#linuxcnc-devel Logs

Oct 19 2017

#linuxcnc-devel Calendar

04:17 AM linuxcnc-build: build #125 of 4030.deb-stretch-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4030.deb-stretch-rtpreempt-i386/builds/125 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
04:23 AM linuxcnc-build: build #125 of 4031.deb-stretch-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4031.deb-stretch-rtpreempt-amd64/builds/125 blamelist: Chris Morley <chrisinnanaimo@hotmail.com>
04:44 AM -!- #linuxcnc-devel mode set to +v by ChanServ
07:21 AM trasz: folks, some time ago i've pushed some patches to make linuxcnc work on freebsd.
07:21 AM trasz: now, those patches were the 'clean' ones.
07:22 AM trasz: there's, however, still a number of patches that i'm not sure about.
07:22 AM trasz: is someone interested in taking a look, to see if there's anything left that could be upstreamed?
07:22 AM trasz: current list of out-of-tree diffs is here: https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/
08:02 AM jepler: trasz: in general I would be happy to see the build script copy and set the #!-line of all python and tcl/wish scripts, which would get rid of https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-scripts_hal-histogram?revision=418465&view=markup
08:03 AM jepler: we really should stop putting that so-called realtime script in /etc/rc.d, it's a bug on linux too https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-src_Makefile.inc.in?revision=418465&view=markup
08:04 AM jepler: weird, linux manpage says fdatasync is in _POSIX_C_SOURCE >= 199309L -- do we not have the enabling defines right or something? https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-src_emc_rs274ngc_rs274ngc_pre.cc?revision=446574&view=markup
08:05 AM jepler: I wonder why this is necessary. but I don't even know what panelui is. https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-src_hal_components_Submakefile?revision=418465&view=markup
08:06 AM jepler: it's obviously an oddity, since the source files are being compiled directly to an executable without making object files. I'd love to see the build of this .. thing .. fixed
08:07 AM jepler: er no it's converting its source files to objects in the normal way. so why is cflags necessary? weird.
08:08 AM jepler: you can't shim out the parport-using drivers by just making the hal_parport calls all fail? https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-src_hal_drivers_hal__ppmc.c?revision=418465&view=markup
08:08 AM jepler: we also have a spot to provide useless non-working inb/outb on linux arm64
08:08 AM jepler: maybe by hooking into that?
08:08 AM jepler: https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-src_hal_hal__lib.c?revision=418465&view=markup is this not otherwise assigned a value?
08:09 AM jepler: would not *reject* a patch to make "make" vs "gmake" configure-time detected: https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-src_hal_utils_halcompile.g?revision=418465&view=markup
08:11 AM jepler: would not reject a patch to make "md5" vs "md5sum" configure-time detected https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-tcl_tooledit.tcl?revision=418465&view=markup
08:12 AM jepler: reportedly, "tcllib" would provide hashes direct from Tcl, could also investigate using it instead
08:13 AM jepler: "comp" should get modes for printing the usual compiler flags, halcompile --print-cc, etc. could use it here https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-tests_build_header-sanity_test.sh?revision=418465&view=markup
08:13 AM jepler: and here https://svnweb.freebsd.org/ports/head/cad/linuxcnc-devel/files/patch-tests_build_ui_test.sh?revision=418465&view=markup
08:13 AM jepler: trasz: anyway there's my off the top of my head feedback on your patches
08:13 AM jepler: trasz: thanks for your contributions!
08:13 AM trasz: jepler: thanks! i'll go through it when i get home.
08:47 AM jepler: jepler@rat:~/src/linuxcnc/master$ git merge origin/2.7
08:47 AM jepler: Segmentation fault
08:48 AM jepler: well that doesn't usually happen
08:49 AM jepler: .. and it didn't repeat
08:53 AM jepler: andypugh: I merge 2.7 to master, so you can go ahead and make the right fix for hm2_7i43/7i90 in master branch without causing trouble
09:48 AM jepler: merge=merged
01:23 PM skunkworks: seb_kuzminsky, how did you do your code formatting on the bug report?
01:23 PM skunkworks: when I do a quote or a code insert - it takes the formatting out
01:23 PM skunkworks: hmm does <pre> work?
01:24 PM skunkworks: heh - if you code a quote it works
01:36 PM jepler: skunkworks: you can put code between "~~~~" lines
01:36 PM jepler: ~~~~
01:36 PM jepler: 10 PRINT HELLO
01:36 PM jepler: 20 GO TO 10
01:36 PM jepler: ~~~~
01:37 PM andypugh: Error: Variable not known “HELLO”
02:06 PM skunkworks: jepler, thanks
02:06 PM * skunkworks waits patiently for jmk to yell at me. sure do miss that
02:38 PM jepler: thanks andypugh
02:41 PM andypugh: How do we feel about multispindle into master?
02:42 PM andypugh: (needs a decision on G-code syntax. I like “$” as a newly-valid character. Which might be missing from CNC keypads, though)
02:42 PM andypugh: And it looks like Ben has given up on G71, so I should perhaps pick that up?
03:00 PM skunkworks: yes and yes. (IMHO)
03:08 PM pcw_mesa: you dont want to buy Patriks G71 add-on?
03:08 PM pcw_mesa: https://www.youtube.com/watch?v=E-P2qL3qkzk
03:08 PM andypugh: Oddly, no.
06:07 PM skunkworks: I really should get my github account setup with a forked linuxcnc repository so I can play with it
06:11 PM hazzy: skunkworks: Absolutely, it is very straight forward to do and makes it much easier to fiddle with things
06:18 PM skunkworks: I think I still have your repository up in one of my tabs.. :)
06:20 PM hazzy: that is not good, you will learn bad coding style just by my code sitting there
06:23 PM skunkworks: I have no style...
06:28 PM hazzy: heh
06:28 PM seb_kuzminsky: http://s2.quickmeme.com/img/cc/ccba76acea0ede1defa681067029430a8d58b1a58def275c34d18756b122daf1.jpg
09:27 PM jepler: that fix can't possibly be right, he said, not understanding what a right fix would be
10:12 PM skunksleep: jepler: I agree
10:15 PM skunksleep: I stared at it long enough to sort of figure out what is going on. Didn't look at the overshoot issue. The wasn't a problem in my application.