#linuxcnc-devel Logs

Mar 21 2022

#linuxcnc-devel Calendar

04:01 AM -!- #linuxcnc-devel mode set to +v by ChanServ
04:50 AM Tom_itx is now known as Tom_L
04:04 PM pere: seb_kuzminsky silopolis[m] andypugh jepler[m]: I hope you got my email about a videochat on monday.
04:26 PM silopolis[m]: @pere received, but was sent in spam by gmail 🤔
04:26 PM silopolis[m]: * pere: received, but was sent in spam by gmail 🤔
04:27 PM silopolis[m]: Also 18pm UTC is too early for me as I finish work at 18h30 😕
04:28 PM silopolis[m]: Let me setup a framadate...
05:04 PM * pere is amazed someone still trust gmail to not hide their email from them. :)
05:05 PM pere: silopolis[m]: I am fine with a bit later, what about 19UTC?
05:17 PM silopolis[m]: Should be ok for me
06:39 PM yeltrow: I'm considering contributing two components, but I want to make sure I test them well and I was considering writing some sort of automated test for them. Should I write a bash script or something that uses halrun and halcmd and then checks results?
06:56 PM andypugh: yeltrow: Have you seem the LinuxCNC testsuite? Copy that structure and add yours to the list.
07:04 PM yeltrow: @andypugh Nope. I know nothing about it. I cloned the github repo. Is it in there?
07:09 PM yeltrow: Found this: http://linuxcnc.org/docs/html/code/building-linuxcnc.html Which is the basic fork/pull-request stuff.
07:30 PM andypugh: yeltrow: Look in linuxcnc/tests
07:31 PM andypugh: yeltrow: the command “runtests” will run them all, or you can add a path to a test to the command.
07:32 PM yeltrow: Thank you--Any really minimal examples you can think of off-hand that I could use for a .comp file?
07:32 PM yeltrow: Just making a test program would be stepping my game up significantly.
07:33 PM andypugh: linuxcnc-dev/tests/abs.0/
07:34 PM yeltrow: Also -- Can they require a gui? -- One of my components is meant to be used with the homing sequence and I see that even using DISPLAY = ngcgui brings up at least a splash screen.
07:35 PM yeltrow: They might be really slow because it is a legit test to see if the homing sequence ends where expected. If everyone wrote tests like that it would take 10 years to use runtests...
07:36 PM andypugh: The “lathe” test runs a home command. It has a Python program as the “gui”
07:42 PM yeltrow: That is kind of what I was thinking of doing -- I was going to run halui. A python program would be responsible for setting things up, checking the results, and deeming pass or fail. Motion and halui so that I get homing to work would be enough (trivkins??)
07:50 PM yeltrow: I will see if I can use the lathe example as a template for the component that interacts with homing.
07:51 PM andypugh: Once motion is loaded (needs an INI) then the “gui” just needs to run “c.home(-1) on the linuxcnc.command object.
07:56 PM yeltrow: Why won't these worms go...back...in...the...CAN!!! Um... "just needs to run c.home(-1) on the linuxcnc.command object." I don't even know where to look up where that came from :) I just got the python hal library doing a few useful things and can write some C code. I know diddly about the linuxcnc internals.
09:05 PM andypugh: yeltrow: https://linuxcnc.org/docs/2.8/html/config/python-interface.html
10:19 PM yeltrow: andypugh: Thank you. The last copy of that I looked at showed it hadn't been updated since 2016, so I was worried it might be deprecated. I see this page was updated yesterday according to the footer.
10:54 PM -!- #linuxcnc-devel mode set to +v by ChanServ