Re: [gentoo-desktop-research] test

[ sorry my mails are not yet signed, but will be soon ]

Paul de Vrieze <pauldv@g.o> said:

> On Thursday 09 October 2003 01:08, Luca Barbato wrote:
>>
>> I suggest to think about the problems first, see how the other distro
>> solved them and then find a solution that takes in account what the
>> other did but also what they fail to accomplish.
>>
>> so Research fields should be, IMHO, problems&solution from the others,
>>
>
> I do not think that we should focus on what other distro's do. We should focus
> on identifying problems that are there. After such a problem is identified we
> should try to determine whether we think this problem needs to be solved by
> gentoo (keeping the gentoo and desktop-research goals in mind). After that we
> need to look at the solution to the problem. At this stage comes looking at
> solutions by other distros.

I agree with that process.

>
> This solving process has a number of stages. The first one being the
> identification of possible solutions, and their costs (how much work are
> they, and how "beautiful" are they etc.). Then we choose one solution to
> implement and test. This involves trying to get approval by the people from
> the projects that will be involved when it is actually put into gentoo. After
> the test is successful, we are going to put it into the main tree, and put it
> into action. Putting it into action involves making sure that our solution
> will be maintained, if necessary by creating a team to do that maintenance.
>
> However before we can think of finding problems, we first need to have a
> policy on the following points:
> - Which problems do we handle? We only handle desktop problems, but when does
>   a problem qualify as being desktop.

For me, a desktop problem is not always a problem wich involve window managers,
or graphical interface.

I would tend to say that desktop means workstation for private use, client
side. That is, everything UI oriented. That includes windows manager, but also
config programs, installation problems, software installations, and simple
(client-side personal use) system managment. By client side I mean everything
that does not use servers. So setting internet/lan connection is desktop, but
setting a ftp server is not.

Maybe you have an other vision?


> - How do we handle problems? Basically, do we follow the process described
>   above, or some other process.

I think the above process is not bad at all. So that could be :

- define the problem
- look if it has not being resolved, or if nobody is working on it, inside gentoo
- look for similar problem inside gentoo
- look for similar problems outside gentoo (official software author, other
  distro)
- find a way to solve it, by discussing it here or #irc (taht would be the
  Evaluation part of gerrynjr)
- allocate people on it, plan the solving time.
- write proposal/recommendations to gentoo main
- fo the main inclusion, I have no precise ideas.

> - What are the conditions for a solution to be acceptable.
doesn't break things, gentoo way of thinking/developping compliant, doesn't
break the desktop taste.

The gentoo desktop taste is something we might want to describe. It can be
vanilla desktop plus some little not much visible features, or (waht I
prefer), vanilla desktop with pluggable features that changes the
appearence/feel. We'll have to make technical coice on it

> - Will we have a seperate cvs testing tree (overlay)?

I don't know, what do you think?

> - How are we going to make sure that the solutions get out of the testing
>   tree, and in to the main distribution? Will we use GLEP's?

why not, but maybe new ebuilds/doc, with official maintainers could suffice?


>
> Paul
>
> ps. I know it is nicer to think of solutions than of problems, and nicer to
> think of problems than of procedures, but first things should go first. And
> procedure comes first.

100% agreed on that, but I don't want to enforce anybody :)

-- 
dams

--
gentoo-desktop-research@g.o mailing list