The help for option -g "Gauge the number of times a package has been merged" is misleading and imho wrong. -t seems to produce output more like what -g should be. The descriptions seem backwards at best, and perhaps the 'gauge' option needs to be rewritten entirely due to its vagueness. Reproducible: Always Steps to Reproduce: 1. qlop -t package-name 2. genlop -t package-name 3. qlop -g package-name Actual Results: 1: qlop -t openoffice-bin openoffice-bin: 55 seconds average for 5 merges 2: genlop -t openoffice-bin * app-office/openoffice-bin Thu Apr 28 04:17:46 2011 >>> app-office/openoffice-bin-3.3.0 merge time: 1 minute and 36 seconds. Tue May 3 13:14:46 2011 >>> app-office/openoffice-bin-3.3.0 merge time: 46 seconds. Tue May 3 13:33:48 2011 >>> app-office/openoffice-bin-3.3.0 merge time: 45 seconds. Tue May 3 14:02:15 2011 >>> app-office/openoffice-bin-3.3.0 merge time: 44 seconds. Wed May 4 00:34:32 2011 >>> app-office/openoffice-bin-3.3.0 merge time: 44 seconds. 3: mike@odin ~ $ qlop -g openoffice-bin openoffice-bin: Thu Apr 28 04:16:10 2011: 96 seconds openoffice-bin: Tue May 3 13:14:00 2011: 46 seconds openoffice-bin: Tue May 3 13:33:03 2011: 45 seconds openoffice-bin: Tue May 3 14:01:31 2011: 44 seconds openoffice-bin: Wed May 4 00:33:48 2011: 44 seconds Expected Results: 1: openoffice-bin: Thu Apr 28 04:16:10 2011: 96 seconds ... 3: 55 seconds average for 5 merges I started out using genlop. Then I found qlop. Works similarly but not entirely the same. With genlop the -t option shows the time it took to merge each package, per merged iteration. The -t option for qlop shows an averaged time and count-of-times-merged. To get a similar output, I tried a few things: I combined this with the -l list command; This produced nothing but the list. I thought perhaps make it more verbose with qlop -t -v. Nope. Its that darned -g option that sounds like what -t actually outputs. "Gauging the number of times a package has been merged" sounds like "5 merges". That's the only output I would expect with that sort of description. The 55 seconds average to me seems like a bonus yet uncalled for tidbit, if we are going strictly on 'gauging number of times'... Incidentally, the -t options provide similar help between both programs: genlop: -t calculate merge time for the specific package(s). qlop : -t, --time * Calculate merge time for a specific package
1: qlop -t automake automake: 51 seconds average for 10 merges 2: qlop -gv automake automake-1.14.1: Tue Jun 14 15:58:01 2016: 29 seconds automake-1.15: Tue Jun 14 17:08:20 2016: 31 seconds automake-1.15-r2: Sun Mar 26 17:31:22 2017: 43 seconds automake-1.14.1-r1: Sun Mar 26 17:32:05 2017: 41 seconds automake-1.15.1: Wed Jun 21 08:54:33 2017: 94 seconds automake-1.15.1-r1: Mon Aug 21 15:54:14 2017: 48 seconds automake-1.15.1-r2: Mon Mar 5 18:47:00 2018: 51 seconds automake-1.16-r1: Mon Mar 5 18:47:51 2018: 44 seconds automake-1.16.1: Tue Mar 13 09:54:00 2018: 75 seconds automake-1.16.1-r1: Wed Mar 14 10:20:25 2018: 63 seconds automake: 10 times 3: similar to 1: different programs have different options. genlop and qlop aren't comparable, IMO. qlop sees "time" as global number for all merges, genlop treats it as "time" per merge.
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/proj/portage-utils.git/commit/?id=fe42f2a99297fed36720e71ba2ed77cf7a9da804 commit fe42f2a99297fed36720e71ba2ed77cf7a9da804 Author: Fabian Groffen <grobian@gentoo.org> AuthorDate: 2018-03-31 19:04:20 +0000 Commit: Fabian Groffen <grobian@gentoo.org> CommitDate: 2018-03-31 19:04:20 +0000 man/qlop: add some more/better documentation Bug: https://bugs.gentoo.org/335453 Bug: https://bugs.gentoo.org/367199 man/qlop.1 | 40 +++++++++++++++++++++++----------------- 1 file changed, 23 insertions(+), 17 deletions(-)}
