1 |
Nathan L. Adams posted <43074CD2.7080505@××××.org>, excerpted below, on |
2 |
Sat, 20 Aug 2005 11:31:30 -0400: |
3 |
|
4 |
> Ciaran McCreesh wrote: |
5 |
>> On Sat, 20 Aug 2005 10:03:18 -0400 "Nathan L. Adams" <nadams@××××.org> |
6 |
>> wrote: |
7 |
>> | Ciaran McCreesh wrote: |
8 |
>> | > No, I'm saying that having a 'team lead' throw some arbitrary stamp |
9 |
>> | > of approval upon bug closures is worthless. |
10 |
>> | |
11 |
>> | So you're problem isn't with the peer review I'm proposing but instead |
12 |
>> | quality of work of the team leads? |
13 |
>> |
14 |
>> Not at all. I'm saying that a) most 'team leads' will not do proper |
15 |
>> checks because they don't have time to and b) the limited time that |
16 |
>> 'team leads' have is better spent elsewhere. |
17 |
> |
18 |
> I really am curious here: |
19 |
> |
20 |
> a) What are the team leads spending most of their time on? |
21 |
> b) What is more important than improving the code? |
22 |
|
23 |
Not to sound harsh, but... |
24 |
|
25 |
I think what many users lose sight of is the fact that 100% of the Gentoo |
26 |
developer team, INCLUDING the team leads, are unpaid volunteers. |
27 |
|
28 |
What most of the team leads, and what everyone else involved, spend most |
29 |
of their time on, therefore, is *REAL* *LIFE*! A wife, a husband, kids, |
30 |
THOSE are "more important" (or /should/ be) than "improving the code". A |
31 |
job, good grades at the uni, THOSE are more important than "improving the |
32 |
code". |
33 |
|
34 |
Are you a Mark Shuttleworth? Do you have a few million dollars sitting |
35 |
around to fund your little distribution? If so, go to it, but it's not |
36 |
going to be Gentoo, because Gentoo is a community distribution. Part of |
37 |
what makes it what it is, is the volunteer efforts of all that pitch in. |
38 |
If you changed that by sponsoring it, paying for development, it would |
39 |
cease to be the Gentoo most of us know and love. (Look up the Zynot fork |
40 |
for more on that.) If you don't have that few million, then perhaps a bit |
41 |
more understanding of the nature of volunteer efforts is in order. This |
42 |
is /not/ to say there isn't room in the open source community for the |
43 |
Ubuntus of the world, because obviously there is. However, Ubuntu is |
44 |
/not/ Gentoo; Gentoo is /not/ Ubuntu. Again, that's been tried before. |
45 |
Go take a look at Zynot. |
46 |
|
47 |
So... we are left with a situation in which every contributor is a |
48 |
volunteer, taking a bit of time here, a bit of time there, to pitch in and |
49 |
make their little corner of Gentoo better. One characteristic of |
50 |
working with volunteers is that the volunteers get to decide what they |
51 |
spend time on. Most of the developers, it would seem, choose to spend |
52 |
their time directly involved with the code, developing and doing primary |
53 |
testing, sure, but the QA testing is left to the ~arch users such as |
54 |
myself, and to the bug system, depending on users to file bugs, then check |
55 |
them and reopen them if necessary. |
56 |
|
57 |
Even if we were to find a number of volunteers that wanted to spend all or |
58 |
most of their Gentoo time on QA peer reviewing the work of others, who's |
59 |
to say the ones actually doing the work would find that situation |
60 |
satisfactory? Keep in mind, once again, that it's volunteers doing the |
61 |
work. They only do it as long as it remains satisfying for them to |
62 |
continue doing it. Fortunately or unfortunately, the types of people that |
63 |
would find constantly peer reviewing the work of others satisfying enough |
64 |
to continue to do it on a volunteer basis, are not generally the types of |
65 |
people that the volunteers actually doing the coding are likely to find it |
66 |
pleasant enough working with to continue to volunteer their own time. |
67 |
Pretty quickly, it would seem too much like a job -- one they aren't |
68 |
getting paid to do -- and too little like the sort of fun that continues |
69 |
to draw them into volunteering. Very likely, it wouldn't be long until |
70 |
it'd all be peer reviewers, with nothing to do, because all the folks |
71 |
doing the work to be peer reviewed had gotten tired of it, and found other |
72 |
"more important" things to do with their time! |
73 |
|
74 |
That would appear to me to be the dynamic that's the problem with your |
75 |
solution, in addition to the fact that Gentoo is constantly |
76 |
"understaffed", that is, there is always more work to be done than there |
77 |
are folks with time to do it. That of course, pretty much by definition, |
78 |
is the nature of a volunteer project. The closer it gets to stasis, the |
79 |
closer it gets to having enough man-hours to match the work available, the |
80 |
less important what is left becomes, so the more likely it becomes for |
81 |
those that /would/ volunteer, to again, find other "more important" things |
82 |
to do with their time. For that reason alone, in addition to the one |
83 |
above, it's relatively unlikely such a QA/peer review system will ever be |
84 |
set up. Why? Because by definition, that's less important than actually |
85 |
having the code there to use or peer review in the first place, and by the |
86 |
time there are enough folks actually doing the coding to make that less |
87 |
urgent than the peer review process, we are down the relative importance |
88 |
levels far enough that other things will by definition be more important |
89 |
than that last little bit of coding OR the peer review stuff, so it'll |
90 |
never get done. |
91 |
|
92 |
My personal view, FWIW as a simple Gentoo user (NOT a dev). That said, |
93 |
it's one way to explain the resistance to your argument, in any case, |
94 |
regardless of whether it's the direct viewpoint of any dev, or not. The |
95 |
devs are volunteers, resulting in a situation rather different than if |
96 |
they were paid to do it. However, if they were paid to do it, it wouldn't |
97 |
be Gentoo as we know it, but something far different, anyway. I |
98 |
personally don't happen to believe that something different would be |
99 |
anywhere close to as good, because, by definition, it would then lose that |
100 |
unique community dynamic that is so much a part of what makes Gentoo what |
101 |
it is today. |
102 |
|
103 |
-- |
104 |
Duncan - List replies preferred. No HTML msgs. |
105 |
"Every nonfree program has a lord, a master -- |
106 |
and if you use the program, he is your master." Richard Stallman in |
107 |
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html |
108 |
|
109 |
|
110 |
-- |
111 |
gentoo-dev@g.o mailing list |