1 |
On Thu, Oct 05, 2017 at 02:32:59PM -0700, Rich Freeman wrote: |
2 |
> On Thu, Oct 5, 2017 at 2:17 PM, Daniel Campbell <zlg@g.o> wrote: |
3 |
> > |
4 |
> > I think your idea is noble, but it comes down to "do QA and Arch testers |
5 |
> > care enough to make their job easier and invite others?" I suspect there |
6 |
> > are territorial behaviors at work that prevent people from learning and |
7 |
> > improving their packages through testing. As long as our culture is |
8 |
> > insular, this problem will remain. |
9 |
> > |
10 |
> |
11 |
> I don't really think this is the problem. If somebody wants to chime |
12 |
> in and say that they have been dying to arch test but some arch team |
13 |
> is keeping them away please do so, but my sense is that there just |
14 |
> isn't a lot of interest in it. |
15 |
> |
16 |
> Face it, arch testing is tedious. There isn't any mystery as to how |
17 |
> it works. You install a package, use it a bit, and generally make |
18 |
> sure it works. Then you commit a keyword change. |
19 |
> |
20 |
> Now, perhaps the whole process could be radically transformed, such as |
21 |
> using automated testing, just compile testing, and improved reporting |
22 |
> tools/etc. The issue there is that this is a completely different set |
23 |
> of skills than what current arch testers are actually doing. It is |
24 |
> also one of those areas that a few people talk about but few put much |
25 |
> time into. Most of the people who do talk about it aren't arch |
26 |
> testers (nor do I imply that they should be, or that arch testers |
27 |
> should be doing this work - I'm just pointing out that blaming arch |
28 |
> testers misses the mark). |
29 |
|
30 |
To clarify, I'm not placing "blame" per se. It's not productive and |
31 |
reduces morale. We all seem in agreement that it's a problem in need of |
32 |
solving, though, so my first thought was "what are arch testers doing to |
33 |
solve their problem?" If anyone took that as blame, it wasn't meant to |
34 |
be. |
35 |
|
36 |
> |
37 |
> The QA team really has nothing to do with the arch testing process. |
38 |
|
39 |
Why not? If someone stabilizes a package and it breaks stable somewhere, |
40 |
you can be sure QA will jump on the developer who caused the problem. |
41 |
They may not be *directly* involved, but QA is necessarily part of just |
42 |
about all development in the main tree. No value judgment on whether |
43 |
that's good or bad, just a simple fact. |
44 |
|
45 |
> |
46 |
> Now, maybe there are some barriers, such as arch teams that don't |
47 |
> really allow non-dev arch testing. Back in the original amd64 arch |
48 |
> test days where the concept was pioneered the typical process was: |
49 |
> |
50 |
> 1. Non-dev trains as arch tester. |
51 |
> 2. Non-dev takes a test to be certified as an arch tester. This was |
52 |
> administered by the arch team - no recruiters/etc involved. The arch |
53 |
> tester status was just an arch team thing - it didn't have any formal |
54 |
> Gentoo-level recognition really. |
55 |
> 3. Arch tester reviews stabilization/testing bugs and marks those |
56 |
> they have tested. |
57 |
> 4. Arch team members monitor bugs tagged as having been tested. They |
58 |
> immediately commit changes without further testing. |
59 |
> |
60 |
> For this to work you really need #4 - if arch testers just submit bugs |
61 |
> for them to end up in a backlog while a dev re-tests them then there |
62 |
> is little benefit to the process. Back in the original amd64 AT days |
63 |
> it was rare for more than a few hours to pass between an AT marking a |
64 |
> bug as tested, and the commit being made. I'm not sure why it wasn't |
65 |
> automated - probably just laziness. |
66 |
> |
67 |
> In any case, for any of this to work people need to WANT to arch test. |
68 |
> I doubt that trying to force people to arch test will help. |
69 |
|
70 |
Agreed. I just took a look at the wiki page for amd64 testing [1] and |
71 |
there are a few pieces of documentation already. I recall mgorny doing |
72 |
an audit of projects to see what they needed; were arch projects part of |
73 |
that? Looks like we have a starting point already. Maybe we should be |
74 |
asking *why* people don't want to arch test. |
75 |
|
76 |
I don't arch test because (a) I don't use stable and (b) I don't want to |
77 |
mess something up. There are also too many ways to do arch testing and |
78 |
it's not clear which method is most reliable. chroots can be problematic |
79 |
since they have bound mounts to the host system and the interaction |
80 |
between systems can present false positives for issues, and are fragile |
81 |
to maintain. A VM comes with its own set of problems (drivers, I/O, |
82 |
admin overhead), and containers are a world all their own. |
83 |
|
84 |
Perhaps arch testing would be more attractive if there was One True Way™ |
85 |
to setup an arch testing system and higher level tools to facilitate bug |
86 |
interaction. |
87 |
|
88 |
How do our downstream forks handle arch testing? Do they have the same |
89 |
problems we do? |
90 |
> |
91 |
> -- Rich |
92 |
> |
93 |
|
94 |
[1]: https://wiki.gentoo.org/wiki/Project:AMD64_Arch_Testers |