1 |
On 5/24/09, Alan McKinnon <alan.mckinnon@×××××.com> wrote: |
2 |
> Dale asked *where* system is defined, not what it consists of. |
3 |
|
4 |
Well, he was asking many things. One can run into several directions |
5 |
with all the questions he asked, depending on the assumptions one |
6 |
makes. ;) |
7 |
|
8 |
> It appears you are completely missing the point. It is indeed very easy to |
9 |
> add things to the @system set, but we are talking about the system set, |
10 |
> and it is broken out of the box as shipped. |
11 |
|
12 |
I'm just focusing on it on a level where we (the users, me, you, Mr. |
13 |
Dalek, or the OP) can do things about it, where we have immediate |
14 |
write access. That rules out the Gentoo repos and leaves the local |
15 |
boxes. |
16 |
|
17 |
We also can and should escalate it to those who can do more, and IIRC |
18 |
some kind person (you? missed it and forgot who it was) already |
19 |
offered to file the bug. Short of becoming a dev, what more can be |
20 |
done in a volunteer-based distro? |
21 |
|
22 |
> Portage will not let you unmerge portage or gcc without a fight. |
23 |
|
24 |
Been there, done exactly that (with gcc), though no t-shirts were |
25 |
awarded. A 10-second countdown (IIRC) is hardly a "fight". (But the |
26 |
reinstall sure was, won't do it ever again if I simply can avoid it!) |
27 |
:) |
28 |
|
29 |
> But it's not working that way today. Ergo, it is broken. |
30 |
|
31 |
I think that is what we agree on, it warrants an escalation (a bug or |
32 |
a feature request). It is the immediate things vs. higher level |
33 |
reporting that we seem to be slightly mixed up about. |
34 |
|
35 |
I already asked quickly about this on #gentoo on Saturday and the |
36 |
resident devs there hadn't heard about this, but didn't really appear |
37 |
that worried either. I won't blame them. This won't destroy your data, |
38 |
this will only make certain breakage cases a bit more complicated, and |
39 |
if you really need the four or five nines uptimes, you can try |
40 |
mitigating the problem by, e.g., going full scale buildpkg for |
41 |
everything. |
42 |
|
43 |
Furthermore, the glep-55 discussion seems to be heating up again on |
44 |
the -dev-list, so I wouldn't expect the devs to be quick to fix |
45 |
anything related to the PMS/multiple package manager issues right now. |
46 |
They're too busy flinging all shades of poo. Summer, they should have |
47 |
lectures and exams during summer at Universities, too. :/ |
48 |
|
49 |
-- |
50 |
Arttu V. |