1 |
On Tue, Jun 11, 2019 at 8:38 PM Michał Górny <mgorny@g.o> wrote: |
2 |
|
3 |
> On Tue, 2019-06-04 at 15:05 +0200, Kristian Fiskerstrand wrote: |
4 |
> > The agenda item "Forums (specifically OTW)" was deferred to further |
5 |
> > discussion in the mailing lists during the 2019-02-10 council meeting. |
6 |
> > The agenda for that meeting can be found at [Agenda] and the tracking |
7 |
> > bug is [Bug 677824]. |
8 |
> > |
9 |
> > This email aims to re-opens the discussion [which was started in a |
10 |
> > previous thread] as per the council decision. |
11 |
> > |
12 |
> > I ask that the discussion remains civil and respectful, while also |
13 |
> > allowing for a high bar for the actual discussion. |
14 |
> > |
15 |
> |
16 |
> I don't think OTW is a major problem. The real problem are *technical* |
17 |
> forums, and those cause two problems. The first of them is that |
18 |
> developers rarely hear of the problems with their packages. The second |
19 |
> of them is that Forums tend to breed very bad 'solutions'. |
20 |
> |
21 |
|
22 |
I'm curious how this is different from other support forums. Are there no |
23 |
bad solutions proposed on the wiki? in #gentoo? In other channels? |
24 |
|
25 |
-A |
26 |
|
27 |
|
28 |
> |
29 |
> I don't mind providing multiple support channels as long as problems |
30 |
> actually reach developers. However, I don't think it's news to most of |
31 |
> the developers who don't actively participate in Forums (read: look |
32 |
> for new threads everywhere) that some problems never leave them. |
33 |
> |
34 |
> During my years in Gentoo I've been pinged a few times over expansive |
35 |
> Forum threads on problems with my packages which never made it to |
36 |
> Bugzilla or anything else that I actually could've noticed. It all |
37 |
> relies on courtesy of accidental developers (who are not Forum |
38 |
> moderators, I should add). I can only imagine how many problems were |
39 |
> never addressed properly because the maintainer never learned of them, |
40 |
> and cheap hacks proposed on Forums were sufficient for the users. |
41 |
> |
42 |
> And no, I don't think that requiring every developer to directly follow |
43 |
> all Forum feeds is a solution. |
44 |
> |
45 |
> A side effect of the former problem is that Forums are home to many |
46 |
> horrible 'solutions'. Sadly, those solutions sometimes involve making |
47 |
> things *much worse* than they were before. This is problem both for |
48 |
> users who end up victims of having their systems broken, and developers |
49 |
> who end up having to help fix the resulting breakage. |
50 |
> |
51 |
> Breakage resulting from use of dev-python/pip is the most prominent |
52 |
> example I know of. Multiple Forum victims ended up using it to 'fix' |
53 |
> problems. As a result, they ended up with obsolete directly installed |
54 |
> packages overriding Gentoo packages and breaking stuff. The scale of |
55 |
> this was so great that I had to actually patch dev-python/pip to block |
56 |
> installing packages system-wide. Which is a technical hack to a social |
57 |
> problem. |
58 |
> |
59 |
> I'm not saying Forums is the only source of the problem, people can |
60 |
> figure out how to break systems themselves. However, Forums is |
61 |
> frequently a source of bad information that is mistakenly trusted |
62 |
> and is not properly verified and rejected. |
63 |
> |
64 |
> To summarize, I think the two major problems with the Forums are: |
65 |
> 1) not passing information properly to package maintainers, |
66 |
> and 2) lack of proper Q/A. If you can solve them, I don't have any |
67 |
> problem with the Forums. |
68 |
|
69 |
|
70 |
> -- |
71 |
> Best regards, |
72 |
> Michał Górny |
73 |
> |
74 |
> |