| 1 |
On 2006.07.31 12:01, Giacomo Cariello wrote: |
| 2 |
> Roy Bamford wrote: |
| 3 |
> > Sunrise attempts to provide the next part of the school system |
| 4 |
> down. A school system should have a well defined syllabus, explicit |
| 5 |
> educational targets and an objective evaluation system. If Gentoo is |
| 6 |
> going to run a Gentoo School with courses for developers, that sounds |
| 7 |
> cool (and may relieve some burden from the developers mentors). |
| 8 |
|
| 9 |
Agreed. |
| 10 |
|
| 11 |
> Probably its objectors just want the project to formalize its methods |
| 12 |
> before official launch. |
| 13 |
|
| 14 |
Possibly. |
| 15 |
|
| 16 |
> |
| 17 |
> In regard to the "rule of thumb" you suggest to verify average |
| 18 |
> quality, |
| 19 |
> I wouldn't consider a reliable way to assess quality. I suppose that |
| 20 |
> part of responsibility of Gentoo towards their users consist in |
| 21 |
> "using caution" in their choices. Caution suggest that you cannot |
| 22 |
> suppose the long-term effects by measuring the current, limited, |
| 23 |
> 150-ebuilds version of this project. |
| 24 |
|
| 25 |
You can only examine now, what exists now. The long term effects can be |
| 26 |
assessed by repeated examinations, much like holders of ISO 9000 (a |
| 27 |
quality standard) undergo to retain their accreditation. |
| 28 |
Going off on a wild tangent for a moment perhaps sunrise and other |
| 29 |
overlays could be accredited by Gentoo using such a system of regular |
| 30 |
and surprise checks. |
| 31 |
|
| 32 |
> Also, it's better to debate this path now, rather |
| 33 |
> than wait until we realize Gentoo cannot handle at least minimal |
| 34 |
> safety of a 100000-ebuilds user-contributed overlay used by |
| 35 |
> hundreds/thousands of people and throw it in the dustbin with a "we |
| 36 |
> said that if it doesn't work, we kill it" quote. |
| 37 |
This in not a realistic claim - look at the way the official portage |
| 38 |
tree has grown with time and that changes made to cope with that |
| 39 |
growth. Your statement implies that sunrise starts out badly, gets |
| 40 |
worse but nobody notices for a long time. That's simply not realistic. |
| 41 |
Sunrise will evolve - like any other OSS project. |
| 42 |
|
| 43 |
I would expect sunrise to spawn both devs and ebuilds and to see the |
| 44 |
more popular ebuilds moved into the official tree as the dev population |
| 45 |
can cope. That's not much different from the present process, where |
| 46 |
ebuilds are in b.g.o. However b.g.o doesn't interactively encourage |
| 47 |
would be devs. |
| 48 |
|
| 49 |
> It would be a lack of respect towards the efforts of users that |
| 50 |
> contributed to it. |
| 51 |
Yes it would and it won't happen. |
| 52 |
|
| 53 |
> |
| 54 |
> - Giacomo 'jwk' Cariello |
| 55 |
> |
| 56 |
> |
| 57 |
> -- |
| 58 |
> gentoo-dev@g.o mailing list |
| 59 |
> |
| 60 |
|
| 61 |
Regards, |
| 62 |
|
| 63 |
Roy Bamford |
| 64 |
|
| 65 |
-- |
| 66 |
gentoo-dev@g.o mailing list |
Archives