| 1 |
On 27/09/2013 00:12, Ian Stakenvicius wrote: |
| 2 |
> -----BEGIN PGP SIGNED MESSAGE----- |
| 3 |
> Hash: SHA256 |
| 4 |
> |
| 5 |
> On 26/09/13 06:51 AM, Michael Palimaka wrote: |
| 6 |
>> On 26/09/2013 17:53, Michał Górny wrote: |
| 7 |
>>> How do we handle packages which install multiple libraries? I'm |
| 8 |
>>> afraid forcing such a policy and/or hurrying developers to adapt |
| 9 |
>>> will only cause more of poppler-like issues to occur. |
| 10 |
>>> |
| 11 |
>> There isn't a 100% perfect solution currently, and I agree that |
| 12 |
>> hurrying people will simply move us from "not enough rebuilds" to |
| 13 |
>> "too many rebuilds". |
| 14 |
> |
| 15 |
> Enforcing consistency is much more important imo than "emerge -uDN |
| 16 |
> @world" efficiency. For those users that need more efficiency they |
| 17 |
> can always get it by upgrading individual packages with |
| 18 |
> '--rebuild-ignore' or '--ignore-built-slot-operator-deps y' after |
| 19 |
> seeing what all is going to be rebuilt via 'emerge -uDNav' |
| 20 |
Why do you think striving for correct subslot usage will reduce |
| 21 |
consistency? I am not saying we should avoid subslots completely for a |
| 22 |
package because of some edge case, but excessive unnecessary rebuilds |
| 23 |
will push users towards disabling the feature. |
| 24 |
|
| 25 |
Subslots for poppler was an improvement on the previous situation and a |
| 26 |
great idea for libpoppler consumers. For packages that that use one of |
| 27 |
the stable interfaces, rebuilding them needlessly is a major annoyance, |
| 28 |
and definitely not what subslots were intended for. Why can't we limit |
| 29 |
our subslot usage to where it's actually useful (consumers of the |
| 30 |
unstable library)? It doesn't have to be all-or-none. |
| 31 |
|
| 32 |
What about when the subslot of boost was equal to ${PV}? Was it really a |
| 33 |
good idea to make everyone rebuild half their system for a bugfix |
| 34 |
release, without even checking if the ABI changed? |
| 35 |
|
| 36 |
>> Poppler was a great example of what can go wrong. Apart from |
| 37 |
>> people being forced to rebuild packages that link only against one |
| 38 |
>> of the stable interfaces, I even saw rebuilds forced for packages |
| 39 |
>> that didn't even link against the libraries. |
| 40 |
> |
| 41 |
> The latter in that case was a mis-use of the ':=' on the poppler atom |
| 42 |
> in *DEPEND. |
| 43 |
Yes, misuse is what I'm complaining about. When used correctly, subslots |
| 44 |
are great. |
Archives