#linuxcnc-devel Logs

Jan 25 2022

03:30 AM linuxcnc-build: build #3493 of 1630.rip-stretch-rtpreempt-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1630.rip-stretch-rtpreempt-amd64/builds/3493 blamelist: Sebastian Kuzminsky <seb@highlab.com>
04:48 AM pere: I just updated <URL: https://github.com/LinuxCNC/linuxcnc/pull/1376 >. Is there any interest in the approach?
04:54 AM linuxcnc-build: build #1537 of 4041.deb-buster-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4041.deb-buster-rtpreempt-amd64/builds/1537 blamelist: Dewey Garrett <dgarrett@panix.com>
05:29 AM linuxcnc-build: build #6639 of 1403.rip-wheezy-amd64 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1403.rip-wheezy-amd64/builds/6639 blamelist: Sebastian Kuzminsky <seb@highlab.com>
05:31 AM linuxcnc-build: build #6638 of 1400.rip-wheezy-i386 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1400.rip-wheezy-i386/builds/6638 blamelist: Sebastian Kuzminsky <seb@highlab.com>
05:43 AM linuxcnc-build: build #6161 of 1402.rip-wheezy-rtpreempt-i386 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1402.rip-wheezy-rtpreempt-i386/builds/6161 blamelist: Sebastian Kuzminsky <seb@highlab.com>
05:56 AM linuxcnc-build: build #6844 of 1404.rip-wheezy-rtpreempt-amd64 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1404.rip-wheezy-rtpreempt-amd64/builds/6844 blamelist: Sebastian Kuzminsky <seb@highlab.com>
05:57 AM linuxcnc-build: build #6314 of 1401.rip-wheezy-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1401.rip-wheezy-rtai-i386/builds/6314 blamelist: Sebastian Kuzminsky <seb@highlab.com>
06:33 AM linuxcnc-build: build #8504 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/8504 blamelist: Sebastian Kuzminsky <seb@highlab.com>
06:43 AM smoe: pere : The transition of man pages to asciidoc I think everyone consider beneficial. It is just a matter of synchronisation with the tool itself. If there was any "help2man" magic involved at some point, then this should be documented somewhere - and maybe the man page should then even be autocreated.
06:44 AM pere: why would sync with the tool it self be involved in such migration?
07:00 AM smoe: Well, if the tool changes, the man page needs to change with it. There is a chance that the man page was autogenerated from the --help output of the command with help2man.
07:03 AM smoe: You are right that the decision to store everything as asciidoc should not be affected by the question how much the man pages are in sync with the tools or not. But if you are touching this process then you may want to have a look.
08:01 AM pere: nope, I do not want that. :) I must admit, I try to keep my change proposals limited and not expand all over the place, to let them have a chance of getting accepted. :)
09:07 AM smoe: I was looking with my 3D-printing friend for drivers for my mill's stepper motors. The max 32V are possibly a bit low but https://www.duet3d.com/Duet3Mainboard6HC would cover that. I would just need to bring one of my many RPi back to life for SPI. Now, somewhat disturbing - this board already comes with a G-code interface that I would then somehow
09:07 AM smoe: want to bypass to keep using LinuxCNC for its overall milling-awareness. Has anybody done this yet?
01:07 PM seb_kuzminsky: smoe: i did something like that with my 3d printer
01:07 PM seb_kuzminsky: oops he left
01:08 PM seb_kuzminsky: it was originally driven by an arduino + RAMPS, replaced the arduino with a PC + mesa card, and wired that new controller into the RAMPS board to drive the existing stepper drivers & power transistors for the heaters & fans etc
01:09 PM seb_kuzminsky: i use a Teensy for analog input, https://github.com/SebKuzminsky/linuxcnc-hal-teensy
01:09 PM seb_kuzminsky: not saying this is the best setup, but it's enough to melt plastic and make bad smells, and occasionally parts ;-)
01:10 PM Tom_L: what more could one ask for!
01:41 PM pere: silopolis[m]: around?
02:17 PM pere: seb_kuzminsky: what is your view on <URL: https://github.com/LinuxCNC/linuxcnc/pull/1470 >?
02:21 PM seb_kuzminsky: pere: i'm in favor of the docs being buildable on older systems
02:21 PM seb_kuzminsky: and i'm certainly in favor of the more explicit image filenames
02:22 PM seb_kuzminsky: so that looks like an all-around win to me
02:22 PM seb_kuzminsky: do you know of any downsides?
02:22 PM pere: seb_kuzminsky: no. I assumed it was a reason it was done the original way, but I have no idea what it could be.
02:24 PM smoe: You have a point with the backports.
02:25 PM seb_kuzminsky: the version currently in master looks ok to me, what am i missing here? http://buildbot.linuxcnc.org/doc/master/v2.9.0-pre0-5622-g610c37379/html/ladder/classic-ladder.html#_data_response
02:25 PM pere: spaces?
02:26 PM seb_kuzminsky: that's built on buster with asciidoc 8.6.10-3
02:26 PM smoe: From how I understand this then this will fail once the .po files want to be applied for all the many languages.
02:27 PM seb_kuzminsky: pere: i'm sorry, i'm still not getting it. what spaces am i missing?
02:27 PM pere: seb_kuzminsky: not sure how your comments seem irrelevant to the problem I tried to solve, so I suspect I failed to explain it.
02:27 PM pere: the patch is to get po4a working for the future.
02:27 PM smoe: The po4a routines apparently fail to accept table rows that are empty in the last field.
02:28 PM smoe: So pere had introduced a blank.
02:28 PM seb_kuzminsky: oh thanks, that unconfused me - i was thinking you were working around an asciidoc problem, but you're working around a po4a problem
02:28 PM seb_kuzminsky: i get it now, thank you
02:28 PM seb_kuzminsky: and it's an issue we don't see currently because we're not using po4a currently
02:29 PM pere: exactly.
02:29 PM * seb_kuzminsky <-- a bit slow sometimes
02:30 PM smoe: What is slow is me getting my LinuxCNC conversion together. Every week brings new ideas. The latest one is to use https://www.duet3d.com/ who apparently reach into the CNC world with their 3D printing background.
02:30 PM seb_kuzminsky: i'm going to fetch that branch and push it to the main linuxcnc git repo, so that the buildbot will build the docs. if they look ok using the current pre-po4a docs build system i'll merge it. sound good?
02:31 PM smoe: Have you both chatted about pere 's change to the build system to get po4a running?
02:31 PM smoe: This is not part of this patch, but since you are here together, I thought you may want to get synced.
02:31 PM seb_kuzminsky: yeah, good idea
02:31 PM pere: nope. Just updated it. my po4a-build branch.
02:32 PM seb_kuzminsky: what's the status of the po4a-build branch now?
02:32 PM pere: did not create a pull request for it yet, as I am waiting for fr.po from silopolis[m] first. and it do not remove the old translated adoc files.
02:33 PM smoe: Hm. But _fr and _es can go, not?
02:33 PM pere: ?
02:34 PM smoe: I mean, we made quite an investment to get the original English and Spanish/French files synced and generate .po files from this syncing. Then po4a should happily regenerate these files using the English as a template.
02:35 PM smoe: The first few weeks will be a bit bad since we need to untranslate all the << >>> references and [[ ]] anchors, but otherwise this should be just fine?
02:35 PM pere: happily, yes, but someone need to review the strings to unfuzzy them. we could of course automatically unfuzzy, but quality would raise with review.
02:36 PM smoe: but that woudl happen within weblate, not?
02:36 PM pere: yes
02:36 PM smoe: So, then, fine. Whoever wants the original files - they are in github.
02:38 PM smoe: The only problem I see is that we will have references that will not exist since we are mixing languages and not everything is translated, so we may have a <<rue>> and only a [[street]] defined.
02:38 PM smoe: But that will be ironed out within two weeks.
02:38 PM pere: note, as build time tripled, I suspect there is something wrong with my make modifications causing repeated rebuilds of something, but the end result look good.
02:39 PM smoe: You just have many more files to run this on since _es _and _fr have never been complete.
02:39 PM pere: could be. did not spend much time debugging it, had other fishes to fry. :)
02:40 PM smoe: What have you done to the build system. I hope you say that it is no longer run from within the $root/src directory.
02:40 PM pere: only adjusted to get the PO based docs working.
02:40 PM smoe: Hm.
02:41 PM pere: as I said, I try to keep my changes small and focused.
02:41 PM smoe: And why exactly are we then still waiting for silopolis[m]?
02:41 PM smoe: He can introduce something to the English bits at any time, can't he? The synchronisation work was otherwise completed.
02:41 PM pere: I have not seen a message from him stating he is done.
02:42 PM smoe: Hm. I do not think that he needs to be done.
02:43 PM smoe: He just adds the missing English texts at some later point to the documentation.
02:44 PM smoe: Fine - we would lose two or three files that could not be matched, but that could be fixed in other ways, still.
02:44 PM pere: well, I want to hear it from him that I should stop waiting.
02:44 PM smoe: I PM him.
02:52 PM pere: hm, rebasing po4a-build on master did not build this time.
02:53 PM pere: I build using 'debuild', if someone want to have a look.