1 |
m h wrote: |
2 |
> I'm not a gentoo dev (just a satisfied user), but I lurk on this list. |
3 |
> |
4 |
> I was at PyCon last month. I would estimate that about 40% of the |
5 |
> people there ran linux on their laptops. The most popular distros |
6 |
> were gentoo and ubuntu. (Not this is not a scientific study, just my |
7 |
> observations from talking to people there). While I was there the |
8 |
> person next to me starting hacking the ebuild classes to handles eggs |
9 |
> (so he could emerge turbogears). I talked to at least 3 others who |
10 |
> were running gentoo. I asked all of them if they had worked on |
11 |
> portage. Most said "No, the code is a little scary". (I'll concur |
12 |
> with that sentiment, as the code doesn't feel very pythonic). |
13 |
> |
14 |
> If you want to attract more developers (python people), a few things are needed: |
15 |
|
16 |
That depends on how they contribute, I personally don't want random |
17 |
python master bob contributing pieces to portage itself. Portage things |
18 |
are not necessarily as simple as people make them out to be. Even |
19 |
developers who know the code well make mistakes in adding and removing |
20 |
code. As solar once pointed out "the only man I trust to touch the |
21 |
resolver is Jstubbs." I realize thats a bit elitest...but at the same |
22 |
time...I am overly cautious ;) |
23 |
|
24 |
However we always accept patches and I think we get most of them |
25 |
critiqued, sometimes it may take an extra prodding mail or two. We |
26 |
usually don't implement your features for you though ;) |
27 |
|
28 |
> |
29 |
> * Portage documentation. How the innards work. There is very little |
30 |
> docs/comments in the portage code |
31 |
|
32 |
Someone has to write them; I have some of it done, it's been a longtime |
33 |
project that I've worked on off and on; I actually had more done last |
34 |
year and I know kutsuya did some as well. However these are not |
35 |
particularly interesting..and no one wants to document the 2.X branch. |
36 |
|
37 |
> * Unittests - without this how do I know that my change to portage |
38 |
> didn't break someone else's corner case |
39 |
No one is writing unittests for the 2.X branch |
40 |
> * Refactoring into a more pythonic style. Note that this is pretty |
41 |
> hard without unittests. |
42 |
See above :) |
43 |
|
44 |
> |
45 |
> Take this as a grain of salt, from an observer, who believes that |
46 |
> there are a lot of potential users (who know python), and who could |
47 |
> easily contribute, if the bar was lowered a bit. (Or steps were |
48 |
> provided to reach a little higher ;)) |
49 |
> |
50 |
> -matt |
51 |
> |