| 1 |
On Wednesday 03 January 2007 15:36, Bo Ørsted Andresen wrote: |
| 2 |
> On Wednesday 03 January 2007 09:17, Alan McKinnon wrote: |
| 3 |
|
| 4 |
[snip] |
| 5 |
|
| 6 |
> > Affected packages are mostly beryl, emerald, gimp, gtk+2 and the |
| 7 |
> > e17 supporting libs. |
| 8 |
> |
| 9 |
> It have been better to include the full output of `revdep-rebuild -i |
| 10 |
> -vp` either inline or attached as a compressed file... |
| 11 |
|
| 12 |
I didn't do that as it's very long, and I was actually looking for a |
| 13 |
deeper understanding of what's going on below the covers. Which I think |
| 14 |
you might have given below |
| 15 |
|
| 16 |
[SNIP] |
| 17 |
|
| 18 |
> > I get the revdep-rebuild output - the .la lists a lib that is not |
| 19 |
> > longer there. But why the compile error? pygtk has no directfb USE |
| 20 |
> > flag, and I can't figure out why it would want to grep |
| 21 |
> > /usr/lib/libdirectfb.la. |
| 22 |
> |
| 23 |
> [SNIP] |
| 24 |
> |
| 25 |
> Because it links against cairo which is still linked against |
| 26 |
> libdirectfb.la. revdep-rebuild makes no attempt (or fails ;) to |
| 27 |
> figure out in which order the packages should remerged. Cairo, pango |
| 28 |
> and gtk+ are probably on the list from revdep-rebuild and should be |
| 29 |
> remerged first (in that order)... |
| 30 |
|
| 31 |
Ah, this makes sense, and I hadn't considered this indirection. emerge |
| 32 |
world is still running so when it finishes I'll let you know if it |
| 33 |
fixes things. |
| 34 |
|
| 35 |
Thanks, |
| 36 |
|
| 37 |
alan |
| 38 |
|
| 39 |
|
| 40 |
-- |
| 41 |
gentoo-user@g.o mailing list |
Archives