#linuxcnc-devel Logs

Jun 01 2017

#linuxcnc-devel Calendar

08:17 AM jepler: seb_kuzminsky: it's nice how things sometimes become clear after 5 years
11:21 AM seb_kuzminsky: i successfully resisted the urge to rewrite all the linuxcncrsh-based tests in python
11:23 AM -!- #linuxcnc-devel mode set to +v by ChanServ
11:29 AM KGB-linuxcnc: 03Sebastian Kuzminsky 05master 8c21a2d 06linuxcnc 10docs/src/config/ini-config.txt 10docs/src/gui/pyvcp.txt 10src/hal/drivers/mesa-hostmot2/hm2_eth.c Merge remote-tracking branch 'origin/2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=8c21a2d
02:51 PM linuxcnc-build: build #100 of 4025.deb-jessie-armhf is complete: Failure [4failed shell_1] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4025.deb-jessie-armhf/builds/100 blamelist: dummy, Dewey Garrett <dgarrett@panix.com>
02:53 PM seb_kuzminsky: linuxcnc-build: force build --branch=2.7 0000.checkin
02:53 PM linuxcnc-build: build #4967 forced
02:53 PM linuxcnc-build: I'll give a shout when the build finishes
04:17 PM linuxcnc-build: Hey! build 0000.checkin #4967 is complete: Success [3build successful]
04:17 PM linuxcnc-build: Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4967
05:26 PM seb_kuzminsky: the jessie armhf buildslave is back
05:26 PM seb_kuzminsky: 1: 0
05:26 PM seb_kuzminsky: 500 http://buildbot.linuxcnc.org/ jessie/2.7-sim armhf Packages
05:26 PM seb_kuzminsky: linuxcnc-build: build --branch=master 0000.checkin
05:31 PM seb_kuzminsky: linuxcnc-build: hello?
05:32 PM seb_kuzminsky: ffs
05:33 PM seb_kuzminsky: it's hearing me, it's just not doing anything:
05:33 PM seb_kuzminsky: 2017-06-01 15:56:36-0600 [IrcStatusBot,client] irc command build
05:33 PM seb_kuzminsky: 2017-06-01 16:01:45-0600 [IrcStatusBot,client] irc command hello?
05:33 PM seb_kuzminsky: oh, because i typoed it
05:33 PM seb_kuzminsky: linuxcnc-build: force build --branch=master 0000.checkin
05:33 PM linuxcnc-build: build forced [ETA 1h23m44s]
05:33 PM linuxcnc-build: I'll give a shout when the build finishes
05:34 PM seb_kuzminsky: that's more like it
05:52 PM andypugh: Great
05:53 PM andypugh: There seems to be interest. Not sure why, but there is
05:54 PM seb_kuzminsky: i think it helps that arm boards are physically small, come with processor and RAM preinstalled, and are inexpensive compared to x86
05:54 PM andypugh: There are some cool little borads out there, but in the context of 2 tons of cast iron, board size is not top of my list.
05:54 PM seb_kuzminsky: most people have .002 tons of aluminum extrusions
05:54 PM seb_kuzminsky: i think
05:55 PM andypugh: But I have started playing with glue-guns. I am not sure we have a “market” there, they just work.
05:55 PM seb_kuzminsky: might be we missed that boat
05:55 PM andypugh: And you won’t spend £250 retrofitting a £250 3D printer
05:55 PM seb_kuzminsky: well i did
05:55 PM seb_kuzminsky: but your point is well taken
05:56 PM andypugh: Luckily we don’t care about that boat :-) Fewer clueless newbies and a focus on what our machines do. I know That’s selfish. But.
05:57 PM seb_kuzminsky: i've got nothing against diy glue gun printers, i'm just disappointed how fragmented the arm hardware ecosystem is
05:58 PM seb_kuzminsky: it's a ton of work to deal with each new board
05:59 PM seb_kuzminsky: debian has made some progress lately on unifying the kernel for the arm boards, but i haven't been following it closely
06:00 PM andypugh: I find myself wishing we could get Yocto on board. Richar Purdie is a friend of mine, and he does own a millling machine he has talked of CNC-ing
06:01 PM andypugh: But he is pulled so many ways that I am not sure he gets to live a life of his choosing.
06:02 PM seb_kuzminsky: is yocto anything like openwrt or gentoo?
06:03 PM seb_kuzminsky: their website looks more like politics than technology
06:03 PM seb_kuzminsky: or more like marketing, maybe
06:04 PM andypugh: Not as I undersand it (dimly). Yocto is a build system than makes hardware-specific builds. It does have it’s own distro (Poky Linux) but it is more aboue the cross-compilers as I understand it. I think it is abour hardware that is too painful to native-compile on.
06:05 PM andypugh: This one? https://www.yoctoproject.org
06:06 PM seb_kuzminsky: i worked on a project using OpenWRT recently, and i liked how it cross-built everything from source, and how the resulting image ran in a robust read-only mounted way
06:06 PM seb_kuzminsky: it was a nice way to build a linux appliance
06:08 PM seb_kuzminsky: but the openwrt project blew up and forked and fell apart
06:09 PM andypugh: RP is no politician. He’s just our average geek rather surprised to find himself a Linux Fellow. I know him though an interest in motorcycling, especially offroad.
06:10 PM seb_kuzminsky: sounds like an interesting fellow!
06:13 PM andypugh: Oh, he is. Which is why I keep nudging him towards CNC-ing his (I think) Bridgeport.
06:28 PM linuxcnc-build: build #3120 of 1405.rip-wheezy-armhf is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1405.rip-wheezy-armhf/builds/3120
06:28 PM linuxcnc-build: build #4968 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4968
06:42 PM seb_kuzminsky: - M66 remapping pre: 5399=0.000000; 100=0.000000; ain0=0.00; din0=0
06:42 PM seb_kuzminsky: + M66 remapping pre: 5399=0.000000; 100=0.000000; ain0=0.00; din0=1
06:42 PM seb_kuzminsky: i dont know what to do with that test failure...
06:44 PM andypugh: is that an actual digital in on a test-slave?
06:44 PM seb_kuzminsky: no, it's just a hal pin
06:44 PM seb_kuzminsky: no real hardware in the buildslave farm
06:45 PM andypugh: Maybe I phrased the question badly.
06:45 PM andypugh: It seems that that is either no problem at all, or really very serious.
06:46 PM andypugh: Unless you can imagine a middle-ground?
06:47 PM seb_kuzminsky: standard test flakiness would be my first guess
06:47 PM seb_kuzminsky: i think that test has run many times since the related code changed any
06:48 PM andypugh: Or it’s unrelated code writing to random memory?
08:23 PM jepler: just for "fun" I spun up a copy of forum and configured it to use php5-fpm and mpm_worker. It doesn't benchmark any faster (ab with concurrency of 1 and 4) than the current configuration. it might be a little easier on memory usage, though, which would mean that workers could be tuned a bit higher
08:23 PM jepler: (since the apache worker threads don't have to have php in them)
08:27 PM seb_kuzminsky: i've not had any failures to fetch https://forum.linuxcnc.org/forum/recent with a 25-second timeout for... days and days
08:28 PM seb_kuzminsky: if you want to try that config change i'll tune the timeout down and we can see how it does
08:29 PM seb_kuzminsky: or if you'd rather do something else, i'm content with this level of performance
08:29 PM jepler: nah
08:36 PM jepler: OK time to break another machine by upgrading to stretch..
08:37 PM jepler: 2166 upgraded, 637 newly installed, 107 to remove and 1 not upgraded.
08:38 PM jepler: .. why an upgrade needs to install over 600 new packages I have no clue
08:48 PM Tom_L: sounds like a windows install
09:04 PM jepler: except the computer has been usable this whole time
09:04 PM jepler: .. still scrolling messages from apt+dpkg merrily along
09:22 PM jepler: OK time to reboot and see what's broken ...
10:12 PM seb_kuzminsky: ... and he was never heard from again
11:00 PM Tom_L: hah
11:31 PM AndChat|144384 is now known as skunkworks_