| 1 |
>>>>> "ZM" == Zac Medico <zmedico@g.o> writes: |
| 2 |
|
| 3 |
ZM> That commit allows backtracking to continue in the event that unsolved |
| 4 |
ZM> blockers are encountered, solving this bug: |
| 5 |
|
| 6 |
ZM> https://bugs.gentoo.org/show_bug.cgi?id=465638 |
| 7 |
|
| 8 |
I read the bug. But it was the only commit that at first glance *could* |
| 9 |
have been the issue. |
| 10 |
|
| 11 |
ZM> That means that emerge may spend some more time backtracking in some |
| 12 |
ZM> cases involving blockers, even though the backtracking may not |
| 13 |
ZM> ultimately lead to a useful solution. |
| 14 |
|
| 15 |
Five hours more, in my case. |
| 16 |
|
| 17 |
ZM> Did your test case involve blockers? Are you using the --backtrack |
| 18 |
ZM> option to increase the amount of backtracking that is allowed? |
| 19 |
|
| 20 |
Yes. There are several blockers and at some point in the fuzzy past I |
| 21 |
added --backtrack=30 to get around some issue which was preventing the |
| 22 |
automated sync && pv from reporting any possible updates. |
| 23 |
|
| 24 |
The larger backtrace allowed pv to complete. Then the scripts use -O |
| 25 |
to merge a few packages at a time in the pv order. |
| 26 |
|
| 27 |
As a side note, my next workstation will split most stuff into smaller |
| 28 |
topic-spcific VMs so as to avoid the annoying conflicts which occur when |
| 29 |
one tries to keep everything one might need available at once. |
| 30 |
|
| 31 |
-JimC |
| 32 |
-- |
| 33 |
James Cloos <cloos@×××××××.com> OpenPGP: 1024D/ED7DAEA6 |
Archives