1 |
On Wednesday 18 December 2013 17:20:32 Greg Turner wrote: |
2 |
> My hunch is that the decision to put the config.{sub,guess} |
3 |
> replacement code in econf was intended as a quick-and-dirty way to |
4 |
> avoid doing the replacements, in cases where no configure script runs |
5 |
> in an ebuild. |
6 |
|
7 |
it was intended as the easy answer so that all packages would get it "for |
8 |
free". previously, we had an eclass, and had to update every single package |
9 |
to call a function in that eclass. it sucked hard. it makes sense to have it |
10 |
in `econf` because that func is only used in conjunction with autotool based |
11 |
packages. |
12 |
|
13 |
> Post EAPI-2, the convention that hacking on the sources in "${S}" is |
14 |
> a "no-no" after src_prepare has clearly crystallized considerably (I'm |
15 |
> guessing the code has EAPI-[01] origins); violating that convention in |
16 |
> econf seems awkward. |
17 |
|
18 |
this code existed long before EAPI was ever a thing |
19 |
|
20 |
> o It doesn't run, if, for some reason, the ebuild must invoke |
21 |
> configure directly rather than use econf |
22 |
|
23 |
this happens in like 3 packages. we can suck it up. |
24 |
|
25 |
> o when econf is invoked repeatedly, it does the same |
26 |
> O(# of dirs in ${S}) noop over and over |
27 |
|
28 |
true, but it hasn't been a big enough deal for us to care |
29 |
|
30 |
> In short... moving the config.{sub,guess} replacement code (but |
31 |
> probably not the shebang patching for reasons of expedience) to some |
32 |
> post-src_prepare place would probably be more elegant and pretty easy |
33 |
> to do. |
34 |
|
35 |
that discussion should happen on gentoo-dev in conjunction with PMS. or file a |
36 |
bug if there isn't one already. |
37 |
-mike |