1 |
On Fri, 2004-07-09 at 13:56, Alexander Gretencord wrote: |
2 |
> Well someone on the kde team should, right? Or if you rely on user input, |
3 |
> automatically push new versions from ~x86 to x86 when no user has complained |
4 |
> (it's easier to complain than to state "everything works fine", especially |
5 |
> when you don't use all of kde, I personally would never find a problem with |
6 |
> kdegames as I don't install it). But I do use kdepim and if I find a problem |
7 |
> with it I report it to the KDE bug system. However, I have not found any |
8 |
> error that was the ebuilds fault. Should I file a bug stating "kdepim merged |
9 |
> fine and seems to work ok a week after merging"? |
10 |
|
11 |
A week is fine... a month is better. |
12 |
|
13 |
> > The fact is that, with the way portage is structured and the commands you |
14 |
> > have to use to properly make the submittals, it takes me well over an hour |
15 |
> > to bump a version for KDE. |
16 |
> |
17 |
> For getting it from ~x86 to x86? If that's really the case then perhaps this |
18 |
> should open a wider discussion: a) Need more developers as most do have a |
19 |
> real life besides gentoo or b) make it simpler! |
20 |
|
21 |
Considering KDE has many dependencies, *and* many things that depend on |
22 |
it, I don't see how it could be made any simpler. More developers are |
23 |
definitely needed, not only on the KDE team, but on most of them. |
24 |
|
25 |
> > Considering the fact that it's volunteer time on my part, it gets put down |
26 |
> > on my priority list. |
27 |
> |
28 |
> Then its the job of someone who does get paid to make it easier for the |
29 |
> volunteers to get the job done, right? |
30 |
|
31 |
Nobody at Gentoo gets paid to work for Gentoo. |
32 |
|
33 |
> > So, if you want to be part of the solution, start being proactive in |
34 |
> > helping to fix bugs and get ebuilds submitted. |
35 |
> |
36 |
> The ebuild is almost always there when I want to upgrade a package it just |
37 |
> sits in ~x86 ... |
38 |
|
39 |
package.unmask and package.keywords are your friends. Learn them! Use |
40 |
them! Love them! |
41 |
|
42 |
> > and it gets to be a bit frustrating to have to try and explain over and over |
43 |
> > again. |
44 |
> |
45 |
> Have a standard reply to paste, also stating you just don't have the time to |
46 |
> answer everyone personally. Its still better than saying nothing because |
47 |
> you're overworked. In the time you explain this to all of those people you |
48 |
> could move lots of packages to x86 :) |
49 |
|
50 |
I'm sure he isn't wasting that much time replying... ;] |
51 |
|
52 |
This tends to come up about every 3 months or so. Perhaps we should |
53 |
just all bookmark this thread on gmane and next time it comes up, simply |
54 |
paste them the URL and end the discussion right there. |
55 |
|
56 |
> > That said, there's absolutely no harm in kde sitting as ~x86 for 3.2.3. |
57 |
> > It's still completely installable, and portage is configurable enough so as |
58 |
> > to not complain to you to downgrade. |
59 |
> |
60 |
> Yes there is harm in it. When I first see a package in ~x86 after the official |
61 |
> release I think there might be a problem with it and don't merge it until I |
62 |
> can't stand it any more and just merge it as I want to have a fix for all the |
63 |
> bugs I reported for the last version of kde :) |
64 |
> |
65 |
> Still, if I just edit the various files in /etc/portage/ I might at some time |
66 |
> get a really unusable version of an ebuild which would not be the case if the |
67 |
> build sat in ~x86 for a week or so before going to x86. |
68 |
|
69 |
No ebuild should ever move to stable after a week. The "Gentoo |
70 |
standard" is 30 days. |
71 |
|
72 |
> I really thought that an ebuild automatically went to x86 after some time of |
73 |
> no complaints, until now... |
74 |
|
75 |
Definitely not. What system would possibly be able to know if there |
76 |
were complaints? Would we have it parse bugzilla? #gentoo? gentoo-user? |
77 |
|
78 |
How about the forums? |
79 |
|
80 |
> All that being said, I do understand that you all have a real life, work and |
81 |
> family to attend to. Maybe gentoo should get more developers and skip ~x86 |
82 |
> until they have :) I would really like to help besides reporting a bug or two |
83 |
> once in a while, if it wasn't so much work. What good would I do when all I |
84 |
> could spend was and hour or two every few days. Or a whole day once in a |
85 |
> while and then nothing for a week? |
86 |
|
87 |
Honestly, if you think ~x86 should be skipped, then perhaps you should |
88 |
be running it, instead. After all, you seem more than willing to help |
89 |
with "testing" which is exactly what the ~arch KEYWORDS are all about. |
90 |
We don't put things that we know are broken into the tree. Not as |
91 |
~arch, and not as stable. |
92 |
|
93 |
Funny enough, your "free time" to help out is about the same as most |
94 |
Gentoo developers. Remember that absolutely *none* of us is paid and we |
95 |
all have jobs/school/families to attend to other than Gentoo. |
96 |
|
97 |
-- |
98 |
Chris Gianelloni |
99 |
Release Engineering QA Manager/Games Developer |
100 |
Gentoo Linux |
101 |
|
102 |
Is your power animal a penguin? |