Gentoo Archives: gentoo-dev

From: George Shapovalov <george@g.o>
To: gentoo-dev@l.g.o
Subject: [gentoo-dev] Scientific Gentoo reorg
Date: Sat, 17 Jun 2006 18:08:39
Message-Id: 200606171959.12144.george@gentoo.org
1 This seems to have never made it to the list, resending..
2
3 Hi everybody.
4
5 Sorry for crossposting, but many people whom I want to catch with this are not
6 on gentoo-science. For the same reason lets keep this initial discussion
7 here, on -dev. If we need to expand, lets take it to the gentoo-science, but
8 then I would expect everybody interested to sign up ;).
9
10 Gentoo Science has been relatively quiet, even though we are making a steady
11 progress on many fronts, but lately a few things has caught my attention and
12 I think we can do better if create a bit more structure.
13
14 At present we count 10 categories containing 309 packages (wow! Considering
15 that it all started with some 20 packages I put in sci back when we did not
16 have a two-tier categories yet, quite impressive :)). However looking at
17 herds.xml I see only 10 devs listed, which just cannot be true (considering,
18 according to my experience with bugzilla, that majority of these packages are
19 actually maintained). The real situation is that many devs are quietly
20 supporting their own packages but are reluctant to join the sci team
21 officially. And it is this situation that I want to address.
22
23 I see one reasonable rationale for this relctance to join: people are
24 afraid "to get too much on theirs hands" by signing up. In reality the
25 gentoo-science mailing list is really low on traffic (quite a relief in
26 present times ;)), but nonetheless the sheer amount of packages may be
27 frightening. Therefore I am proposing to recognize the fact that we
28 (Scientific Gentoo project) became big and act accordingly:
29
30 1. We need more herds. The easiest possibility is to simply split them
31 accordingly to categories. However, with 10 categories, this may be an
32 overkill, or, with some categories having >50 packages an underkill, or
33 simply may not correspond to maintenance reality (it will definitely fail on
34 sci-libs for example). It is hard to tell without seing who does what, so I
35 am going to ask for some feedback here (see below).
36
37 2. Should we create some subprojects? This really will have to be discussed in
38 more detail when people respond and join corresponding teams I guess.
39 However, while at it we may as well become a top-level project of our own.
40 Right now Scientific Gentoo is under Dektop, which is at the very least
41 strange (but all the other options back then were even less fitting..)
42
43 3. Mail aliases. Right now we have sci@g.o, which we should keep as an all
44 encompassing alias for announcements or, well, I am not sure yet what else,
45 but time will tell. In addition we should create new ones, one per herd. If
46 by chance there are people interested in seeing *all* the bugs (which I
47 somehow doubt, but theoretically?), we can reuse sci@g.o for that as well..
48
49 So, right now I would like to ask for the feedback on the following:
50
51 Q1) I would like to hear about the reasons why people are afraid to join the
52 sci team. You may respond to me personally or raise it on the list, but
53 please let me/us know about the problems in any case, so that we may address
54 them!
55
56 Q2) Please let me know if you are supporting or occasionally touching some
57 package under sci-* and, assuming we create more herds, which herd it should
58 belong to (just make it up as you see fit right now) and whether you would be
59 willing to add yourself to the alias of that herd or join some subteam if we
60 create one. I will collect the responces and then compile a proposal for the
61 new structure.
62
63 Q3) Not relevant to this restructuring, but always usefull: if you know of
64 some package that you think should really go under sci-something, please let
65 us know!
66
67 And to finish it all up :)
68 Q4) If you are a user but would like to be involved more actively, or you have
69 to run that particular package for your work but it sits in bugzilla for ages
70 and no developer seems sensible enough to take it up, please let us know too.
71 Best of all - subscribe to that gentoo-scie mailing list and ask somebody to
72 mentor you. By the time it will be over we should have a new structure, so
73 you won't end up with the whole 300+ sci packages on your hands (this was
74 holding some people with whom I discussed it too).
75
76 This should be enough to start with, so, bring it on! :)
77
78 George
79 --
80 gentoo-dev@g.o mailing list