| 1 |
Dnia 11 stycznia 2018 07:54:40 CET, Mike Gilbert <floppym@g.o> napisał(a): |
| 2 |
>On Wed, Jan 10, 2018 at 11:10 PM, Michał Górny <mgorny@g.o> |
| 3 |
>wrote: |
| 4 |
>> W dniu śro, 10.01.2018 o godzinie 21∶45 -0500, użytkownik Mike |
| 5 |
>Gilbert |
| 6 |
>> napisał: |
| 7 |
>>> On Wed, Jan 10, 2018 at 5:56 PM, Zac Medico <zmedico@g.o> |
| 8 |
>wrote: |
| 9 |
>>> > On 01/10/2018 02:24 PM, Michał Górny wrote: |
| 10 |
>>> > > Enable repoman checks on exp profiles by default to improve |
| 11 |
>>> > > the dependency graph integrity on those profiles and help them |
| 12 |
>on their |
| 13 |
>>> > > way towards stable status. This is possible now that the |
| 14 |
>dependency |
| 15 |
>>> > > graph problems are warnings rather than errors. |
| 16 |
>>> > > --- |
| 17 |
>>> > > repoman/pym/repoman/argparser.py | 2 +- |
| 18 |
>>> > > 1 file changed, 1 insertion(+), 1 deletion(-) |
| 19 |
>>> > > |
| 20 |
>>> > > diff --git a/repoman/pym/repoman/argparser.py |
| 21 |
>b/repoman/pym/repoman/argparser.py |
| 22 |
>>> > > index f32972288..d49147366 100644 |
| 23 |
>>> > > --- a/repoman/pym/repoman/argparser.py |
| 24 |
>>> > > +++ b/repoman/pym/repoman/argparser.py |
| 25 |
>>> > > @@ -167,7 +167,7 @@ def parse_args(argv, qahelp, |
| 26 |
>repoman_default_opts): |
| 27 |
>>> > > |
| 28 |
>>> > > parser.add_argument( |
| 29 |
>>> > > '-e', '--include-exp-profiles', choices=('y', |
| 30 |
>'n'), metavar='<y|n>', |
| 31 |
>>> > > - default=False, |
| 32 |
>>> > > + default='y', |
| 33 |
>>> > > help='include exp profiles in dependency checks') |
| 34 |
>>> > > |
| 35 |
>>> > > parser.add_argument( |
| 36 |
>>> > > |
| 37 |
>>> > |
| 38 |
>>> > We have dev and exp profiles disabled by default because the time |
| 39 |
>>> > consumed by repoman is proportional to the number of profiles. |
| 40 |
>>> > |
| 41 |
>>> > The current counts are as follows: |
| 42 |
>>> > |
| 43 |
>>> > stable 87 |
| 44 |
>>> > dev 88 |
| 45 |
>>> > exp 149 |
| 46 |
>>> |
| 47 |
>>> Yeah, I really don't like the idea of making repoman even slower by |
| 48 |
>default. |
| 49 |
>> |
| 50 |
>> The alternative is to go for all profiles being stable. Because |
| 51 |
>> accepting developers randomly breaking profiles 'because repoman |
| 52 |
>speed' |
| 53 |
>> is not acceptable. |
| 54 |
> |
| 55 |
>I disagree. Many of these are profiles that are seldom used and very |
| 56 |
>few users are affected by any breakage. |
| 57 |
> |
| 58 |
>There's a balance to be struck here, and I think it's pretty good |
| 59 |
>right where it is. |
| 60 |
|
| 61 |
Just to be clear, the current counts are irrelevant. The goal is to have a second status that covers profiles on their way to becoming stable. This patch is part of that thread. |
| 62 |
|
| 63 |
|
| 64 |
-- |
| 65 |
Best regards, |
| 66 |
Michał Górny (by phone) |
Archives