| 1 |
On Sun, Mar 29, 2015 at 8:23 PM, Rich Freeman <rich0@g.o> wrote: |
| 2 |
> I think we really need to either stabilize 4.8.6, or backport |
| 3 |
> qtchooser/multilib/etc to the current stable version. |
| 4 |
> |
| 5 |
|
| 6 |
Backporting is not an option. The introduction of multilib support in |
| 7 |
qt4 required extensive changes to the eclass (substantial portions of |
| 8 |
code were rewritten) and the ebuilds. |
| 9 |
|
| 10 |
Therefore stabilizing 4.8.6-r1 is the way to go. We had surprisingly |
| 11 |
few regressions, but the ones that are still unsolved are proving |
| 12 |
quite hard to debug. Please help with the blockers of bug #530238. |
| 13 |
|
| 14 |
As usual, at least for me, the problem is the lack of time for gentoo |
| 15 |
development, plus the fact that very few people are actively working |
| 16 |
on qt{4,5} packaging (kensington, me, and more recently yngwin). |
| 17 |
|
| 18 |
> qt is a pretty significant package to have break with multilib, and |
| 19 |
> trying to run qt-5 on a stable system is already a nightmare with the |
| 20 |
> qtchooser switch (in my case I ended up abandoning qt5 as I didn't |
| 21 |
> need it that badly). |
| 22 |
> |
| 23 |
|
| 24 |
I'm not sure how qt5 is relevant to this discussion. |
| 25 |
|
| 26 |
Thanks, |
| 27 |
Davide |
Archives