was trying to avoid that role.
>
> I am of course not an authoritative source.  But that is my
> impression, yes (and it's bourne out by dfarning's reply, IMO).

Well, David gave his personal opinion (that I happen to share), but I
would like to note that SLs is going in the opposite direction right
now by developing Soas. When we were supposed to be making Sugar
stable, we were instead doing integration work in Fedora 12, much
before fedora people started to test the live bits. So an organization
with half a dozen of active, non-payed developers ended up doing the
job of an organization with hundreds of developers.

It's normal that people think "SugarLabs is shipping Soas and it
doesn't work on the XO, are they abandoning their only user base?".

I think we should be more careful in thinking where we spend the
resources or we won't be able to focus on what we think is important.
Also, as a community we should accept that the people that have done
task A in the past are the ones that can better appreciate the amount
of work A will take. We have right now lots of people saying what
should be done and how (and that's great!), but actually little people
doing it (not so great).

Similarly to how a software developer should double his initial
estimation of work for a task, someone who isn't a developer should
multiply by ten his own estimation. This also counts for software
maintainership, when a coder asks for some piece of code to be merged
in the mainstream sources, it's completely understandable that she
will underestimate the amount of work it will take to ship it and
maintain it afterwards. So please make an effort to understand the
maintainer's point of view.

Also think that the more busy developers are, the less time they will
have to participate in discussions like this one, so the chances of
bikeshedding increase. And I guess we want to be a community that does
awesome things, and not that only discusses awesome things.

Sorry about the rant, what I mean here is that what matters isn't if
SugarLabs officially declares that wants to replace OLPC or not. What
matters here is that whoever wishes to take up tasks left behind by
OLPC, can find here people that have the knowledge and the will to
help them reach their goals, and that work will end up helping lots of
real children.

Regards,

Tomeu


>> But there is _nobody
>> else_ really looking at how (whether!) the integration works. That a
>> desktop works or doesn't work is a job of integration.
>
> Hence the reason for me asking the questions.  I think SoaS, as a
> Fedora-derivative anyway, is the right place to do it.
>
>> So somebody has to do it.
>
> Indeed.  I don't think it's actually that big of a technical deal to
> *do* the things I asked about, but *supporting* them is a pretty big
> deal.  That's why I asked.
>
>> If SL does it, SL has a chance to ensure the quality. If SL doesn't
>> do it, who will? Maybe SL should do it until a focused team takes
>> over the distro side?
>
> Good subjects for discussion.
>
>> cheers,
>> m
>
> Martin
>
> _______________________________________________
> Sugar-devel mailing list
> Sugar-devel at lists.sugarlabs.org
> http://lists.sugarlabs.org/listinfo/sugar-devel
>
>