| 1 |
On Sat, Sep 05, 2015 at 02:42:15PM +0200, Ulrich Mueller wrote: |
| 2 |
> >>>>> On Sat, 5 Sep 2015, Rich Freeman wrote: |
| 3 |
> |
| 4 |
> > I certainly support the principle, but for the sake of transparency |
| 5 |
> > can we try to coordinate this so that the setting name doesn't |
| 6 |
> > change when this moves into the package manager for EAPI6? |
| 7 |
> |
| 8 |
> So far, the EAPI 6 draft says [1]: |
| 9 |
> |
| 10 |
> eapply_user |
| 11 |
> Takes no arguments. Package managers supporting it apply |
| 12 |
> user-provided patches to the source tree in the current working |
| 13 |
> directory. Exact behaviour is implementation defined and beyond |
| 14 |
> the scope of this specification. Package managers not supporting |
| 15 |
> it must implement the function as a no-op. Only available in |
| 16 |
> EAPIs listed in table [...] as supporting eapply_user. |
| 17 |
|
| 18 |
Is there a reason to pick eapply_user rather than epatch_user? I think |
| 19 |
the second one would fit better with what we already have. |
| 20 |
Alternatively, if we would like to avoid adding a new function, it could |
| 21 |
be something like epatch --user or epatch -d $DIRECTORY_WITH_PATCHES, |
| 22 |
where $DIRECTORY_WITH_PATCHES either defaults to something like |
| 23 |
$EPATCH_SOURCE_USER, or can be specified via package.env. This takes the |
| 24 |
burden away from the package manager for where to put patches, and lets |
| 25 |
the user make that choice. |
| 26 |
|
| 27 |
Cheers, |
| 28 |
—Guilherme |
Archives