#linuxcnc-devel | Logs for 2014-02-25

[06:50:10] <automata> is there any plan to make axis acceleration variable at run time? (maybe through HAL pins)
[06:53:25] <skunkworks> ooh - acceleration override!
[06:53:26] <skunkworks> ;)
[06:54:01] <automata> so is there any plans?
[06:54:14] <automata> or have I hit a raw nerve?
[06:55:33] <skunkworks> heh - I don't know - you might want to ask Rob - maybe on the devel email list..
[07:00:30] <automata> yup.. i figure since rob is already neck deep into this stuff.. maybe he can quickly give us pros and cons of doing this...
[07:00:45] <automata> and suggest a path to get there...
[07:22:29] <emcPT|2> Since yestarday I am trying to work out where self.tools[pocket] present in line 150 of lib/python/rs274/interpret.py
[07:22:45] <emcPT|2> is "constructed".
[07:23:37] <emcPT|2> It fetches the status, I presume from what I have been told, but I have no clear ideia where the table is constructed.
[07:28:47] <emcPT|2> on lib/python/rs274/interpret.py it is present self.tools = list(s.tool_table)
[07:29:13] <emcPT|2> so it seams that the source would be s.tool_table
[07:35:31] <emcPT|2> and still I do not understand where the tool_table gains it structure from
[07:39:47] <emcPT|2> maybe it is from io.tool.toolTable
[07:44:27] <emcPT|2> ?
[08:54:03] <dgarr> jepler Any ideas about http://www.panix.com/~dgarrett/stuff/fd2_connection.txt
[09:39:15] <emcPT|2> Hello Kwallance. Not sure if you can continue to provide (from yestarday) some help, but it would be appreciated
[09:40:10] <emcPT|2> in lib/python/rs274/interpret.py, the tool list is loaded though self.tools = list(s.tool_table)
[09:41:38] <emcPT|2> still I do not figure out from where the s.too_table gains it struture from
[09:42:37] <emcPT|2> on the C part of things I have all built with another column in the tool table (and I can access it in several parts of the code), but on the python code no
[09:43:15] <emcPT|2> It seams that I am missing some part of the code where the status channel creates the tool list
[10:33:34] <emcPT|2> where is the python interface, that provides import linuxcnc, located? (Dumb question maybe...)
[10:38:22] <dgarr> import linuxcnc is from: src/emc/usr_intf/axis/extensions/emcmodule.cc
[11:10:17] <emcPT|2> Nice. I did get it going... it was quite hard to figure out where the issue was. At the end was just changing a fixed value from 14 to 15 in the function static PyStructSequence_Desc tool_result_desc
[11:22:18] <KGB-linuxcnc> 03Dewey Garrett 05master 5f22663 06linuxcnc 10scripts/sim_pin sim_pin: accomodate longer pin names * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5f22663
[11:22:18] <KGB-linuxcnc> 03Dewey Garrett 05master 4f677a9 06linuxcnc 10scripts/sim_pin sim_pin: try to use sig if pin connected already * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4f677a9
[11:45:27] <seb_kuzm1nsky> folks are asking for debs of Rob Ellenberg's TP branches...
[11:45:40] <seb_kuzm1nsky> maybe it's time to give him push access and ask him to do his work in feature branches on glo?
[12:03:04] <cradek> seb_kuzminsky: I am pretty sure I offered him push access and didn't hear back. If you want to try again, by all means do.
[13:30:45] <skunkworks> I emailed rob. He mentioned that he wanted to get it to build against 2.5.3
[13:41:33] <pcw_home> does that mean there will be a 2.5.4 release?
[13:43:34] <cradek> yes there will be a 2.5.4 release, but it will not contain rob's new planner
[13:46:37] <cradek> if he rebases his changes onto 2.5_branch it'll be easily usable by people who are using 2.5, but it won't be in an official point release
[13:47:54] <pcw_home> 2.6?
[13:48:13] <cradek> that's a good question
[13:48:38] <cradek> I'd love it if the TPs were pluggable, and I had intended to work on doing that, but I haven't
[13:49:05] <cradek> I am not sure if rob has worked on that
[13:49:34] <cradek> since (if I understand right) he has a fallback mode to the old planner, maybe it is easy to make that configurable even without full pluggability
[13:50:01] <cradek> I'm not really in the loop here
[13:51:35] <pcw_home> Yeah but it looks really impressive (a big thanks to Rob for the TP work and Skunkworks for banging on it)
[13:52:00] <cradek> yes I appreciate them both very much
[13:53:49] <seb_kuzminsky> i asked rob to keep the new tp out of 2.6
[13:54:00] <seb_kuzminsky> i'm overwhelmed enough as it is with the stuff that's supposed to go into 2.6
[13:55:51] <cradek> I sure sympathize, but that makes me sad
[13:57:51] <cradek> the 2.6 situation is pretty demoralizing
[14:06:38] <PCW> it it merging the UBC stuff that's a problem?
[14:07:34] <cradek> yes that is the main obstacle
[14:07:56] <PCW> arghh
[14:10:59] <Connor> I'll ask again.. Anyone know anything about emc.halio ? Trying to get camview working on 2.6 I have it working, except for the labels for the X Y and Z corods..
[14:16:40] <cradek> PCW: looking at http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Todo-2.6 -- what is the status of getting the hm2 packages from the buildbot up to date?
[14:17:14] <seb_kuzminsky> Connor: sorry, i dont know anything about that
[14:17:52] <Connor> psha ever login anymore ?
[14:18:06] <seb_kuzminsky> sometimes, yeah
[14:18:17] <Connor> It's his code..
[14:18:21] <Connor> camview that is.
[14:31:21] <cradek> Connor: I'm not promising I can help, and I doubt anyone else can promise, but there's a possibility that if you share a lot more about what exactly you've done and what specific problem you have, someone can help. As is, I have no idea what I might be able to do to answer your question, because I don't really know what it is.
[14:31:55] <cradek> all I know is you've done *something* and some labels (?) are wrong
[14:32:31] <Connor> okay. one sec.. let me find the code.. brb
[14:32:54] <cradek> maybe share what you've done on a public git or (if more appropriate) a pastebin
[14:34:33] <skunkworks> cradek, Do you remember what emc1 used for TP? I remember you showing a picture of it following a path oddly
[14:35:26] <skunkworks> like cutting a segment out entirely
[14:37:06] <cradek> yeah, its blending would sometimes get overzealous and you'd get a fairly mangled path
[14:39:03] <Connor> run this from command line.
[14:39:07] <Connor> camview-emc -C camviewcfg.bak -g togle-caa.ui -H campins.hal
[14:39:15] <Connor> parse error?
[14:39:16] <Connor> Unable to instantiate [emc.halio]
[14:39:49] <Connor> http://pastebin.ca/2646420
[14:40:00] <Connor> the camviewcfg.bak file.
[14:40:09] <Connor> <unit id="emc.halio" width="640" height="480" pixelformat="CAM_PIXEL_FORMAT_RGB" format_name="640x480 RGB 24bpp">
[14:40:16] <Connor> is the problem..
[14:40:30] <Connor> I think maybe I'm missing a plugin?
[14:41:27] <cradek> does this work with modern (post-rebrand) 2.5 branch?
[14:41:40] <Connor> I was in the process of checking.
[14:42:20] <Connor> let me spin up the vm with 10.04 and 2..3
[14:42:22] <Connor> 2.5.3
[14:44:18] <cradek> skunkworks: I think this might be a test case that made emc1 (or pre-tp-rewrite emc2) misbehave: http://timeguy.com/cradek-files/emc/nasty-blend.png
[14:44:47] <dgarr> Connor: there was very long thread on camview on the user list in nov13 (maybe other threads too): http://thread.gmane.org/gmane.linux.distributions.emc.user/47679
[14:45:33] <cradek> skunkworks: and then fixed: http://timeguy.com/cradek-files/emc/whee.png and http://timeguy.com/cradek-files/emc/whee2.png
[14:48:48] <cradek> skunkworks: these are from Mar '06
[14:55:31] <skunkworks> wow
[14:57:52] <Connor> cradek: Honestly, I can't say that it does..
[14:57:58] <skunkworks> cradek, http://imagebin.org/295789
[14:58:09] <Connor> let me look through this thread on camview
[14:59:37] <cradek> skunkworks: those ends on the upper right look a lot like nasty-blend.png
[15:00:02] <skunkworks> that is what made me think of it.
[15:00:11] <cradek> Connor: if camview is pre-rebrand you will probably have to update it
[15:00:39] <Connor> Yea.. I'm thinking it's some sort of add on or something.. still looking..
[16:13:23] <PCW> cradek: probably needs a day or so work with someone familiar with how the build scripts work
[16:14:39] <PCW> I have some batch files that include new build options for SP6 based cards (with fallback config and all that)
[16:16:28] <PCW> item 18.3, 7i90 flash via mesaflash is fixed according to micges
[16:16:44] <PCW> (flash over EPP port)
[16:31:59] <seb_kuzminsky> we should put the mesaflash debs in our debian archive so they're easy for our users to get
[16:32:21] <cradek> yes yes
[16:33:25] <PCW> sure (I think there's some generally useful HM2 stuff embedded in mesaflash as well)
[16:40:38] <PCW> maybe the first step for hostmot2 buildbot updating would be to get one of the older card bitfiles (say 5I20) to build with the current source
[16:41:09] <seb_kuzminsky> that might be as simple as pushing the new source to the hm2 git repo
[16:41:18] <PCW> (and the attendant script massaging needed)
[16:42:08] <PCW> (include lists, new pin and module types for the xml files etc)
[16:45:16] <PCW> I have been trying to make the top level files have @Card@ and @Pin@ as the selected options so they dont need to be edited
[17:27:37] <skunkworks> logger[mah]_:
[17:27:37] <logger[mah]_> skunkworks: Log stored at http://linuxcnc.mah.priv.at/irc/%23linuxcnc-devel/2014-02-25.html
[17:33:03] <Tom_itx> PCW how's that goin?