| 1 |
Dnia 2015-09-16, o godz. 15:49:24 |
| 2 |
Rich Freeman <rich0@g.o> napisał(a): |
| 3 |
|
| 4 |
> On Wed, Sep 16, 2015 at 3:31 PM, Michał Górny <mgorny@g.o> wrote: |
| 5 |
> > |
| 6 |
> > As for virtuals and eclasses, I don't really understand why anyone |
| 7 |
> > thinks they are special in any regard. In both cases, we're talking |
| 8 |
> > about regular dependency change in metadata, and we need to understand |
| 9 |
> > the consequences. And they're going to be the same whether we change |
| 10 |
> > dep A to B, A to virtual, and whether we change it directly or via |
| 11 |
> > eclass. |
| 12 |
> |
| 13 |
> Sure, but a developer KNOWS when their RDEPENDS change when they're |
| 14 |
> modifying it directly in an ebuild. If they inherit an eclass and it |
| 15 |
> sets an RDEPEND and the eclass changes, then it effectively changes |
| 16 |
> the dependency for every ebuild that inherits it even though there |
| 17 |
> weren't any commits to any of those ebuilds. |
| 18 |
|
| 19 |
If you modify an eclass, you're responsible for the outcome. Even if |
| 20 |
means revbumping hundreds of ebuilds for the sake of it. Note that this |
| 21 |
is the kind of revbump that wouldn't require resetting stable keywords |
| 22 |
as long as deps are satisfied. |
| 23 |
|
| 24 |
> > 2. Dependency changes that don't need to apply immediately don't need |
| 25 |
> > revbump. For example, if foo.eclass raises minimal required version of |
| 26 |
> > a dependency but all packages built so far will work with the old one. |
| 27 |
> > |
| 28 |
> |
| 29 |
> Are we talking about a build dependency or a run-time dependency? I |
| 30 |
> don't get why we'd increase the minimal required version of a run-time |
| 31 |
> dependency if everything built so far still works with the old |
| 32 |
> version. |
| 33 |
|
| 34 |
Runtime. The minver can be raised for developer's convenience -- like |
| 35 |
when a large number of packages is expected to require a new version |
| 36 |
soon, and the developers would otherwise have to override the eclass- |
| 37 |
specified versions. Instead, the eclass is updated and new requirements |
| 38 |
apply. |
| 39 |
|
| 40 |
> Also, assuming that there is a case where this actually happens, does |
| 41 |
> this have any impact on running --depclean or any other obscure |
| 42 |
> blocker issues because the version in VDB is no longer in the tree? |
| 43 |
|
| 44 |
It shouldn't have. And if it would, the issues with the whole 'do |
| 45 |
stupid stuff and not bump' policy would be much worse. |
| 46 |
|
| 47 |
-- |
| 48 |
Best regards, |
| 49 |
Michał Górny |
| 50 |
<http://dev.gentoo.org/~mgorny/> |
Archives