| 1 |
>>>>> On Mon, 7 May 2012, Ciaran McCreesh wrote: |
| 2 |
|
| 3 |
> On Mon, 7 May 2012 02:12:14 +0200 |
| 4 |
> Ulrich Mueller <ulm@g.o> wrote: |
| 5 |
>> | Furthermore, for these EAPIs, if the function is overridden, it |
| 6 |
>> | shall be a fatal error if the apply_user_patches command has not |
| 7 |
>> | been called at least once by the end of the phase. |
| 8 |
>> |
| 9 |
>> Wouldn't it make more sense to call apply_user_patches implicitly at |
| 10 |
>> the end of the phase, if it hasn't been called before? |
| 11 |
>> |
| 12 |
>> Otherwise, a call to that function would have to be added to every |
| 13 |
>> ebuild that defines src_prepare. |
| 14 |
|
| 15 |
> That was the point. The discussion on gentoo-dev suggested that "at |
| 16 |
> the end" is often the wrong place to put it, due to eautoreconf etc. |
| 17 |
> We need people to be explicit about where it goes. |
| 18 |
|
| 19 |
Yes, so apply_user_patches gives ebuilds the possibility to specify |
| 20 |
the exact place. I still think that a fallback to calling it at the |
| 21 |
end of the phase would be better than aborting with a fatal error. |
| 22 |
|
| 23 |
After all, this functionality is just a stop-gap measure for users to |
| 24 |
apply quick bug fixes, so I don't expect that it will be used very |
| 25 |
often. Even fewer cases will require that eautoreconf is called. Do we |
| 26 |
really want to force developers to put this function call into every |
| 27 |
ebuild? That would be out of proportion, IMHO. |
| 28 |
|
| 29 |
Ulrich |
Archives