#linuxcnc-devel | Logs for 2015-08-24

[07:02:45] <skunkworks> cradek, the drives in jr are +/-10v - correct?
[07:03:50] <skunkworks> the manuals for the drives are odd. they say 'differential input adjustable range 4 to 10v'
[07:04:33] <skunkworks> http://www1.yaskawa.com/site/dmservo.nsf/f5c9b60441c8a80286256e61004f64b9/86256ec30065f21886256e99004f6aaa/$FILE/TSE-C717-13E.pdf
[07:07:40] <skunkworks> must be TG and then the tack goes into the 'speed reference..'
[07:54:54] <cradek> yes, they were normal to hook up, no problems.
[07:55:37] <cradek> they have jumpers or resistors or something for scaling
[07:55:45] <cradek> forget exactly
[08:11:37] <skunkworks> Do you remember having to adjust the scaling?
[08:41:16] <cradek> I think so, but two of the amps weren't original to the machine, and one I modified for simulated tach
[08:41:31] <cradek> so don't trust my memory to guide you
[08:41:35] <cradek> what are you working on?
[08:44:34] <skunkworks> getting things ready to convert the mattsura
[08:44:50] <skunkworks> it looks if the spindle drive keeps running - it will be pretty easy
[08:44:59] <skunkworks> it takes care of orient and such
[08:45:05] <cradek> neat
[10:14:48] <skunkworks> pcw_home, does the db style mesa cards ie 5i25 have issues driving the g540 watchdog?
[10:14:58] <skunkworks> that you know of?
[10:15:31] <skunkworks> why does the latest release still say 2.6.9? ;)
[10:31:13] <seb_kuzminsky> skunkworks: i got distracted trying to add a bunch of jessie buildslaves instead of being responsible and pushing 2.7.0 out the door this weekend :-/
[10:31:25] <seb_kuzminsky> then i fell into the libvirt/kvm configuration rabbit hole
[10:31:31] <skunkworks> heh
[10:32:08] <seb_kuzminsky> my current yak is trying to enable multicast routing between libvirt virtual networks and the hypervisor host's real networks :-/
[11:48:50] <dgarr> i see netcat listed in debian/control.in as a Build-Depends Is it used somehow by the buildbot? i can't find references to it
[11:49:59] <cradek> tests/t0/shared-test.sh: if nc -z localhost 5007; then
[11:50:16] <cradek> at least some of the tests use it
[11:50:40] <dgarr> ok - i didn't know about nc -- thanks
[11:51:01] <cradek> ahh, yeah, that's how it's spelled
[11:54:34] <dgarr> on debian jessie it seems to be a package named netcat6 and the program is nc6, the package netcat is some docs only
[11:55:47] <cradek> oh good grief
[11:56:13] <cradek> are you sure it's not just an ipv6 version of nc?
[11:56:30] <dgarr> dpkg -L netcat |grep bin returns nil
[11:56:49] <cradek> ok but it might depend on nc4 or something?
[11:56:52] <dgarr> dpkg -L netcat6|grep bin returns /bin/nc6
[11:58:39] <dgarr> now i see there is a package named netcat-traditional and dpkg -L netcat-traditional |grep bin returns nc.tratitional
[11:58:44] <cradek> lrwxrwxrwx 1 root root 20 Nov 7 2013 /bin/nc -> /etc/alternatives/nc
[11:58:47] <cradek> lrwxrwxrwx 1 root root 15 Sep 15 2014 /etc/alternatives/nc -> /bin/nc.openbsd
[11:58:57] <cradek> (wheezy)
[12:01:29] <jepler> hm my jessie system is suspended but I don't remember any particular problem due to nc
[12:01:34] <jepler> I'm sure to have run the testsuite there
[12:02:05] <cradek> I'd guess that if you do apt-get install netcat, you'd get sensible defaults and a program called nc. is that not true?
[12:04:17] <dgarr> i can't try runtests at this moment but will try later to see if the jessie package named netcat works with the runtests, thanks for the info
[12:53:12] <seb_kuzminsky> dgarr: 'dpkg -l' only searches installed packages (plus packages that have been installed and then removed but not purged, or something like that)
[12:53:20] <seb_kuzminsky> you want "apt-cache search --names-only":
[12:53:30] <seb_kuzminsky> 0 11:28:52 seb@jessie-i386 /home/seb> apt-cache search --names-only netcat
[12:53:33] <seb_kuzminsky> netcat6 - TCP/IP swiss army knife with IPv6 support
[12:53:36] <seb_kuzminsky> netcat - TCP/IP swiss army knife -- transitional package
[12:53:38] <seb_kuzminsky> netcat-traditional - TCP/IP swiss army knife
[12:53:41] <seb_kuzminsky> netcat-openbsd - TCP/IP swiss army knife
[12:54:24] <seb_kuzminsky> my jessie machine, with minimal futzing after installing off the iso image, has netcat-traditional
[13:18:41] <dgarr> i was using -L (cap) not -l to see what bin/ files were there and the result confused me not knowing about the /etc/alternatives, i have now done runtests jessie with package netcat, no problems
[13:19:45] <dgarr> dpkg -L netcat (cap L) does not show any bin/ hence the confusion on my part
[13:20:28] <skunkworks> ship it!
[13:30:22] <seb_kuzminsky> ah
[13:32:39] <jepler> yeah, netcat-traditional is also initially installed in this cloud vm of jessie I checked with
[13:34:08] <jepler> .. because it is priority: important ?
[15:22:37] <andypugh> pcw_home: Do you know what might cause the “/DONE is not low after CPLD reset! “ message? This is someone else’s 7i43 installed in the place of my (working) one to test it. What does it take to test it in USB mode?
[15:54:10] <skunksleep> Ugh again.. Last post http://www.cnczone.com/forums/dynomotion-kflop-kanalog/280144-forum.html
[15:55:57] <cradek> in these situations a coworker used to say "why do you think you want that?"
[15:56:41] <andypugh> Why should we care what he wants or chooses? The fewer users we have the less trouble they cause us :-)
[15:56:51] <cradek> I'm with andy
[16:06:14] <skunksleep> You guys!
[16:08:54] <cradek> he'll come here and lecture us about how we need to support k-flippers or what have you
[16:09:41] <cradek> I have decided to have 4 AVRs and a kflipper on each one, and punchcards with gcode on them
[16:10:00] <cradek> because it's an architecture that makes sense to me
[16:10:10] <skunksleep> You forgot the arduinos
[16:10:30] <cradek> did you see the video of the glass 3d-printer? it's actually really neat.
[16:10:47] <skunksleep> On
[16:11:11] <cradek> https://vimeo.com/136764796
[16:11:35] <skunksleep> No
[16:13:50] <seb_kuzminsky> neat, looks like honey
[16:13:57] <seb_kuzminsky> glowing white hot honey
[16:16:22] <seb_kuzminsky> ah, mit center for bits & atoms, should have known ;-)
[16:18:17] <skunksleep> I would think they would have to keep that whole area at close to glass melting temp
[16:19:30] <seb_kuzminsky> looks like the print volume is inside a box of refractory brick
[16:23:39] <skunksleep> Ah.. Didn't watch it that close
[16:26:45] <seb_kuzminsky> bbl soft serve
[16:29:35] <skunksleep> Yum
[16:30:33] <skunksleep> Or is that hacker for something?
[16:36:05] <seb_kuzminsky> the melty glass swirls made me think of it
[17:28:05] <PCW> andypugh: /done not low after CPLD reset
[17:28:06] <PCW> probably means not jumpered right or bad chips (assumed everything else is right)
[17:28:08] <PCW> Are the initial LEDs the same as yours? (in parallel mode with just power you should get a yellow and 2 reds)
[17:28:26] <PCW> s/assumed/assuming/
[17:32:43] <andypugh> Yes, yellow and two reds, and the jumpers are right. (both down)
[17:33:11] <PCW> if you move the OPT1 jumper up and cycle the power, the 7I43 should self config and extinguish the 2 red LEDs
[17:33:13] <PCW> (not guaranteed but all 7I43s had their EEPROM written with a test config and part of the test procedure)
[17:33:14] <PCW> If this doesn't work its usually bad news
[17:33:28] <PCW> as part of
[17:35:40] <andypugh> I just tried it, and it worked.
[17:35:56] <andypugh> (suggests that the jumpers work too)
[17:36:45] <PCW> so you are using all your stuff (cables etc) except he 7I43 card?
[17:37:39] <andypugh> I didn’t have it connected to the machine IO, just plugged into the same parallel cables.
[17:38:06] <PCW> same FPGA size as yours?
[17:38:33] <andypugh> I think so, 7i43-U-4
[17:39:16] <andypugh> With the cable flipped (as a test) I got the wrong-size message, but not with it the right way round.
[17:42:53] <PCW> hmm not sure I guess it could be a bad bit on the EPP interface
[17:46:10] <PCW> was this from a working system that quit?
[17:47:04] <andypugh> I don’t think so, I don’t think it has ever worked
[17:48:22] <andypugh> It belongs to cncbasher
[17:52:30] <PCW> if its under warranty we will replace it
[17:54:22] <PCW> strange problem though, most often bad 7I43s have damaged CPLDs or FPGAs and get hot
[17:54:57] <andypugh> In USB-mode only one light goes off when it is powered up
[17:55:26] <PCW> Thats correct
[17:56:03] <PCW> all sounds good except EPP comms
[17:57:05] <PCW> Dont think we have any way to talk to it over USB without Windows
[17:58:16] <PCW> the USB interface shares the 8 EPP data bits with the parallel interface so it might be broken also
[17:58:57] <andypugh> I have a windows VM. Can you email me some clues? I need to sleep now :-)
[18:09:21] <micges> PCW: mesaflash can access 7i43 via usb
[18:39:58] <PCW> Ah I had forgotten that
[21:14:14] <cradek> http://timeguy.com/cradek-files/emc/.old/dscn6186.jpg
[21:14:43] <cradek> first thread cut by emc2 on fred p's lathe, on loan from nist
[21:15:46] <cradek> ooh, vintage video: http://timeguy.com/cradek-files/emc/.old/dscn6185.mov
[21:15:46] <Tom_itx> nice
[21:16:17] <cradek> Apr 2006 (I just found the email to fred sharing these, while looking for something else)
[21:28:55] <seb_kuzminsky> linuxcnc-build: force build --branch=2.6 0000.checkin
[21:28:56] <linuxcnc-build> build #3385 forced
[21:28:56] <linuxcnc-build> I'll give a shout when the build finishes
[21:28:58] <seb_kuzminsky> here goes nothing
[21:31:50] <linuxcnc-build> build #0 of 1500.rip-jessie-i386 is complete: Failure [4failed install-missing-build-dependencies configuring] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1500.rip-jessie-i386/builds/0
[21:31:52] <linuxcnc-build> build #0 of 1502.rip-jessie-amd64 is complete: Failure [4failed install-missing-build-dependencies configuring configure] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1502.rip-jessie-amd64/builds/0
[21:35:09] <jepler> > libgnomeprintui2.2-dev
[21:35:13] <jepler> afaik no alternative is available
[21:36:15] <jepler> and 2.6 branch doesn't know about "uspace"
[21:36:53] <jepler> I don't think building 2.6 is worth the time without uspace
[21:38:02] <seb_kuzminsky> yep
[21:58:02] <jepler> thanks for working on buildbot!
[21:59:15] <seb_kuzminsky> sure thing
[22:03:19] <jepler> goodnight
[22:05:22] <seb_kuzminsky> nighty night
[22:11:57] <seb_kuzminsky> finally
[22:11:58] <seb_kuzminsky> geez
[22:18:33] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.5-for-jessie 7dac8eb 06linuxcnc 10debian/configure 10debian/control.in packaging: jessie doesnt have libgnomeprintui2.2-dev * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7dac8eb
[22:22:40] <linuxcnc-build> build #1 of 1500.rip-jessie-i386 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1500.rip-jessie-i386/builds/1 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[22:35:03] <linuxcnc-build> build #1 of 1502.rip-jessie-amd64 is complete: Failure [4failed compile] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1502.rip-jessie-amd64/builds/1 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[22:40:29] <seb_kuzminsky> ok i'm starting to feel silly
[22:40:34] <seb_kuzminsky> 2.5 on jessie isn't worth it
[22:40:36] <seb_kuzminsky> wonder if 2.6 works
[22:52:43] <seb_kuzminsky> looks like 2.6 builds, with only minor tweaks (the libgnomeprintui thing)
[22:57:44] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.5-for-jessie 49c7fa2 06linuxcnc 10scripts/platform-is-supported packaging: dont build 2. on debian jessie * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=49c7fa2
[22:58:10] <seb_kuzminsky> that turns off builds on jessie for 2.5
[22:58:32] <seb_kuzminsky> good night, more to come tomorrow
[23:10:50] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/debs_wodocs cefa3a2 06linuxcnc 10debian/control.in debian/control.in: reformat depends line only * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=cefa3a2
[23:10:50] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/debs_wodocs ebc6514 06linuxcnc 10debian/control.in debian/conrol.in jessies pkt name: tcl-tclreadline * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ebc6514
[23:10:50] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/debs_wodocs be45be4 06linuxcnc 10(7 files) debian support deb build without doc packages * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=be45be4