1 |
David Shakaryan wrote: |
2 |
> Marijn Schouten (hkBst) wrote: |
3 |
>> Can't we already (automatically) remove masks for which the relevant |
4 |
>> ebuilds have been removed? Is there another situation for which you |
5 |
>> think this is useful? |
6 |
> |
7 |
> While it might be possible, we currently aren't doing that. It shouldn't |
8 |
> be too difficult to implement, but it would have to be semi-smart and |
9 |
> check if that package ever existed, as some developers add masks before |
10 |
> adding packages and we don't want those automatically removed. |
11 |
> |
12 |
I thought a package removal was a slightly more complex thing than a simple |
13 |
dir removal? Even if it isn't, that's the point to tie in (ie when it's |
14 |
removed from the tree) imo. |
15 |
|
16 |
> Without saying something on whether it would be useful or not, I think |
17 |
> that we should change the format to sth which can be processed easily by |
18 |
> programs (XML for instance). To ensure human readibility, one could keep |
19 |
> the text file and have the xml one addtionally. |
20 |
> |
21 |
> Then the whole process of adding fields is rather simple :) |
22 |
> |
23 |
XML is much harder to process in scripts than plain-text. Having a simple |
24 |
ini file format (eg [atom]), or a start and stop sequence (eg ## start |
25 |
atom) is much easier both to read and to process. As for parsing, ini file |
26 |
format parsers and the like are common libs. There are two different libs for |
27 |
it in gentoo already[1] |
28 |
|
29 |
I agree that appending would be easier; it might also mean older masks don't |
30 |
just get forgotten, since anyone looking at the file would have to scroll |
31 |
past them, decreasing the likelihood of cruft. |
32 |
|
33 |
I concur that the expiration date sounds risky (as another poster |
34 |
described.) Personally I'd prefer all masks to have open bugs for as long |
35 |
as they are in the file, assigned to the dev who made the entry. This would |
36 |
be specific to the mask, and not any issues it is supposed to address |
37 |
(which would be dependent bugs?) |
38 |
|
39 |
Good luck with automation! |
40 |
|
41 |
[1] http://article.gmane.org/gmane.linux.gentoo.devel/45762 |
42 |
-- |
43 |
gentoo-project@g.o mailing list |