| 1 |
Donnie Berkholz wrote: |
| 2 |
> I hacked together a script this afternoon to find any packages that are |
| 3 |
> not yet ported to modular X. It will only work on systems _with modular |
| 4 |
> X installed_. This is because it works by using emerge to check for |
| 5 |
> blockers, then resolving those blockers down to a single package. |
| 6 |
> |
| 7 |
> Try it out and let me know how it works. If you want to watch it at work |
| 8 |
> and have confidence that it didn't freeze or anything, add "set -o |
| 9 |
> xtrace" up near the top. For my Athlon 2000+, it takes about 1.5 minutes |
| 10 |
> to track down 2 packages in my 650-package world file. |
| 11 |
> |
| 12 |
> If you find bugs, please report them with a patch attached. If you can't |
| 13 |
> come up with a patch, definitely include your world file so others can |
| 14 |
> try to reproduce. |
| 15 |
|
| 16 |
My current script seems to miss some things. The reason this happens is |
| 17 |
that the group of applications they're emerged in also contains some app |
| 18 |
that breaks the emerge, perhaps because it's in package.mask or isn't |
| 19 |
keyworded at least ~x86. Broken emerges don't get far enough to produce |
| 20 |
blocking output, so it's falsely assumed to be a success for that group. |
| 21 |
|
| 22 |
marienz suggested an idea for error handling last night, and I'll work |
| 23 |
to figure something out. |
| 24 |
|
| 25 |
Thanks, |
| 26 |
Donnie |
Archives