1 |
Catalin Grigoroscuta wrote: |
2 |
|
3 |
>> I've just rebuilt my system without "motif" use flag (with emerge |
4 |
>> --newuse, etc). |
5 |
>> Everything works fine, except that equery shows stale data. |
6 |
>> For example: |
7 |
>> 1. "equery depends emacs" shows openmotif as dependency, but emacs is |
8 |
>> rebuilt without motif USE ("emerge -pv emacs" clearly shows this). |
9 |
|
10 |
The depends query shows all depends regardless of whether they are enabled or not (by USE flags). |
11 |
|
12 |
>> 2. "equery hasuse motif" also shows emacs |
13 |
>> Is there any caching done by equery? How can I invalidate it? |
14 |
|
15 |
The hasuse query shows all USE flags regardless of whether they are enabled or not. |
16 |
|
17 |
There is no cache, it queries the database directly. |
18 |
|
19 |
> I did not use esync, I did not even know about it, and I do not recall |
20 |
> seeing it in install docs... |
21 |
> It is quite strange why equery (which is in gentoolkit package) relies |
22 |
> on eupdatedb (which is in esearch package), and gentoolkit does not |
23 |
> depend on esearch. |
24 |
> If it were not for your answer, I wouldn't have known about esync and |
25 |
> eupdatedb. |
26 |
> I think maybe the install docs should be updated, or at least put |
27 |
> esearch as a dependency of gentoolkit. |
28 |
|
29 |
Gentoolkit and equery do not depend on esearch. |
30 |
|
31 |
> Even with eupdtaedb, same result. |
32 |
> I think this might be a bug in equery. Should I put it in bugzilla? |
33 |
> Here is the proof for what I think is a bug: see the output of "equery |
34 |
> hasuse motif" and "equery uses emacs". The results are contradictory. |
35 |
|
36 |
I explained this above. If you file a bug then AFAIKT it's a feature request. |
37 |
|
38 |
Zac |
39 |
-- |
40 |
gentoo-user@g.o mailing list |