#linuxcnc-devel Logs

Dec 13 2017

08:23 AM george_: Hi guys!!!
08:25 AM george_: Any suggestion for a micro camera like endoscope,knowing that works on LinuxCnc?
09:26 AM max12345: heyo, I just looked at the code-notes, there is a disclaimer at the start that it's not accurate. How not accurate is it? :)
09:27 AM archivist: measure
09:30 AM max12345: so you don't know?
09:33 AM archivist: when I want to know something I measure
09:33 AM max12345: but you didn't want to know, so you don't know, but you can give me the advice to go measure myself?
09:35 AM cradek: do you mean the range cap should be -pi/2 to +pi/2 instead of -2pi to +2pi?
09:35 AM cradek: it would be good for your bug report if you gave input for which you think the output is wrong
09:40 AM cradek: other nearby functions use +-2pi, still others use 0..pi
09:40 AM cradek: I don't know enough to evaluate this, if you do please write a fuller description of the problem you see so it's clearer
09:53 AM cradek: max12345: wait, are you lethang12cdt? if not, I've been talking to the wrong person
09:56 AM max12345: cradek: no I am not. did you talk to me? :)
10:00 AM max12345: But I don't know what he means with that bugreport either.
10:01 AM max12345: that said if p is in radians, the conditions don't make much sense.
10:02 AM max12345: assuming PM_PI is pi
10:02 AM cradek: yeah, it certainly is radians
10:03 AM max12345: what is the function supposed to do? extract the rotation data from a matrix?
10:05 AM cradek: rpy->p will always be in -pi..pi because atan2() so none of those will ever even fire
10:06 AM cradek: so it has a lot of unnecessary code, but otherwise no obvious bug that I see
10:06 AM cradek: so I'm even more puzzled about the bug report
10:51 AM max12345: hm, I'm trying to read the code, are there maybe some other guids than the one on the website? on blogs maybe?
10:54 AM cradek: sorry, probably not
11:05 AM max12345: I guessed as much. I'm going to take notes trying to understand it, maybe I can help update the code notes later?
11:06 AM max12345: the makefile is generated, right?
11:09 AM cradek: do you mean src/Makefile?
11:09 AM max12345: ye
11:09 AM cradek: that file is in revision control and maintained by humans
11:10 AM cradek: the files rewritten by configure come from the files with the .in extension
11:10 AM max12345: that's a pretty long makefile then...
11:11 AM cradek: it's a normal size for a big complex program
11:16 AM max12345: if you say so. It's not easy to read anyway :P
11:17 AM cradek: yeah they never are
11:17 AM cradek: makefiles are not exactly write-only but close
11:18 AM max12345: so to find out how it all ties together my best hope is look at the includes and go from there?
11:20 AM cradek: I don't know how to learn generally how a code base works by looking at it. I wish I had that skill. I have been able to make improvements to lots of projects without understanding the whole big picture, though, which is how I think most people handle it, for better or for worse.
11:21 AM archivist: I think many cherry pick a small part of a project to work on
11:21 AM max12345: that's fair. Most of the time I imagine people fix the bug they find and need fixed and move on :D
11:21 AM cradek: yes, that's a pragmatic way to accomplish a goal
11:22 AM cradek: yeah exactly
11:22 AM cradek: and some of us repeat that many times :-)
11:27 AM cradek: a nice wire edm would sure save me trouble with a part I need to make
11:31 AM archivist: what part?
11:32 AM cradek: the finger half of a stop work, square hole
11:33 AM * archivist points at a square file
11:33 AM cradek: yeah yeah
11:33 AM archivist: :)
11:33 AM cradek: the hole is the easy part
11:33 AM cradek: getting the outside shape nice seems harder
11:35 AM archivist: I have done parts manually turning that part of a radius I can, then file
11:45 AM cradek: archivist: http://timeguy.com/plot.png
11:46 AM cradek: instantaneous rate over a full run of an elgin bw raymond 21 jewel, dial up, freshly serviced
11:46 AM archivist: odd I get The connection to the server was reset while the page was loading
11:46 AM cradek: the main wiggle (black line) is at the rate of center wheel pinion teeth
11:47 AM cradek: huh it's just a (biggish) png
11:47 AM cradek: this included the heat going on and off in the room - the temperature compensation is really good
11:48 AM rene-dev: max12345 anything in particulat you are looking for?
11:48 AM archivist: been getting odd connection errors today, I can get the root site but not the png
11:49 AM cradek: http://www.letmecheck.it/mtu-test.php
11:51 AM archivist: from me The maximum MTU size for 81.149.189.98 is: 1500
11:51 AM max12345: rene-dev: not really.
11:51 AM cradek: that's not the problem then
11:52 AM cradek: huh
11:52 AM rene-dev: http://linuxcnc.org/docs/html/code/code-notes.html
11:52 AM max12345: I 'inherited' responsibility for a CNC in the basement of my student association / workshop. The previous computer died and I'm setting it up again, figured I could dive into stuff a bit.
11:53 AM max12345: Yep, I saw the code-notes, thanks. :)
11:56 AM max12345: Six months ago I learned my way into controlling a stepper motor with my raspi. I just want to see how the real deal works and what goes into it.
01:30 PM skunkworks__: you know what halscope needs? Channel invert!
03:54 PM Roguish: halscope could also use an FFT and a BODE plot
03:55 PM andypugh: seb_kuzminsky: Idn’t you know about Todd and his PCL720 cards?
03:57 PM andypugh: Here’s the driver: https://github.com/LinuxCNC/linuxcnc/blob/master/src/hal/drivers/pcl720.comp
03:57 PM andypugh: You must do, you have committed shanges to that driver :-)
08:37 PM -!- #linuxcnc-devel mode set to +vv by card.freenode.net
09:09 PM Tom_itx is now known as Tom_L
10:33 PM dgarr: seb_kuzminsky: the buildbot may need some attention:Start Mon Dec 11 23:09:41 2017/End Not Finished/Elapsed 45 hrs, 10 mins, 16 secs