1 |
On Thu, 26 Mar 2009 20:59:02 -0300 |
2 |
Jorge Morais <please.no.spam.here@×××××.com> wrote: |
3 |
|
4 |
> > It's still a way to check out these new features hands-on :) |
5 |
> |
6 |
> It's a kernel. |
7 |
|
8 |
Exactly. |
9 |
That's definitely one of the thing you have to know about, especially |
10 |
if this one's got some unpredictable stuff in it. |
11 |
|
12 |
> Even if it already released, it has a higher chance of bugs than a |
13 |
> more established kernel. |
14 |
|
15 |
Agreed, but from my experience, it's more like 'features' with kernel - |
16 |
some stuff just change, and most of this isn't quite obvious from |
17 |
release info, even with such commit-by-commit teardown like |
18 |
kernelnewbies. |
19 |
http://kernelnewbies.org/Linux_2_6_29 |
20 |
|
21 |
> Waiting for the package to become stable on Gentoo is not just about |
22 |
> ebuild bugs; it is also about waiting for enough users in general to |
23 |
> test the upstream package, and Gentoo users in particular to test the |
24 |
> package within Gentoo. |
25 |
|
26 |
There's a race condition and quite unpredictable dependency :) |
27 |
|
28 |
> Of course, if you want to actively help, that is a reason to test |
29 |
> the latest software... but I am afraid that a kernel bug could lead |
30 |
> to unpredictable behavior, data loss and other problems I can't |
31 |
> tolerate. Also, I think a kernel has a higher chance of hidden |
32 |
> bugs (bugs that don't stand in your face). |
33 |
|
34 |
But that's fine with desktop systems, especially if you have full |
35 |
net backup as a daily cronjob (which is a great idea, btw). |
36 |
I tend to use each new kernel for at least a few weeks, before |
37 |
deploying it anywhere, and no bugreports or security advisory papers |
38 |
are substitute for that. |
39 |
|
40 |
But this isn't really a discussion, since I certainly don't speak for |
41 |
production systems and you probably mean just a general deployment |
42 |
everywhere, so everyone's right in their own way. |
43 |
Besides, stability vs innovation is too much a matter of personal |
44 |
preference to discuss it w/o starting an endless holywar. |
45 |
|
46 |
-- |
47 |
Mike Kazantsev // fraggod.net |