1 |
On Wed, Sep 16, 2015 at 4:44 PM, Ian Stakenvicius <axs@g.o> wrote: |
2 |
> |
3 |
> - - emerge -uD @world would update the dep anyhow |
4 |
> |
5 |
> - - emerge -u @world wouldn't rebuild the package if that package |
6 |
> didn't change, and if the package did change then the new dep would |
7 |
> get built. |
8 |
|
9 |
Just to be clear, my point was that if the reason the eclass was |
10 |
updated was because some new package version in the tree needed it, |
11 |
and you update that other package in the tree, then that will update |
12 |
the dependency, and that would in turn rebuild anything with a slot |
13 |
operator dep on it. |
14 |
|
15 |
Which is of course exactly what should happen if the soname/ABI needs to change. |
16 |
|
17 |
> |
18 |
> So I think I can safely drop my concern. Care still needs to be |
19 |
> given, certainly, as if the in-tree package isn't revbumped and gets |
20 |
> a patch that only supports the new dep, then suddenly systems will |
21 |
> fail when said package re-emerges. But that seems bad practice anyhow |
22 |
> . |
23 |
|
24 |
Right, a patch change would always require a revbump and that is |
25 |
nothing new. The only case that doesn't require a revbump is a |
26 |
build-time change. And if somebody makes a build-time change with the |
27 |
assumption that the new minimum depencency version is in effect it is |
28 |
fine, since anybody with it already installed won't be rebuilding it, |
29 |
and if anybody does rebuild it then portage will check and force the |
30 |
dependency update so everything is fine at build time. |
31 |
|
32 |
So, assuming we want to entertain the "only revbump if necessary" |
33 |
eclass wording, do we think we can actually come up with something |
34 |
that not only makes technical sense, but where we can actually expect |
35 |
developers to get it right 99% of the time? Are we encouraging |
36 |
dangerous behavior just because a few of us might or might not be able |
37 |
to get it right? |
38 |
|
39 |
I suppose if somebody does mess up we can go in and revbump a bunch of |
40 |
ebuilds. The thing is that such problems are very hard to detect - |
41 |
they usually manifest as users posting bizarre portage output on lists |
42 |
and forums and being showered with a lot of bad advice before somebody |
43 |
who knows what they're talking about notices the thread, and it can |
44 |
happen a while after the commit that introduced the problem. |
45 |
|
46 |
So, part of me really wonders if it is worth it just to save a bunch |
47 |
of revbumps that probably could be done by a script and with git it |
48 |
can even happen atomically and with the possibility of review or |
49 |
testing. |
50 |
|
51 |
-- |
52 |
Rich |