| 1 |
On Wed, 2006-04-12 at 17:27 +0200, Paul de Vrieze wrote: |
| 2 |
> While this also does not stand any chance to win a beauty contest, this |
| 3 |
> would also be acceptable to me. The only thing I want is things not |
| 4 |
> starting to fail without a good reason. A new version of a library is on |
| 5 |
> itself not a good enough reason. |
| 6 |
|
| 7 |
It is just a matter of fact that a few upstream authors don't really get |
| 8 |
libtool versioning right, but to say a new .so nr. means a new slot |
| 9 |
version is wrong. Packages should only be slotted if every single bit of |
| 10 |
them is versioned, check glib, gtkhtml, etc. |
| 11 |
|
| 12 |
The proper course of action is making upstream aware of how they should |
| 13 |
do libtool versioning. Adding workarounds like preserve_old_lib* should |
| 14 |
only be a temporary measure. |
| 15 |
|
| 16 |
- foser |
Archives