#linuxcnc-devel | Logs for 2013-11-16

[05:20:50] <norbert> Halo, I have written some new custom glade widgets, I would like to share. Now as I have tested them and find them usefull, I would like to implement also the documentation to linuxcnc-dev(docs/src/gui/gladevcp.txt. I just copied one section and modified it to my needs. Now my two questions: Is it possible to "compile just the one file to see how it looks later in HTML? and as I do not understand very good the formating code used in tha
[05:20:51] <norbert> t file, where can I find documentation to that? I hope that my questions are not to stupid, I so, please just ignore me. Regards Norbert
[05:36:08] <rob_h> hi norbert might want to ask later on when some of the Devs are about they can help you then.
[05:38:21] <norbert> Hope they do read than my message, or I will post it later again. Thanks
[07:46:40] <mhaberler> hallo borbert
[07:46:44] <mhaberler> norbert, sorry
[07:46:58] <mhaberler> gehts nur ums document formatieren?
[07:48:24] <mhaberler> mache folgendes: ./confgure --enable-build-documentation=html ; lass das einmal laufen (braucht länger); beim nächsten make wird nur das veränderte dokument formatiert und die links gechecked
[07:52:06] <norbert> Hallo Michael, das habe ich schon gemacht, aber dann funktionieren interne Links nicht mehr, ich denke, da mache ich was in der Formatierung falsch.
[07:52:39] <mhaberler> ja, das kann nur passieren wenn Deine Links falsche syntax haben oder ins Nichts zeigen
[07:52:56] <mhaberler> schau Dir in anderen Dokumenten an wie das mit den links geht
[07:53:19] <norbert> Genau und dazu suche ich eine DOKU, um nachzuschlagen was ich falsch mache, z.B. wie fgt man einen Zeilenumbruch ein ?
[07:53:24] <mhaberler> veränderst du das gladevcp dokument oder machst einen neuen abschnitt?
[07:53:34] <mhaberler> asciidoc manual
[07:53:37] <norbert> Ich verndere das Dokument
[07:53:43] <mhaberler> ah ok
[07:53:55] <norbert> AHA ich schaue mal in die Man rein, danke
[07:54:39] <mhaberler> minimalguide: http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=blob;f=docs/asciidoc-markup.txt;h=8eda65cc733fddf4018460520af83d4ce605b3f7;hb=739df958aca9d246daad36f439c82bfbeac681b9
[07:55:38] <mhaberler> achtung - asciidoc hat siche rasch verändert und linuxcnc verwendet eine ältere versions; insbes bei captions etc musst aufpassen
[07:56:01] <norbert> Werde mir Mhe geben ;-)
[07:57:14] <mhaberler> das ist ein bissl steil am anfang aber das wird schon
[11:17:38] <ju-emb> I know, I asked that before here but I forgot:
[11:17:40] <ju-emb> how do I know from inside a python script where the actual used config files reside
[11:17:42] <ju-emb> I like to loadusr a python script and access files that I put in the same directory as the config files are
[11:47:35] <kwallace2> This may be off base but maybe try "import os" and "lcnc_base_dir = os.environ['EMC2_HOME']"
[11:50:22] <kwallace2> There may be other EMC2_XXX keywords. "print os.environ" might find them.
[11:51:26] <kwallace2> ju-emb: ^^
[12:08:20] <ju-emb> yes, I'm playing around with that right now,
[12:08:22] <ju-emb> I'm absolute python noob and right now still fighting with it
[12:10:27] <ju-emb> who's a python geek here?
[12:13:42] <kwallace2> There should be plenty except it's Saturday.
[12:14:51] <kwallace2> Did you try putting a 'print os.environ' in your script?
[12:15:54] <ju-emb> > it's Saturday that's right, sadly not for me
[12:15:56] <ju-emb> so, I don't let google have it's Saturday too
[12:18:06] <ju-emb> Yes I tried print os.environ, but it seems that are specific variables coming out of the .ini
[12:18:54] <ju-emb> there is an example also to parse the ini with import linuxcnc,
[12:20:06] <ju-emb> so I try to go the way using cmd-line parameters
[12:20:58] <kwallace2> I would think EMC2_HOME plus '/configs' should put you in the config directory.
[12:21:49] <ju-emb> but python is tough or I'm to stupid to make python accept the same parameter several times
[12:22:17] <ju-emb> yes, EMC2_HOME plus '/configs works
[12:22:22] <kwallace2> Oops, that's right you want the particular file active .ini name.
[12:22:43] <ju-emb> that's the point
[12:25:04] <ju-emb> OK, for my tests right now I give the dir as parameter,
[12:25:06] <ju-emb> till Monday it will work that way
[12:26:01] <ju-emb> in C you can give a parameter several times like
[12:26:03] <ju-emb> programname -d -d -d
[12:26:35] <kwallace2> http://www.tutorialspoint.com/python/python_command_line_arguments.htm
[12:26:42] <kwallace2> Maybe.
[12:27:11] <ju-emb> python seems to not accept that
[12:27:24] <ju-emb> this page I visited
[12:28:29] <ju-emb> kwallace2: you also doesnt have Saturday?
[12:29:49] <kwallace2> I'm a freelancer so I work all the time or whenever I can.
[12:36:50] <ju-emb> that's the package and the good for freelancers,
[12:36:52] <ju-emb> I share that with you
[12:42:32] <kwallace2> http://www.linuxcnc.org/docs/html/common/python-interface.html#_reading_ini_file_values_a_id_sec_python_reading_ini_values_a
[12:45:48] <ju-emb> that page I had before, it leads me to the idea using program args
[12:46:38] <kwallace2> Does it work?
[12:47:50] <kwallace2> print "--> Active INI =", linuxcnc.ini(sys.argv[1])
[12:50:13] <ju-emb> in general yes,
[12:50:15] <ju-emb> I try to figure out, how I can use program parameters in python and how can I parse arguments like
[12:50:17] <ju-emb> programname -d<dir> 1=file1 2=file2 5=file5
[12:51:39] <mhaberler> see configparser
[12:51:46] <mhaberler> example in gladevcp.py
[12:53:19] <ju-emb> ahh OK, that's a good idea
[12:53:21] <ju-emb> not using program parameter but config file entries
[12:53:55] <mhaberler> you just talked about program parameters?
[12:54:06] <mhaberler> which I understand are command line args
[12:55:19] <ju-emb> that's rigth, I tried to use command line args
[12:55:53] <ju-emb> problem is: same arg several times
[12:56:29] <ju-emb> the c way:
[12:57:24] <ju-emb> programname -d<dir> -cfile1 -cfile2 -cfile5
[12:57:41] <ju-emb> or
[12:58:21] <ju-emb> programname -d<dir> --c1=file1 --c2=file2 --c5=file5
[12:58:34] <mhaberler> that is unusal
[12:58:50] <mhaberler> usual one distinguishes between args and options
[12:59:18] <mhaberler> like prog -opt1 -opt2=optval arg1 arg2 arg3
[13:00:00] <ju-emb> yes, that way it is ok
[13:00:33] <mhaberler> that is the standard way options and args are handled in unixland, and configparser is just fine for that
[13:00:53] <ju-emb> and
[13:00:55] <ju-emb> arg1 can be
[13:00:57] <ju-emb> 1=file1
[13:00:59] <ju-emb> arg2 2=file2 etc
[13:02:07] <ju-emb> Ok I'll have a look at configparser
[13:02:52] <mhaberler> ah sorry configparser is for inis. I mixed that up with optionparser, hold on
[13:03:04] <ju-emb> hal_parport has also a intersting construct: cfg="ffff fff f "
[13:03:36] <ju-emb> or hostmot2
[13:04:13] <ju-emb> I try to figure out how those guys have done that
[13:04:29] <mhaberler> just look at src/hal/user_comps/gladevcp.py and look at the OptionParser usage there
[13:04:52] <ju-emb> ok, I'll do that
[13:04:59] <ju-emb> thanks
[13:05:15] <mhaberler> re hal_parport: you dont want to know.. this is dictated by a several restricted environment, kernel modules, and not a great guide for general command line handling
[13:07:06] <ju-emb> thanks, for mentioning that
[13:08:09] <ju-emb> but first I have to lift the glucose level in my blood a bit
[13:08:58] <mhaberler> oh, just read optparse is deprecated, use argparse
[13:09:01] <mhaberler> http://docs.python.org/2/library/argparse.html#module-argparse
[13:09:30] <ju-emb> yes, argparse sounds interesting
[13:10:33] <mhaberler> this is a step-by-step tutorial on argparse: http://docs.python.org/2/howto/argparse.html
[13:12:28] <ju-emb> bbl
[13:12:30] <ju-emb> -> on the way -> kitchen
[13:13:39] <kwallace2> It seems to me since G38 can put successive probe locations in the PROBEOPEN file, there should be an option to add a tag to the file line on each G38 call.
[13:15:22] <skunkworks> kwallace2, you don't ask for much ;)
[13:18:36] <kwallace2> Just "G38.3 X 2.222 p west_face", then have the the file write add "west_face" to the beginning or end of the line.
[13:20:08] <kwallace2> That way I can pull west_face data from the file and display it in the GUI.
[13:22:19] <kwallace2> It looks like I need to molest emccanon.cc to do it.
[13:27:16] <mhaberler> please.. if you are dying to do this, do it without patching around in C
[13:27:27] <mhaberler> you can use Python / linuxcnc module
[13:27:46] <mhaberler> or a new G-code in Python which will jump any hoop
[13:28:16] <mhaberler> this is the kind of thing which embedded Python via a remapped code does perfectly
[13:29:24] <mhaberler> the remapping manual has lots of examples for inspiration, for instance the spiral example
[13:30:40] <kwallace2> Okay, I'll look at that.
[13:31:09] <mhaberler> 3.3.4. Advanced example: Remapped codes in pure Python
[13:31:24] <mhaberler> somewhere here: http://www.linuxcnc.org/docs/devel/html/remap/structure.html#_the_em_argspec_em_parameter_a_id_remap_argspec_parameter_a
[13:32:45] <mhaberler> when doing probes, you will need to use the 'yield' construct (dealing with queuebusters), see 9.4.5. Dealing with queuebuster: Probe, Toolchange and waiting for a HAL pin
[13:34:46] <mhaberler> the rest is normal Python file I/O
[15:41:28] <KGB-linuxcnc> 03Sebastian Kuzminsky 05v2.5_branch 2ba400b 06linuxcnc 10docs/src/code/Contributing-to-LinuxCNC.txt docs: give some examples of public git servers * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2ba400b
[17:14:03] <cradek> http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2ba400b
[17:23:52] -holmes.freenode.net:#linuxcnc-devel- [freenode-info] channel flooding and no channel staff around to help? Please check with freenode support: http://freenode.net/faq.shtml#gettinghelp
[19:16:36] <cradek> jepler: I'm getting AXIS segfault in ja4 branch; my attempt at a backtrace: http://pastebin.com/raw.php?i=FK6415mV
[19:17:24] <cradek> this is sim on x86_64, approximately ubuntu precise (actually mint)
[19:20:27] <cradek> commenting the three _tk_seticon.seticon calls in axis.py lets it start
[19:30:40] <KGB-linuxcnc> 03Chris Radek 05joints_axes4 3046dd3 06linuxcnc 10src/emc/motion/control.c Fix: teleop jog speed uncontrollable from gui * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=3046dd3
[19:30:40] <KGB-linuxcnc> 03Chris Radek 05joints_axes4 0b55aba 06linuxcnc 10configs/sim/axis/rdelta.ini rdelta: set more appropriate axis jog speed limit * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0b55aba
[19:52:52] <jepler> cradek: hmph. we tweaked that code around a couple different ways but ja4 seems to have the latest iteration
[19:53:41] <jepler> cradek: hmph. we tweaked that code around a couple different ways but ja4 seems to have the latest iteration
[19:53:44] <jepler> 88eec4be7005a5b1431562c4b3e185b914a05769
[19:54:10] <jepler> but that commit's in ja4, so...
[20:31:45] <KGB-linuxcnc> 03Jeff Epler 05master 7668d91 06linuxcnc 10src/emc/usr_intf/axis/scripts/axis.py Merge branch 'axis-no-seticon' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7668d91
[20:32:39] <jepler> cradek: commit 1402d8a should be suitable to merge into ja4 as well, though I think that seb_kuzminsky has asked us not to push merges to ja4
[20:32:50] <jepler> cradek: but for your integration testing it may be useful
[22:35:47] <seb_kuzminsky> i only asked for no merges in order to make rebases easier
[22:36:03] <seb_kuzminsky> i'd prefer to branch ja5 from ja4, then rebase ja5 on top of master
[22:36:10] <seb_kuzminsky> it
[22:36:23] <seb_kuzminsky> it's an experiment in alternative feature branch management
[22:36:47] <seb_kuzminsky> cradek: thanks for fixing that ja4 todo!
[22:43:47] <KGB-linuxcnc> 03Sebastian Kuzminsky 05v2.5_branch 6de6a0d 06linuxcnc 10docs/src/code/Contributing-to-LinuxCNC.txt docs: use imperative form consistently * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6de6a0d
[23:54:34] <KGB-linuxcnc> 03Sebastian Kuzminsky 05master 490e035 06linuxcnc 38 commits pushed, 104 files changed, 031112(+)