1 |
(apologies for the messed up time in my last message) |
2 |
|
3 |
On Friday 18 November 2005 06:53 pm, Ciaran McCreesh wrote: |
4 |
> We've seen why this won't work in the past... Too few users know how to |
5 |
> do proper testing. We've had "please keyword, works for me" bugs for |
6 |
> things that will always segfault on startup. We've had several people |
7 |
> who think it'd be clever to automate testing reports. We've got enough |
8 |
> ricers out there that clearly broken things would end up getting "works |
9 |
> for me" spammed even more than they are already... |
10 |
|
11 |
Yeah, it's not a perfect solution, but nothing is. |
12 |
|
13 |
I think having users systems would be profiled may help ease the ricer issue. |
14 |
fex, user A has 3 systems, and marks package B as "!WFM" on one. devs can |
15 |
cross link that negative mark to the system profile and note that it's "-O12 |
16 |
--omg-itsofast", and disregard the negative mark. You could even take it a |
17 |
step further and setup ratings for the registered users, and those who end up |
18 |
with a set negativity don't count or something (for the ricers).. |
19 |
|
20 |
Not saying this is something that stability or instability should be |
21 |
automatically assumed from, but that it be used as another tool. Something |
22 |
to bridge that "poweruser" - "dev" gap. |
23 |
|
24 |
Just openly brainstorming here.. |
25 |
|
26 |
-- |
27 |
Corey Shields |
28 |
Gentoo Linux Infrastructure Team |
29 |
Gentoo Foundation Board of Trustees |
30 |
http://www.gentoo.org/~cshields |