1 |
On Wed, 2019-05-08 at 11:29 +0200, Alexis Ballier wrote: |
2 |
> On Tue, 07 May 2019 23:47:30 +0200 |
3 |
> Michał Górny <mgorny@g.o> wrote: |
4 |
> |
5 |
> > While the large number of flags is practically invisible to user with |
6 |
> > all the USE_EXPAND hiding, it negatively impacts pkgcheck. When |
7 |
> > the number reached 10, CI became unusable. We're currently back down |
8 |
> > to 8, thanks to powerpc team, but the problem is going to happen again |
9 |
> > sooner or later. Ideally we'd improve pkgcheck but I'm not aware of |
10 |
> > anyone having a good idea how to do it. |
11 |
> |
12 |
> While I don't disagree with your rationale below, I think this |
13 |
> motivation is the wrong one: What sort of algorithm does it use |
14 |
> to explode when going from 8 to 10 flags ?!? |
15 |
|
16 |
A poor one. |
17 |
|
18 |
> |
19 |
> There's multilib that adds a lot of flags with a single eclass change, |
20 |
> but I'd guess the number of packages and flags is constantly growing, |
21 |
> so sooner or later you'll be hit by this again and no multilib killing |
22 |
> will help you then. |
23 |
> |
24 |
> I think it is more future proof to use the addition of multilib flags |
25 |
> to fix pkgcheck rather than actively reducing the number of multilib |
26 |
> flags to cope with its limitations. |
27 |
|
28 |
Then please do it, by all means. The reality is simple. If the tool is |
29 |
broken, you either fix it or stop doing what you know that breaks it. |
30 |
Being unable to do the former, and having no good replacement, I'd go |
31 |
for the latter. |
32 |
|
33 |
> Also, remember that multilib is not entirely about skype or slack, |
34 |
> this was made with multibin in mind too: for example an ABI may perform |
35 |
> better than another one on specific workflows (x32) and it may make |
36 |
> sense to use this abi for a specific binary (which would be manually |
37 |
> built for now). |
38 |
|
39 |
No, it weren't. Just because someone found it convenient to use the |
40 |
design beyond its original purpose doesn't mean it had a different |
41 |
purpose. Besides, how many 'multibin' packages do we have right now? |
42 |
|
43 |
That said, yes, I am also concerned about people proactively adding |
44 |
multilib to all random libraries which probably will never have any real |
45 |
multilib use case and only waste time of people who have abi_x86_32 |
46 |
enabled globally. |
47 |
|
48 |
-- |
49 |
Best regards, |
50 |
Michał Górny |