| 1 |
-----BEGIN PGP SIGNED MESSAGE----- |
| 2 |
Hash: SHA256 |
| 3 |
|
| 4 |
On 14/06/14 10:41 AM, Micha³ Górny wrote: |
| 5 |
> Hi, |
| 6 |
> |
| 7 |
> Some time ago we've got bug #510780 [1] asking us to bump subslot |
| 8 |
> on LLVM even though the new version was ABI-compatible with |
| 9 |
> previous one. It was because it introduced new APIs which |
| 10 |
> applications could make use of. Since I believe this is a wider |
| 11 |
> issue, I would like to know the opinion of our community about |
| 12 |
> this. |
| 13 |
> |
| 14 |
> More specifically: do we want subslots to change only when |
| 15 |
> backwards- incompatible ABI changes are done -- alike SONAME -- or |
| 16 |
> whenever any ABI change is done? The problem seems a bit complex. |
| 17 |
> |
| 18 |
> Considering the libtool versioning, there are two kinds of library |
| 19 |
> bumps relevant to us: |
| 20 |
> |
| 21 |
> 1) when ABI is altered in backwards-compatible way (so old stuff is |
| 22 |
> not touched), |
| 23 |
> |
| 24 |
> 2) when ABI is altered in backwards-incompatible way. |
| 25 |
> |
| 26 |
|
| 27 |
I vote that as primary policy/general practice, it only be bumped for |
| 28 |
(2) -- the primary purpose of subslot rebuilds is to allow portage to |
| 29 |
figure out the deptree order when a dependency upgrade is going to |
| 30 |
break a package that may or may not be emerged later. "break" is the |
| 31 |
key term here. If users want to re-emerge the rdeps of a package on |
| 32 |
upgrade they can certainly do so, but I don't see this as something we |
| 33 |
want to force on everybody just because we can... |
| 34 |
|
| 35 |
|
| 36 |
-----BEGIN PGP SIGNATURE----- |
| 37 |
Version: GnuPG v2.0.22 (GNU/Linux) |
| 38 |
|
| 39 |
iF4EAREIAAYFAlOccrEACgkQ2ugaI38ACPBu1AD+LNiTezb0nnGtGoVW4AHjAMk7 |
| 40 |
sMxoTYTvcYn2MLfYrrAA/iXLTPsTdGUSQcWnq40zz5yK09RljYMlI7f2bk5SlWIt |
| 41 |
=x/MD |
| 42 |
-----END PGP SIGNATURE----- |
Archives