| 1 |
-----BEGIN PGP SIGNED MESSAGE----- |
| 2 |
Hash: SHA256 |
| 3 |
|
| 4 |
On 05/08/13 01:07 PM, Rich Freeman wrote: |
| 5 |
> On Mon, Aug 5, 2013 at 12:15 PM, Alexis Ballier |
| 6 |
> <aballier@g.o> wrote: |
| 7 |
>> On Mon, 5 Aug 2013 18:10:46 +0200 Michał Górny |
| 8 |
>> <mgorny@g.o> wrote: |
| 9 |
>>> We can simply have multiple virtual versions, each depending on |
| 10 |
>>> the proper jpeg & jpeg-turbo versions. |
| 11 |
>> |
| 12 |
>> you can do it that way, yes. |
| 13 |
>> |
| 14 |
>> what will you do when jpeg 10 is out or when libjpeg-turbo |
| 15 |
>> changes its abi ? wait for each provider to have a release that |
| 16 |
>> changes the abi ? this doesnt sound sane |
| 17 |
> |
| 18 |
> I think a thorough solution needs to handle the passthrough |
| 19 |
> situation, and ideally the multiple libs per package situation. |
| 20 |
> |
| 21 |
> Our whole versioning system wasn't really set up with this stuff |
| 22 |
> in mind either. We're dealing with ABI/SONAME versions here - not |
| 23 |
> functionality versions. A package with jpeg-turbo SONAME 0.6 might |
| 24 |
> be a superset of all the functionality in SONAME 0.9 - the larger |
| 25 |
> number doesn't mean better, just different. |
| 26 |
> |
| 27 |
> It might still be possible to handle this with subslots as they |
| 28 |
> are currently implemented, but I'd have to really mess around with |
| 29 |
> it to be sure. The logic of higher-number = more-desirable is |
| 30 |
> likely to cause problems - you don't want your system to force you |
| 31 |
> to move from jpeg-turbo to jpeg because only the latter is a dep of |
| 32 |
> a newer virtual. |
| 33 |
|
| 34 |
This is the primary reason as to why sub-slots were designed as they |
| 35 |
were, to be an independently set value -- it allows the maintainer of |
| 36 |
say, jpeg-turbo to bump the subslot value only when necessary (for |
| 37 |
instance, in the case of SONAME 0.6 being a complete superset of 0.9, |
| 38 |
there is no need to bump sub-slot as long as it is known that all |
| 39 |
consumers are only consuming the subset of functions available in 0.9 |
| 40 |
(and if they aren't, I expect there's going to be compile-time |
| 41 |
breakage anyhow). |
| 42 |
|
| 43 |
This is starting to go off-topic, but it's one of the reasons I don't |
| 44 |
like seeing packages with SLOT="0/${PV}" or similar -- to me, a |
| 45 |
sub-slot bump should really be evaluated on its own merits and not |
| 46 |
just for any version bump. |
| 47 |
|
| 48 |
-----BEGIN PGP SIGNATURE----- |
| 49 |
Version: GnuPG v2.0.20 (GNU/Linux) |
| 50 |
|
| 51 |
iF4EAREIAAYFAlIBBZwACgkQ2ugaI38ACPCuuAD9HhPrGA5A7BN4FM8X58la/r0+ |
| 52 |
wpzueqNlPFAk9WCZkScA/0gfqkYP9GuAFBtQPq8zbEiEqX8DDRZ03FQYISSlrv8B |
| 53 |
=bSZH |
| 54 |
-----END PGP SIGNATURE----- |
Archives