| 1 |
>>>>> On Thu, 1 Oct 2009, Ciaran McCreesh wrote: |
| 2 |
|
| 3 |
>> EAPI 3 is reopened for the purpose of including one sole feature, |
| 4 |
>> namely preservation of file modification times, as outlined in |
| 5 |
>> bug 264130, option B of comment 26 [1]. Both Portage and Pkgcore |
| 6 |
>> already comply with this, so it would be zero implementation cost. |
| 7 |
|
| 8 |
> Why go with an inferior solution? Why not go with a solution that |
| 9 |
> requires the package manager to fix broken mtimes? |
| 10 |
|
| 11 |
Because it would be non-zero implementation cost for Portage, so |
| 12 |
probably out of question for EAPI 3. And it's not at all clear if the |
| 13 |
solution is inferior. Since half a year, nobody cared to answer the |
| 14 |
question of comment 25 of mentioned bug. |
| 15 |
|
| 16 |
But if you want, the council can also vote if it should be option |
| 17 |
A (current Portage and Pkgcore behaviour, all mtimes are preserved), |
| 18 |
B (optional update of "old" mtimes), or C (mandatory update). |
| 19 |
|
| 20 |
@betelgeuse: Could you please add this to the agenda, too? |
| 21 |
|
| 22 |
> Also, what are the rules regarding this and things like stripping |
| 23 |
> and other fixes and changes that the package manager performs upon |
| 24 |
> files before merging them? |
| 25 |
|
| 26 |
This is outside the scope of this proposal, and (at least for now) I'm |
| 27 |
not going to work anything out. |
| 28 |
|
| 29 |
Ulrich |
Archives