1 |
Thomas Cort wrote: |
2 |
> There have been a number of developers leaving Gentoo in the past 6 |
3 |
> months as well as a number of news stories on DistroWatch, Slashdot, |
4 |
> LWN, and others about Gentoo's internal problems. No one seems to have |
5 |
> pin pointed the problem, but it seems glaringly obvious to me. We |
6 |
> simply don't have enough developers to support the many projects that |
7 |
> we have. Here are my ideas for fixing this problem: |
8 |
> |
9 |
> - Cut the number of packages in half (put the removed ebuilds in |
10 |
> community run overlays) |
11 |
|
12 |
I doubt this'll work. I sorta see the portage tree like a starfish -- cut it in |
13 |
half, and within time, you get two starfish. Cut it into three parts, and you |
14 |
eventually get three starfish. You wind up back at square one with double the |
15 |
trouble and none of the fun. |
16 |
|
17 |
|
18 |
|
19 |
> - Formal approval process (or at least strict criteria) for adding |
20 |
> new packages |
21 |
|
22 |
This might work, but it depends on what criteria/process, and how well its enforced. |
23 |
|
24 |
|
25 |
> - Make every dev a member of at least 1 arch team |
26 |
|
27 |
This won't work -- especially if the dev lacks access to the hardware. Some |
28 |
arches are so complex, you need several types of hardware. In mips, for |
29 |
example, if a dev's got access to a low-end box like an Indy or an O2, then |
30 |
letting them help out on basic keywording on common packages probably won't |
31 |
hurt, but it would be much better if they had access to say, more than one type |
32 |
of mips hardware (say, an Octane, and a Cobalt). |
33 |
|
34 |
Also, not every dev would want to have to maintain another box of some |
35 |
obscure/strange arch. It's opposite in my case -- I have 1 x86 box running |
36 |
Linux (not counting my main desktop since its in windows), and everything else |
37 |
is an SGI box (or my one cobalt). I've got spare parts lying around to build |
38 |
two more functional x86 systems, but I've never seen a need to put'em together |
39 |
and run them continuously. |
40 |
|
41 |
|
42 |
> - Double the number of developers with aggressive recruiting |
43 |
|
44 |
This can become a slippery slope real fast. |
45 |
|
46 |
|
47 |
> - Devs can only belong to 5 projects at most |
48 |
|
49 |
I can see this having its uses, but this is more of a personal thing on a |
50 |
per-dev basis. |
51 |
|
52 |
|
53 |
> - Drop all arches and Gentoo/Alt projects except Linux on amd64, |
54 |
> ppc32/64, sparc, and x86 |
55 |
|
56 |
Uh, no? Although we sometimes seem as inactive as hell, mips is very much an |
57 |
alive arch. We're a tad guilty of going off and doing our own thing sometimes, |
58 |
but then again, most of us are guilty of that at some point in our devship. |
59 |
|
60 |
I would instead opt for more interaction among archs, probably through dev |
61 |
sharing and such. sparc and mips share several developers (or did, I think I'm |
62 |
one of the few left), and encouraging more publicity for the lesser archs. I |
63 |
occasionally post an announcement about some neat new whizbang thing I do (like |
64 |
the X LiveCD for SGI systems I might post about tomorrow), and though I rarely |
65 |
see a response, I feel it gets the word out. |
66 |
|
67 |
|
68 |
> - Reduce the number of projects by eliminating the dead, weak, |
69 |
> understaffed, and unnecessary projects |
70 |
|
71 |
Depends on the definition of "unnecessary". |
72 |
|
73 |
|
74 |
> - Project status reports once a month for every project |
75 |
|
76 |
Hmm, could be useful. Depends on whether one defines a report as needing to |
77 |
match some obscure DoD specification, or whether a simple paragraph or two works |
78 |
fine. |
79 |
|
80 |
|
81 |
|
82 |
--Kumba |
83 |
|
84 |
-- |
85 |
Gentoo/MIPS Team Lead |
86 |
|
87 |
"Such is oft the course of deeds that move the wheels of the world: small hands |
88 |
do them because they must, while the eyes of the great are elsewhere." --Elrond |
89 |
-- |
90 |
gentoo-dev@g.o mailing list |