1 |
On Sat, 2004-06-19 at 22:46 +0200, Danny van Dyk wrote: |
2 |
> You always talk about the "maintainer arch". With the exception of |
3 |
> packets that _are supposed_ to work only one special archs (yaboot or |
4 |
> milo comes to my mind), speaking of a "maintainer arch" is nonsense in |
5 |
> my eyes. 99% of all packages shall work on every arch, but they |
6 |
> naturally don't work equally well on them. |
7 |
|
8 |
No they shall not and as an arch maintainer you don't know how well it |
9 |
behaves in general now don't you ? What's the logical conclusion from |
10 |
that fact... |
11 |
|
12 |
> Best example is (again; Lv |
13 |
> mentioned it already) gcc-3.4. [Please don't try to tell me there is a |
14 |
> maintainer arch for gcc !] gcc-3.4 is currently marked stable on amd64, |
15 |
> *but* is is hardmasked in case you haven't switched your profile over to |
16 |
> gcc34-amd64-2004.1, so no user who doesn't want to get's tampered with |
17 |
> gcc-3.4 problems. And you know what ? Our users rushed out to switch |
18 |
> profiles less than 5 minutes after Lv announced this solution on |
19 |
> #gentoo-amd64. Why ? They were not confident with the way gcc-3.3* works |
20 |
> on amd64. There are plenty of things gcc-3.3 lacks, starting with a |
21 |
> missing "-march=[k8, amd64]" up to severe miscompilation on all kind of |
22 |
> code when using "-Os" or "-O3". |
23 |
|
24 |
What is your point here really.. I've never said that arch maintainers |
25 |
can't do things different from package maintainers. I even agreed that |
26 |
in cases of base system stuff it's probably better, but this doesn't |
27 |
scale at all to all of Gentoo QA wise. |
28 |
|
29 |
> amd64 is a moving target, it's a relative new arch that does not run w/o |
30 |
> any problems, but it runs really good for day to day work and there are |
31 |
> even people who use Gentoo/amd64 on production systems. Lv and many |
32 |
> others (including me) try to keep this arch stable and to improve things |
33 |
> in the best way we can. Honestly, I really get annoyed when you |
34 |
> critisise our efforts w/o any insight in our daily work. Your examples |
35 |
> libtool, gcc and glibc *ALL* had severe problems that _all_ improved |
36 |
> after switching to new versions, yes even by marking prereleases/cvs |
37 |
> snapshots and development versions _stable if necessary_. |
38 |
|
39 |
If amd64 is such a moving target, then what does it do in the main |
40 |
tree ? I've asked similar questions before, because I didn't believe it |
41 |
is the best place to develop unstable new features. If all that it has |
42 |
brought to us is spreading bad developer practice all over the tree and |
43 |
even defending it, then I tend to think I was right about my initial |
44 |
objections. |
45 |
Oh, I should not have said that, now I'm going to be demonized all the |
46 |
way. This must've been my big plan behind the scenes all along... |
47 |
|
48 |
> "'the maintainer's arch' should know about _all_ problems with the |
49 |
> packages."... Well, okay... screw the whole gentoo structure, we only |
50 |
> need one releng-dev to build all arch-livecds and the package |
51 |
> maintainers... I hope you recognized my sarcasm. I really hope so. |
52 |
|
53 |
I'm not sure what the connection is to live-cd's, releng, etcetera... a |
54 |
snapshot could be taken at _any_ point from the stable tree and it |
55 |
should work. What I suggest is only a way to ensure that this will still |
56 |
be the case in the future, moving stable before the maintainer is |
57 |
confident about a package is a way to break this. |
58 |
|
59 |
> On the contraty, in my point of view your arguments seem based on the |
60 |
> fear to break/to question hierarchical structures. And please don't call |
61 |
> Lv's vision "shortshighted amd64 only". He's been working to improve the |
62 |
> toolchains not only of amd64, but also from mips (with 3 ABIs) and |
63 |
> ppc64. |
64 |
|
65 |
*64 then. My arguments are based solely on my fear that the quality of |
66 |
Gentoo is declining by certain practices that are now commonplace in the |
67 |
tree. |
68 |
|
69 |
> You are continously trying to talk bad Lv's efforts. Last month |
70 |
> he did 111 commits, you did 50, this month he did 166 commits, you did |
71 |
> 6. |
72 |
|
73 |
It makes me grin that you actually bring this up as a measuring scale. |
74 |
Someone who fixes all copyright headers makes a lot of commits, but |
75 |
doesn't add much. |
76 |
|
77 |
> It's naturally that with more commits there sneak in more errors. In |
78 |
> my eyes, you should try to take up a bit of Lv's vision. His "vision" is |
79 |
> to improve Gentoo. Please don't call that "shortsighted" ! |
80 |
|
81 |
I'm sort of sad to hear that you do not think my point is not meant to |
82 |
improve Gentoo or at least keep it at the level that it is at. |
83 |
|
84 |
> Foser: this *no battle* between amd64 and x86, and this is especially |
85 |
> *no crusade* against you (or the gnome-herd). But you try hard to let it |
86 |
> look that way. |
87 |
|
88 |
Do I ? I think you pull more into this than I do, I didn't set off this |
89 |
whole discussion, I just replied with what I think is the right thing to |
90 |
do. |
91 |
Anyway, I'm sort of getting targeted by one dev of a certain arch after |
92 |
another, who often tend to flip out at me personally, while I have been |
93 |
only on topic so far. So as far as crusading goes, you might want to ask |
94 |
someone else about the right direction to Jerusalem. |
95 |
|
96 |
> IMHO the above comments had to be said, everything's been comments on |
97 |
> foser's statement. However, I'd like state something to the |
98 |
> "stabilization" process on my own, too. We urgently need a policy on |
99 |
> this, and i want to suggest a system where the ebuild maintainers as |
100 |
> well as the arch developers both get actively get involved. |
101 |
> |
102 |
> 1. The ebuild maintainers should have the possibility to say: "ARCHs, |
103 |
> feel free to mark stable. All major bugs that i know of are closed, only |
104 |
> arch dependent bugs are left." |
105 |
|
106 |
Ehm, this is basicly what I've said so far. And this point is when the |
107 |
maintainers arch goes stable. |
108 |
|
109 |
> 2. The arch maintainers should ask the package maintainers first before |
110 |
> marking a package stable *before* the maintainer's approval. However, |
111 |
> there are package maintainers that aren't on IRC all the time/that check |
112 |
> their dev-mails regularly, but probably only once or twice a week. In |
113 |
> this case, arch maintainers should be allowed to just check bugs.g.o on |
114 |
> major bugs and mark stable if the arch decides so. |
115 |
|
116 |
I really don't see what the difference is with my suggested policy all |
117 |
along. Only some arches do not adhere to this currently, thats my point. |
118 |
|
119 |
> This is pretty much the way Donnie Berkholz proposed (Please correct me |
120 |
> if I'm wrong), and in my eyes it's the only way that makes sense. |
121 |
|
122 |
Well, i hope i corrected it enough. But I'm glad you agree with me in |
123 |
the end. Don't let the pope's propaganda fool ya. |
124 |
|
125 |
No harm to you, |
126 |
- foser |