| 1 |
Simon Stelling wrote: |
| 2 |
> I hate self-replying, but oh well: |
| 3 |
> |
| 4 |
> Simon Stelling wrote: |
| 5 |
> |
| 6 |
>> Hrm, I can't find one at best will. Could you show me such a bug? |
| 7 |
> |
| 8 |
> |
| 9 |
> I finally found bug 8423 which provides a neat script as well as a patch. |
| 10 |
> |
| 11 |
|
| 12 |
Bug 90343 would go along well with 8423 in order to reduce false positives when trying to identify orphans. It uses sandbox debug logging to provide an "installwatch" feature for ebuild pkg_preinst and pkg_postinst. |
| 13 |
|
| 14 |
I'm currently working on a portage database indexer. It uses a forked version of portage's depcache plugin framework. One of the features is a tool that makes the contents/installwatch data useful for cleaning up orphans. It get's quite good performance with it's sqlite backend. |
| 15 |
|
| 16 |
My goal is to generalize the plugin framework some more so that it's useful for many different types of portage queries such as those provided by equery and eix/esearch. |
| 17 |
|
| 18 |
Zac |
| 19 |
-- |
| 20 |
gentoo-portage-dev@g.o mailing list |
Archives