#linuxcnc-devel | Logs for 2015-11-15

[10:49:40] <micges> jepler: I've got 'invalid security code' while sending email to forum user
[11:06:09] <jepler> micges: same here, I don't know why.
[11:07:52] <jepler> micges: ah, it was due to the captcha plugin
[11:08:04] <jepler> it was set not to display on that page, but apparently it checks that the captcha is correctly filled out anyway
[11:08:09] <jepler> so I put the captcha on that page
[11:08:15] <jepler> does it work for you now?
[11:11:35] <micges> I've got mail from you so seems to work
[11:11:39] <micges> thanks
[11:23:20] <micges> jepler: yep works
[17:06:48] <seb_kuzminsky> i dislike how halui's check_hal_changes() manages the old (previous) and new (current) hal pin states
[17:07:28] <seb_kuzminsky> it transitions information from new to old as it checks things, so inside the function, old is halfway updated, and which parts are updated depends on where in the function you are
[17:07:45] <seb_kuzminsky> i'd prefer if it left old alone until the very end, and then memcpy
[17:07:55] <seb_kuzminsky> memcpy'ed new to old
[17:08:30] <seb_kuzminsky> but that's too big a change to do in the middle of 2.6 and 2.7
[17:08:48] <seb_kuzminsky> maybe in master
[17:16:08] <KGB-linuxcnc> 03Dewey Garrett 05features_preview cac0ffd 06linuxcnc 10(20 files in 14 dirs) features_preview: relocate placeholders, +cleanups * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=cac0ffd
[17:16:08] <KGB-linuxcnc> 03FernV 05features_preview fc7e8f7 06linuxcnc Merge branch 'features_preview' of ssh://git.linuxcnc.org/git/linuxcnc into features_preview * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=fc7e8f7
[17:18:28] <seb_kuzminsky> aww yeah
[20:00:22] <dgarr> seb_kuzminsky: i am not understanding what fern's commit did http://www.panix.com/~dgarrett/stuff/fmsg.txt can you help?
[20:04:33] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 05c98cd 06linuxcnc 10tests/halui/jogging/postgui.hal 10tests/halui/jogging/test-ui.py halui/jogging test: change which joint is selected while jogging * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=05c98cd
[20:04:33] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 69aa9e6 06linuxcnc 10src/emc/usr_intf/halui.cc halui: fix some jogging bugs * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=69aa9e6
[20:05:35] <seb_kuzminsky> dgarr: looking...
[20:07:57] <seb_kuzminsky> dgarr: yeah, something went wonky with fern's push
[20:11:14] <seb_kuzminsky> his commit cac0ffd8 has many of the same changes as you made in 0b349088, but some different things too, and (as you say) your commit message and author info and things
[20:11:27] <seb_kuzminsky> i'm not sure how he ended up in that situation
[20:11:35] <seb_kuzminsky> git's sure not easy to pick up and start using :-(
[20:12:28] <dgarr> my suggestion is to first learn how to make patches with git format-patch and let folks like you look at them first. Do you think that would help?
[20:13:19] <seb_kuzminsky> git format-patch makes patch files out of git commits
[20:13:27] <seb_kuzminsky> so a prerequisite for that is to make proper git commits
[20:13:48] <dgarr> yes -- making proper commits before pushing is what i'm suggesting
[20:14:07] <seb_kuzminsky> i think a good step for fern would be to have his own git repo that he can push to, for example a clone of jepler's linuxcnc repo on github
[20:14:34] <seb_kuzminsky> he can do his work, push it there for review, and we (you or i) can merge it into your branch when it's ready
[20:15:29] <seb_kuzminsky> i think it's more convenient for us to work with his commits when they're in the form of a branch we can pull and inspect with gitk or git log or whatever, rather than a bunch of attachments to an email on the mailing lists
[20:17:07] <dgarr> from his second post to the list, it is not clear he knows you to run and test using rip build (yet)
[20:17:08] <seb_kuzminsky> you can see what's different between you 0b349088 and his cac0ffd8 with this command: git diff --stat -p cac0ffd8..0b349088
[20:20:43] <seb_kuzminsky> dgarr: i think jepler pointed him at the run-in-place instructions in the top-level README of our git repo
[20:21:00] <seb_kuzminsky> i'm not sure if he had any questions about it after that...
[20:27:05] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 d5f635c 06linuxcnc 10src/emc/usr_intf/halui.cc Merge remote-tracking branch 'origin/2.6' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d5f635c
[20:36:19] <skunkworks> jepler: did you see https://forum.linuxcnc.org/forum/20-g-code/29685-g76-pause-in-z-during-exit-taper-linux-cnc-2-7-0?start=10#65302
[20:37:04] <dgarr> seb_kuzminsky: using the git diff command you mentioned, i was able to reset the branch and apply the diff , should i go ahead and force push that to the branch?
[20:37:14] <skunkworks> rob made spindle synced motion fall back to parabolic blending. I think that would be just fine. (don't remeber ever hearing about someone having problems with it.
[20:37:16] <seb_kuzminsky> skunkworks: that's great
[20:38:00] <seb_kuzminsky> dgarr: is your goal to reset the features_preview branch to be identical to your commit 0b3490888789cce1f7f40a08605fea382d4284af?
[20:38:12] <seb_kuzminsky> (that seems like a good goal to me, fwiw)
[20:38:22] <seb_kuzminsky> if so, i suggest this sequence of commands:
[20:38:25] <seb_kuzminsky> git fetch origin
[20:38:33] <seb_kuzminsky> git checkout features_preview
[20:39:07] <seb_kuzminsky> git merge --ff-only origin/features_preview
[20:39:41] <seb_kuzminsky> git reset --hard 0b3490888789cce1f7f40a08605fea382d4284af
[20:39:52] <seb_kuzminsky> git push --force origin features_preview
[20:39:55] <seb_kuzminsky> that should do it
[20:40:58] <seb_kuzminsky> the fetch just retrieves what's in that branch on origin (glo), you probably have this already
[20:41:12] <seb_kuzminsky> the checkout moves your local working dir to that branch
[20:41:24] <seb_kuzminsky> the merge --ff-only moves your branch ahead to be the same as what's on glo
[20:41:49] <seb_kuzminsky> the 'reset --hard' changes your branch to point to the named commit, and changes your working tree to that commit too
[20:42:23] <seb_kuzminsky> then finally the 'push --force' pushes that branch to glo, overriding git's natural disinclination to rewinding branches (since that loses commits from the history)
[20:43:01] <seb_kuzminsky> then we can work with fern to get his stuff into shape
[20:43:04] <seb_kuzminsky> sound good?
[20:44:00] <dgarr> no -- resetting to 0b349088 had conflicts when i tried to apply his diffs, however, resetting to cac0ffd worked and i could apply the diff
[20:44:26] <seb_kuzminsky> which diff are you applying?
[20:45:05] <dgarr> i created a patch file from git diff --stat -p cac0ffd8..0b349088 >/tmp/fern/diff.txt
[20:45:15] <skunkworks> the fusee was probably the worse case - and that worked great
[20:46:12] <seb_kuzminsky> dgarr: 0b349088 has you as the author and commiter, cac0ffd8 has you as author but fern as committer
[20:46:38] <dgarr> yeah - i totally don't know what he did to do that
[20:46:48] <seb_kuzminsky> yeah that a wacky commit for sure
[20:49:16] <seb_kuzminsky> the diff command i gave earlier is from his commit to your commit, so it's maybe the reverse of the diff we should be working with
[20:49:23] <seb_kuzminsky> sorry i pointed you i the wrong direction...
[20:49:46] <seb_kuzminsky> try git diff $YOURS..$HIS, ie git diff 0b349088..cac0ffd8
[20:50:37] <seb_kuzminsky> so that's the diff that you'd apply to $YOURS, in order to make the tree into $HIS
[20:51:41] <KGB-linuxcnc> 03Sebastian Kuzminsky 05master fb1afd4 06linuxcnc 10tests/halui/jogging/test-ui.py Merge remote-tracking branch 'origin/2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=fb1afd4
[20:53:06] <dgarr> the branch looks ok with the way i did it -- so i think i should just force push it and move on (it will have me as author with credit to fern)
[20:53:08] <dgarr> http://www.panix.com/~dgarrett/stuff/0001-features_preview-catch-up-with-current-development-b.patch
[20:56:57] <seb_kuzminsky> dgarr: is the final result identical to commit fc7e8f7a4db19442d145477cda5fd04504ede28d?
[20:57:04] <seb_kuzminsky> is, to the merge-commit that fern made
[20:58:01] <seb_kuzminsky> *ie
[20:58:23] <skunkworks> I love that recent topics are on every page on the new forum.
[20:58:28] <seb_kuzminsky> you can check with "git diff $THIS..$THAT"
[21:03:48] <dgarr> seb_kuzminsky: i can't make sense of those diff outputs but i am pretty sure the force push i intend will incorporate his patch and make the branch right
[21:06:05] <seb_kuzminsky> i think the tree you push (ie, the contents of your working dir) should be identical to the tree fern pushed, is that right?
[21:06:16] <seb_kuzminsky> i may be misunderstanding your intent here
[21:07:12] <dgarr> my intent is to just incorporate his intended changes on the branch and it will require a force push
[21:07:28] <dgarr> i'm running out of time this evening
[21:08:07] <dgarr> its a mystery what he did, i just want to fix it now
[21:08:31] <seb_kuzminsky> is the final thing that fern pushed the thing you want? ie, does it have all the changes you intended, and all the changes you think he intended?
[21:08:45] <dgarr> yes
[21:08:52] <seb_kuzminsky> ok, then i wouldn't worry about it
[21:09:01] <seb_kuzminsky> i'd just leave the branch the way it is
[21:09:14] <seb_kuzminsky> i thought we were working towards this:
[21:09:15] <seb_kuzminsky> 19:13:50 <+seb_kuzminsky> dgarr: is your goal to reset the features_preview branch to be identical to your commit 0b3490888789cce1f7f40a08605fea382d4284af?
[21:09:18] <seb_kuzminsky> 19:14:02 <+seb_kuzminsky> (that seems like a good goal to me, fwiw)
[21:09:34] <seb_kuzminsky> if you're happy with the tree as it is now, i would not take any action
[21:09:50] <dgarr> the tree now does not have his changes in it
[21:10:07] <dgarr> if you look at the files he wanted to change they are not changed
[21:14:12] <seb_kuzminsky> dgarr: if i run "git diff --stat -p 0b3490888789cce1f7f40a08605fea382d4284af..cac0ffd85526578ece22a8da4dcd43dc3404a1aa
[21:14:27] <seb_kuzminsky> ie, from your commit to the commit he made before his merge commit
[21:14:32] <seb_kuzminsky> then the diffstat has this:
[21:14:46] <seb_kuzminsky> configs/sim/gmoccapy/features_demo/gmoccapy.ini | 1 +
[21:14:46] <seb_kuzminsky> lib/python/gladevcp/features.py | 801 ++++++++++++++++++++++++++------------------------------------------------------
[21:14:49] <seb_kuzminsky> share/features/catalogs/mill/defaults.ngc | 38 +---
[21:14:51] <seb_kuzminsky> share/features/features.glade | 184 +++++++++----------
[21:14:54] <seb_kuzminsky> 4 files changed, 358 insertions(+), 666 deletions(-)
[21:14:57] <seb_kuzminsky> those are the four files he mentioned in his email
[21:15:21] <seb_kuzminsky> i dont know if the changed files have the contents you both intended
[21:15:24] <seb_kuzminsky> i wish fern was here
[21:16:12] <seb_kuzminsky> i'll send fern an email with the things we've talked about
[21:16:22] <seb_kuzminsky> i'll cc you, and we can try to figure out together what should be done
[21:16:25] <dgarr> i have to go, if this push is wrong , apologies in advance
[21:16:38] <seb_kuzminsky> and maybe give fern some git tips at the same time ;-)
[21:16:43] <KGB-linuxcnc> 03Dewey Garrett 05features_preview 653a4c5 06linuxcnc 10configs/sim/gmoccapy/features_demo/gmoccapy.ini 10lib/python/gladevcp/features.py 10share/features/catalogs/mill/defaults.ngc 10share/features/features.glade features_preview:catch-up with current development * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=653a4c5
[21:16:52] <dgarr> thanks
[21:17:00] <seb_kuzminsky> i'll fetch it and check it out
[21:17:06] <seb_kuzminsky> thanks dgarr
[21:26:07] <micges> seb_kuzminsky: halui is a real mess :|
[21:28:02] <seb_kuzminsky> micges: yeah... if it didn't successfully make awesome parts for tons of users i wouldn't like it at all ;-)
[21:28:31] <micges> hehe
[21:30:20] <seb_kuzminsky> dgarr: what you pushed (653a4c54) is identical to the thing you pushed before (0b349088), but it gets there by a different path (fern's commit, then the diff to make it equal to your original commit)
[21:30:32] <seb_kuzminsky> i think that means it subtracted out fern's changes
[22:05:21] <jepler> > No multiply instruction. Less RAM than an Apollo Guidance Computer. Lame.
[22:25:51] <linuxcnc-build> build #1504 of 4017.deb-wheezy-amd64 is complete: Failure [4failed shell_1] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-amd64/builds/1504 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[22:25:51] <linuxcnc-build> build #1502 of 4016.deb-wheezy-i386 is complete: Failure [4failed shell_1] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4016.deb-wheezy-i386/builds/1502 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[22:51:48] <KGB-linuxcnc> 03Dewey Garrett 05features_preview_2 0b34908 06linuxcnc 10(17 files in 11 dirs) features_preview: relocate placeholders, +cleanups * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0b34908
[22:51:48] <KGB-linuxcnc> 03Sebastian Kuzminsky 05features_preview_2 600192f 06linuxcnc 10configs/sim/gmoccapy/features_demo/gmoccapy.ini 10lib/python/gladevcp/features.py 10share/features/catalogs/mill/defaults.ngc 10share/features/features.glade Fern's changes from cac0ffd8, I hope * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=600192f
[23:00:08] <seb_kuzminsky> i feel like such a jerk trying help people use git: "you have to understand the internal data model! use it daily for a couple of years and it'll all make sense!"