| 1 |
begin quote |
| 2 |
On Thu, 5 Feb 2004 20:34:53 +0100 |
| 3 |
Paul de Vrieze <pauldv@g.o> wrote: |
| 4 |
|
| 5 |
> On Thursday 05 February 2004 19:45, Spider wrote: |
| 6 |
> > |
| 7 |
> > But in a case like this it isnt a SLOT. All openssl builds have |
| 8 |
> > SLOT="0", they just provide diffrent library versions (0.9.6 and |
| 9 |
> > 0.9.7) which cause this breakage. Same goes against other such |
| 10 |
> > interesting"known to break" things. (libpng was another pretty |
| 11 |
> > example. ) |
| 12 |
> |
| 13 |
> In that case we might use something like LIBRARYVERSION/LIBVER (we |
| 14 |
> might even by default calculate it from the soname of the first |
| 15 |
> library found) |
| 16 |
> |
| 17 |
|
| 18 |
Could work, but isn't this a clumsier version of : |
| 19 |
track ldd of foopack. |
| 20 |
resolve all .so files |
| 21 |
store both lists, and if the .so aren't there, spit out the "last seen |
| 22 |
as "<package>" as a hint and refuse to install? |
| 23 |
|
| 24 |
|
| 25 |
(still dirty.. but better than what RPM does ;) |
| 26 |
|
| 27 |
//Spider |
| 28 |
|
| 29 |
|
| 30 |
-- |
| 31 |
begin .signature |
| 32 |
This is a .signature virus! Please copy me into your .signature! |
| 33 |
See Microsoft KB Article Q265230 for more information. |
| 34 |
end |
Archives