1 |
"Alec Warner" <antarus@g.o> posted |
2 |
b41005390809211835h39f5a359k571e9693f7db1630@××××××××××.com, excerpted |
3 |
below, on Sun, 21 Sep 2008 18:35:04 -0700: |
4 |
|
5 |
> gentoo-x86 uses bash; the ebuilds, the eclasses, they all rely on it. |
6 |
> I'm pretty sure most package managers rely on bash as well, but I have |
7 |
> not looked at the code outside of portage to verify. |
8 |
> |
9 |
> I really dislike ideas where the compelling argument is 'in the future |
10 |
> we may make a specific decision and that makes that one choice easier.' |
11 |
|
12 |
> If you have a compelling argument for switching the entire tree to POSIX |
13 |
> then give it; however I'm pretty sure it is a difficult argument to make |
14 |
> (Uberlord tried to make it in the past and did not succeed). Otherwise |
15 |
> lets just roll with the bash implementation. |
16 |
|
17 |
++ |
18 |
|
19 |
This seems to be what it comes down to. Based on past discussion, while |
20 |
individual devs can go POSIX and nobody's going to complain, indeed, |
21 |
they're likely to be respected for taking that position in regard to |
22 |
/their/ /own/ /code/, there's sufficient resistance to making /all/ devs |
23 |
favor POSIX over BASH that it's effectively not even rational to |
24 |
contemplate it at this time, nor is it likely to be for a dev generation |
25 |
or more. Trying to do otherwise is /perceived/ as (note I didn't say it |
26 |
was the /intent/ of, only /perceived/ as) trying to force POSIX down the |
27 |
throat of others, and given the current state requiring bash, turns that |
28 |
respect for devs doing it for their own work on its head -- they're then |
29 |
seen as being an active danger to the ability of devs who don't |
30 |
differentiate between POSIX sh and BASH to continue as devs in good |
31 |
Gentoo standing, with the entirely predictable reaction being to oppose |
32 |
them at nearly any cost. |
33 |
|
34 |
Which pretty much leaves Gentoo depending on BASH now and for the |
35 |
foreseeable future, and there's little point in debating it further or |
36 |
indeed, in "artificially" trying to reduce that dependence, except in |
37 |
one's own work if desired. It's just not worth the fractious debates it |
38 |
causes, particularly when the conclusion is predetermined based on past |
39 |
iterations. We've lost very good developers on this issue in the past. |
40 |
Let's not make it any more, OK? |
41 |
|
42 |
-- |
43 |
Duncan - List replies preferred. No HTML msgs. |
44 |
"Every nonfree program has a lord, a master -- |
45 |
and if you use the program, he is your master." Richard Stallman |