1 |
On nie, 2017-07-09 at 10:29 +0200, Ulrich Mueller wrote: |
2 |
> > > > > > On Sun, 09 Jul 2017, Michał Górny wrote: |
3 |
> > On nie, 2017-07-09 at 09:22 +0200, Ulrich Mueller wrote: |
4 |
> > > Second, and more important, introduction of an automatic solver |
5 |
> > > would inevitably lead to proliferation of REQUIRED_USE in the tree. |
6 |
> > > However, nothing would guarantee that the package manager on the |
7 |
> > > user's side is capable of solving the constraints automatically. |
8 |
> > > The result would be more emerge failures and asking for more |
9 |
> > > micro-management of flags by users. |
10 |
> > Think of dynamic deps. We were able to eventually contain them, and |
11 |
> > teach developers not to account for them even though they are still |
12 |
> > enabled by default, I think. |
13 |
> > I don't see why optional autosolving of REQUIRED_USE could not be |
14 |
> > contained by a policy as well. |
15 |
> |
16 |
> Then can you please confirm that the policy outlined in |
17 |
> https://devmanual.gentoo.org/general-concepts/use-flags/index.html#conflicting-use-flags |
18 |
> can stay in place indefinitely, and that your GLEP doesn't intend to |
19 |
> change it? |
20 |
|
21 |
The GLEP does not mention that policy at all, so it's not affected. |
22 |
If we decide to change it, it will be done independently of the GLEP. |
23 |
|
24 |
> > Of course, there will be some people who will violate it but it's |
25 |
> > not something that doesn't happen anyway right now. |
26 |
> |
27 |
> A policy that isn't enforced is useless. |
28 |
|
29 |
Say that to the people who invented USE=gtk2,gtk3 instead of USE=gtk. Oh |
30 |
wait... |
31 |
|
32 |
> > Are you suggesting that we introduce half-tested feature in EAPI 7, |
33 |
> > then spend a few years figuring out that it doesn't work as |
34 |
> > expected? |
35 |
> |
36 |
> No, I am suggesting that we introduce a new package manager feature in |
37 |
> a well defined way, so that ebuild authors can rely on it. We have a |
38 |
> mechanism for that, and I don't see a good reason not to follow it. |
39 |
> |
40 |
> > Because I don't see how we get it tested properly without having |
41 |
> > users actually test it and report the results. |
42 |
> |
43 |
> It shouldn't be necessary for the spec to specify all details of the |
44 |
> algorithm. It should catch the basics though in the next EAPI, like |
45 |
> leftmost preferred, banning empty groups, and banning USE conditionals |
46 |
> inside groups. |
47 |
> |
48 |
|
49 |
Works for me. However, the problem is whether i'll live to see the day |
50 |
when I can realistically use it. |
51 |
|
52 |
python-single-r1 needs this urgently. Today. Not 15 years from now when |
53 |
we can drop support for EAPI <=6. Presuming Gentoo will be anything but |
54 |
a huge pile of defunct policies and bureaucracy 15 years from now. |
55 |
|
56 |
-- |
57 |
Best regards, |
58 |
Michał Górny |