#linuxcnc-devel Logs

Sep 25 2017

10:31 AM seb_kuzminsky: that commit message is a thing of beauty we should all aspire to
10:33 AM hazzy: hear! hear!
10:52 AM cradek: umm does anyone here github?
10:52 AM cradek: so there's commit a6e417a5
10:52 AM cradek: it has a log message "fix some more bugz"
10:52 AM cradek: I wanted to find a link for that commit on github
10:53 AM cradek: but it seems like github search never helps me find what I want
10:53 AM cradek: I tried searching for the ref, and the log message
10:53 AM seb_kuzminsky: https://github.com/LinuxCNC/linuxcnc/commit/a6e417a5
10:54 AM hazzy: https://github.com/LinuxCNC/linuxcnc/commit/a6e417a59f94d8b5c93e943166a2080dd9b93799
10:54 AM cradek: ooh, you picked any commit and edited the url?
10:54 AM cradek: or is there actually a search?
10:54 AM seb_kuzminsky: history search on github does seem pretty terrible
10:54 AM seb_kuzminsky: i just remembered that you can say github.com/$ORG/$REPO/commit/$SHA and get it that way :-/
10:56 AM hazzy: I searched repo for "fix some more bugz" and it came right up :)
10:56 AM seb_kuzminsky: i searched for that too and nothing came up
10:56 AM cradek: where? I typed it into "This repository:" at the top
10:57 AM hazzy: cradek: "This repository:" and then select commits tab
10:58 AM hazzy: https://github.com/LinuxCNC/linuxcnc/search?q=fix+some+more+bugz&type=Commits&utf8=%E2%9C%93
10:58 AM seb_kuzminsky: oooh
10:58 AM seb_kuzminsky: the search results has many tabs for different kinds of results
10:58 AM seb_kuzminsky: there was 1 hit in the "commtis" tab
10:58 AM seb_kuzminsky: thanks hazzy!
10:58 AM hazzy: sure!
11:12 AM cradek: oooh
11:14 AM jepler: That particular contributor did not have his heart in the right place when making those commits.
11:14 AM jepler: but happily that's all ancient historyn ow
01:15 PM skunkworks: jepler, thanks! I will play with it some more
01:43 PM jepler: I understand, you'd like to give up as little travel as possible while keeping the home switch outside the soft limits
01:44 PM jepler: I assume it's things like feedback position error that affect it more than floating point inaccuracy, at the scale of .005 you mentioned earlier
01:44 PM jepler: but even so, .005 (inch) is a lot compared to the feedback position error you'd hope for .. so maybe it's something else
02:19 PM skunkworks: I would not have a problem giving up .050 or more
02:19 PM skunkworks: not a big deal
02:42 PM seb_kuzminsky: maybe we should ignore soft limits during the end of homing, while traversing from HOME_OFFSET to HOME?
02:54 PM jepler: seb_kuzminsky: yes, if we're not now I think we should.
03:13 PM seb_kuzminsky: one of the guys at work recommended this website for beginners to learn git: http://learngitbranching.js.org/
04:01 PM cradek: seb_kuzminsky: it feels like all the arrows point the wrong way (backward in time) and the diagrams are upside-down (opposing the presentation in gitk)
04:02 PM seb_kuzminsky: the arrows match the git commit's "Parent" field, so those seem ok to me
04:02 PM cradek: I really like the idea of it, but it seems confusing.
04:02 PM seb_kuzminsky: i agree that it's upside down, with time increasing as you go down, instead of time increasing as you go up (like it does in "git log" output)