1 |
On Mon, 7 May 2012 17:17:10 +0200 |
2 |
Ulrich Mueller <ulm@g.o> wrote: |
3 |
> >>>>> On Mon, 7 May 2012, Ciaran McCreesh wrote: |
4 |
> > On Mon, 7 May 2012 02:12:14 +0200 |
5 |
> > Ulrich Mueller <ulm@g.o> wrote: |
6 |
> >> | Furthermore, for these EAPIs, if the function is overridden, it |
7 |
> >> | shall be a fatal error if the apply_user_patches command has not |
8 |
> >> | been called at least once by the end of the phase. |
9 |
> >> |
10 |
> >> Wouldn't it make more sense to call apply_user_patches implicitly |
11 |
> >> at the end of the phase, if it hasn't been called before? |
12 |
> >> |
13 |
> >> Otherwise, a call to that function would have to be added to every |
14 |
> >> ebuild that defines src_prepare. |
15 |
> |
16 |
> > That was the point. The discussion on gentoo-dev suggested that "at |
17 |
> > the end" is often the wrong place to put it, due to eautoreconf etc. |
18 |
> > We need people to be explicit about where it goes. |
19 |
> |
20 |
> Yes, so apply_user_patches gives ebuilds the possibility to specify |
21 |
> the exact place. I still think that a fallback to calling it at the |
22 |
> end of the phase would be better than aborting with a fatal error. |
23 |
|
24 |
Why? That error will only happen at most once, and users will never |
25 |
see it. |
26 |
|
27 |
> After all, this functionality is just a stop-gap measure for users to |
28 |
> apply quick bug fixes, so I don't expect that it will be used very |
29 |
> often. Even fewer cases will require that eautoreconf is called. Do we |
30 |
> really want to force developers to put this function call into every |
31 |
> ebuild? That would be out of proportion, IMHO. |
32 |
|
33 |
"It not being used very often" is the key issue: developers are likely |
34 |
to forget about it if they're not forced to remember. It's especially |
35 |
complicated when eclasses are involved, so even if developers do |
36 |
remember, they may not be sure whether they have to specify it in a |
37 |
particular case. |
38 |
|
39 |
We can avoid a huge pile of "blah doesn't work properly with user |
40 |
patches!" bugs here, essentially for free. |
41 |
|
42 |
-- |
43 |
Ciaran McCreesh |