1 |
(Apologies in advance if it seems like I'm picking on you. I want to |
2 |
make a point more than criticize you in particular, because I'm sure |
3 |
your motivation is care and not impediment.) |
4 |
|
5 |
On Thu, Oct 5, 2017 at 3:54 PM, Daniel Campbell <zlg@g.o> wrote: |
6 |
> |
7 |
> I don't arch test because (a) I don't use stable and (b) I don't want to |
8 |
> mess something up. |
9 |
|
10 |
And if everybody else does this then nobody will use stable, because |
11 |
it will be messed up. :) |
12 |
|
13 |
> There are also too many ways to do arch testing and |
14 |
> it's not clear which method is most reliable. |
15 |
|
16 |
The problem is that we need people do do some testing. If nobody |
17 |
tests anything because nobody has found the "most reliable" way to |
18 |
test things, then everybody will end up using completely untested |
19 |
software by default. |
20 |
|
21 |
IMO this is one of those situations where the perfect is the enemy of the good. |
22 |
|
23 |
> chroots can be problematic |
24 |
> since they have bound mounts to the host system and the interaction |
25 |
> between systems can present false positives for issues, and are fragile |
26 |
> to maintain. A VM comes with its own set of problems (drivers, I/O, |
27 |
> admin overhead), and containers are a world all their own. |
28 |
> |
29 |
|
30 |
Obviously testing grub in a chroot isn't going to work. However, |
31 |
unless you're doing testing of drivers/kernel/bootloader/etc a chroot |
32 |
will be fine for just about anything. A container would be even |
33 |
better, and the difference between a container and a chroot is about 2 |
34 |
shell commands. |
35 |
|
36 |
> Perhaps arch testing would be more attractive if there was One True Way™ |
37 |
> to setup an arch testing system |
38 |
|
39 |
I think arch testing would be more effective if people stopped |
40 |
worrying about the One True Way and just did some arch testing, any |
41 |
way they can. Just about every stable keyword in the tree is the |
42 |
result of this, and this is how it has always been on Gentoo. |
43 |
|
44 |
The moment I post a suggested One True Way somebody is going to point |
45 |
out one bug in the last 10 years that would have escaped detection, |
46 |
and suggest that nothing should be keyworded until a rigorous 300 step |
47 |
process has been run, featuring tinderboxes that nobody is bothering |
48 |
to build, and automated tests that nobody is going to write. Then |
49 |
we'll bikeshed for six months about the pros and cons of 14 different |
50 |
methods of tinderbox log analysis, and people will start wondering why |
51 |
the subject line says something about stable keywords. :) |
52 |
|
53 |
> and higher level tools to facilitate bug |
54 |
> interaction. |
55 |
|
56 |
That would certainly be lovely, but it requires somebody to build |
57 |
those tools, and I don't see that as a short term solution. |
58 |
|
59 |
I think that encouraging devs to be a little less afraid of mistakes |
60 |
is probably the right way to go. There is a balance between paralysis |
61 |
and running around just keywording everything in sight without even |
62 |
bothering to build it. If you get feedback that something broke just |
63 |
do things a little differently next time. The fact that some dev got |
64 |
yelled at by QA because he ran some shell script that wrought havoc |
65 |
upon the tree 6 years ago isn't a reason why everybody should be |
66 |
afraid to make a good-faith effort. |
67 |
|
68 |
If the Council wants proposals for a resolution to define for once and |
69 |
all what "stable" means I can toss something out there. Short of that |
70 |
I suggest that the people actually bothering to keyword stuff are |
71 |
going to get to write the rules... |
72 |
|
73 |
-- |
74 |
Rich |