| 1 |
Thilo Bangert wrote: |
| 2 |
>> So I'd be in favour of getting rid of them, if we make sure that |
| 3 |
>> everybody always commits to the ChangeLog (Make it a repoman failure). |
| 4 |
>> Side benefit of removing the need to double-commit from the hashes |
| 5 |
>> changing. |
| 6 |
>> |
| 7 |
> |
| 8 |
> i have never understood why repoman doesn't automatically put the commit |
| 9 |
> message into the ChangeLog.... (share your use case!) |
| 10 |
|
| 11 |
Yeah I would like at least a switch that would call echangelog first and |
| 12 |
then do its stuff, sunrise-commit which I use for overlays has -c for |
| 13 |
this. Hm well I can make myself a wrapper but if it was already there, |
| 14 |
it would be better :) |
| 15 |
|
| 16 |
> taking this one more step ahead, the ChangeLog could perhaps be made a |
| 17 |
> virtual file, which on demand is extracted from VCS metadata... now |
| 18 |
> _that_ would save some bandwidth and space (no numbers, sorry). |
| 19 |
|
| 20 |
Interesting idea, if that's possible with CVS... but I don't see how it |
| 21 |
saves space and bandwith for rsync users. |
| 22 |
|
| 23 |
> i am all for the removal of $Header:$, btw. the current double commits |
| 24 |
> simply suck! |
| 25 |
|
| 26 |
I would leave it as long as we use CVS, for the reasons others already |
| 27 |
said (syncing changes to overlays which I myself used to do). But if we |
| 28 |
move to some other VCS, it would destroy the beauty of atomic commits... |
| 29 |
-- |
| 30 |
Vlastimil Babka (Caster) |
| 31 |
Gentoo/Java |
| 32 |
-- |
| 33 |
gentoo-dev@g.o mailing list |
Archives