Back
[00:27:31] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/debs_wodocs 9bdd12d 06linuxcnc 10debian/control.in debian/control.in: reformat depends line only * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9bdd12d
[00:27:31] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/debs_wodocs 7d0335e 06linuxcnc 10debian/control.in debian/control.in jessie pkg name: tcl-tclreadline * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7d0335e
[00:27:31] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/debs_wodocs 16fca77 06linuxcnc 10(7 files) debian: support deb build without doc packages * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=16fca77
[01:15:09] <seb_kuzminsky> linuxcnc-build: force build --branch=2.7 0000.checkin
[01:15:15] <linuxcnc-build> The build has been queued, I'll give a shout when it starts
[01:17:10] <seb_kuzminsky> all the slaves on that hypervisor use cache=writeback now, that's the least flushy-looking virtio disk cache mode of the options in virt-manager
[02:27:37] <linuxcnc-build> build forced [ETA 1h33m12s]
[02:27:37] <linuxcnc-build> I'll give a shout when the build finishes
[02:56:49] <linuxcnc-build> build #1302 of 4017.deb-wheezy-amd64 is complete: Failure [4failed shell_3] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-amd64/builds/1302 blamelist: dummy, Sebastian Kuzminsky <seb@highlab.com>, Dewey Garrett <dgarrett@panix.com>
[02:57:20] <linuxcnc-build> build #683 of 4014.deb-wheezy-rtpreempt-i386 is complete: Failure [4failed shell_3] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/4014.deb-wheezy-rtpreempt-i386/builds/683 blamelist: dummy, Sebastian Kuzminsky <seb@highlab.com>, Dewey Garrett <dgarrett@panix.com>
[02:57:33] <linuxcnc-build> build #722 of 4015.deb-wheezy-rtpreempt-amd64 is complete: Failure [4failed shell_3] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/4015.deb-wheezy-rtpreempt-amd64/builds/722 blamelist: dummy, Sebastian Kuzminsky <seb@highlab.com>, Dewey Garrett <dgarrett@panix.com>
[02:58:27] <linuxcnc-build> build #1300 of 4016.deb-wheezy-i386 is complete: Failure [4failed shell_3] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/4016.deb-wheezy-i386/builds/1300 blamelist: dummy, Sebastian Kuzminsky <seb@highlab.com>, Dewey Garrett <dgarrett@panix.com>
[04:10:11] <linuxcnc-build> Hey! build 0000.checkin #3398 is complete: Success [3build successful]
[04:10:11] <linuxcnc-build> Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/3398
[07:14:49] <KGB-linuxcnc> 03John Thornton 052.7 03a2c21 06linuxcnc 10docs/src/gui/pyvcp.txt Docs: add missing parameter and clean up * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=03a2c21
[07:24:31] <jepler> cradek: the first time we ever had to resize it, I tried to lay some ground work to allow it to be runtime tunable, but I don't think we're there yet
[07:24:48] <jepler> and at this time there are some uses of that macro outside of hal_lib.c too
[07:24:51] <jepler> hal/utils/halcmd_main.c: mem_id = rtapi_shmem_new(HAL_KEY, comp_id, HAL_SIZE);
[07:24:54] <jepler> hal/utils/halrmt.c: mem_id = rtapi_shmem_new(HAL_KEY, comp_id, HAL_SIZE);
[07:29:30] <jepler> anyway you'd insert a hal shared memory size member near the beginning of hal_data_t and everybody but the actual creator of the shared memory region would map with a default size, read the memory size member, and remap with the right size. this is a dance that is done now with e.g. the userspace part of streamer/sampler because they don't know the buffer size that was set as a loadrt parameter of
[07:29:36] <jepler> the realtime part
[07:30:14] <jepler> then you just have to decide how the user will specify the shared memory size. As a commandline argument or envionment variable when "realtime start" is invoked?
[08:48:22] <cradek> modprobe.d/linuxcnc.conf: options hal_whatever ...?
[08:48:37] <jepler> uspace.
[08:48:38] <cradek> it would be very rare that someone needs to adjust it
[08:48:46] <jepler> uspace has no modules
[08:48:47] <cradek> oh
[08:48:58] <cradek> it's time for coffee anyway
[08:49:04] <jepler> yay
[08:49:17] <skunkworks> zlog
[13:33:48] <skunkworks> mesa cards can step into the mhz - correct?
[13:40:20] <PCW> ~clock low/4 so
[13:40:22] <PCW> 8 MHz for 5I25
[13:40:23] <PCW> 12 MHz for 7I43
[13:40:25] <PCW> 25 MHz for Ethernet cards
[13:43:12] <skunkworks> wow
[13:43:23] <skunkworks> hey pcw - how is it going?
[13:46:22] <PCW> aight
[14:24:31] <jepler> and somewhere in there you have to start worrying whether your step and direction signals are going to be intact at the other end of the world
[14:24:49] <jepler> at least they have interleaved grounds on the 50-pin cables
[14:25:18] <cradek> yeah, feed that 25 MHz into some optos for good measure
[14:25:58] <cradek> and listen to it on your shortwave receiver
[14:27:13] <jepler> call it software defined radio and you can get on hackers15minutesoffame.io too
[14:27:39] <jepler> I am still tempted to try the raspberry pi hack that is supposed to broadcast a stereo FM audio signal out one of the GPIOs
[14:28:46] <cradek> if you're itching for a project I would love to have an FM IF alignment wobbulator
[14:29:42] <cradek> it'd wobble around 10MHz with the wobble being a secondary output (for feeding to X on an XY scope)
[14:29:53] <jepler> I would probably have to learn too much to do that
[14:30:41] <cradek> well yeah, me too
[14:30:47] <cradek> which is why I don't have one
[14:30:47] <jepler> there are no real analog outputs on a rpi so I'm not sure how one would do the secondary output
[14:31:34] <jepler> a trigger output would likely be feasible though
[14:31:37] <cradek> it could probably have just a few steps, and a poor-man's D-to-A using some digital outputs and binary resistors
[14:31:48] <cradek> oh yeah, that'd work fine too
[14:32:09] <cradek> sawtooth frequency instead of wobble would be fine
[14:32:34] <cradek> (wobble is easier than sawtooth with vacuum tubes)
[14:32:51] <jepler> what is wobble, sine-ish ?
[14:33:16] <cradek> yeah probably
[14:33:29] <cradek> it doesn't matter, since it just goes to a scope
[14:33:45] <cradek> it even might have been done with a motor
[14:34:08] <micges> jepler: hi, do you have somewhere configs for 7i90 spi setup?
[14:34:57] <jepler> micges: no. I don't think there's anything special, just use any mesa config and change the card prefix
[14:35:40] <micges> thnaks
[14:37:57] <jepler> .. I have not used that setup even for testing in ~1 year now
[14:39:48] <micges> yeah I saw github logs, exactly 1 year
[19:33:32] <arrowbook> the stepgen's author is john kasunich? does he still take part in the project?
[19:34:39] <skunkworks> if you ask a question on the mailing list - he may answer
[19:35:50] <skunkworks> what are you trying to do?
[19:37:08] <arrowbook> i want to understand some equation
[19:48:10] <archivist> cradek, use the scopes x to vco a local osc out of an old rx, adjust it to be mid FM band or use another to mix down to the IF of interest
[20:03:21] <cradek> archivist: I'm not quite seeing it. it's true it could sweep at RF and not IF frequency, but then you have the RF amp's selectivity added in there too
[20:04:33] <archivist> on my signal generator I would just widen the FM and tune it to a suitable place
[20:04:34] <cradek> it would be nice if (modern) scopes had a sweep output
[20:05:09] <cradek> I think I've seen that on some old tube scopes
[20:05:33] <archivist> you can use the local AF osc in an fm generator as the sweep
[20:05:57] <cradek> I don't have a fm generator at all
[20:06:28] <cradek> I do have an AM that goes up to 2x MHz
[20:07:02] <cradek> by unbalancing the discriminator I can hear the AF from it
[20:07:10] <cradek> it's not a great setup :-/
[20:09:25] <archivist> a homebrew sweep osc using a 1N4000 series (as a varicap ) at 10.7 should be easy enough
[20:10:37] <cradek> hmm, I have a simpson genescope on the shelf. I bought it many years ago at fair radio and don't think I ever even plugged it in.
[20:10:56] <cradek> maybe I should quit screwing around and rebuild it
[20:11:10] <archivist> http://www.next.gr/inside-circuits/10.7mhz-sweep-generator-l9369.html
[20:11:29] <cradek> http://www.electronicproducts.com/images2/fasj_worth_simpson-genescope_may2012.gif
[20:11:54] <archivist> ant teeeeeek
[20:12:23] <cradek> my AM is a simpson 315 and it works perfectly
[20:12:44] <cradek> pretty sure I recapped it etc, though
[20:13:21] <cradek> "from cro", hmm
[20:13:25] <archivist> my stuff is a bit later, mostly 1970's and some 1980's
[20:13:27] <cradek> surely my tek doesn't have that output does it?
[20:13:57] <archivist> which tek?
[20:14:38] <cradek> 2232
[20:14:47] <cradek> "aux connector" has x & y marked, but surely they are inputs?
[20:15:11] <archivist> might depends on mode
[20:15:35] <cradek> I've got a 466 somewhere too, and maybe a 465b
[20:15:45] <archivist> some scopes has extra connectors on rear
[20:16:25] <cradek> 485 has sweep outputt
[20:17:40] <archivist> but using a wall wart as a sine sweep source is an option
[20:17:56] <archivist> or any old AF generator
[20:18:30] <cradek> true, 60Hz would be fine, faster than any AGC
[20:18:34] <archivist> wobulators often used mains as the sweep signal
[20:19:01] <cradek> X, Y, RELAY go to "analog plotter" (?)
[20:20:20] <archivist> scope those and see
[20:20:40] <archivist> probably were plotter outputs
[20:31:08] <cradek> 466 has some more outputs but I am having trouble finding the right manual to see what they are (it's out at the shop)
[20:32:21] <cradek> so useful:
http://www.modularparts.com/ELECTRONICS-/Outlets-/Old-/Tektronix-466-100MHZ-oscilloscope-with-dm-44-and-probes-displayImage.jpg
[20:34:13] <cradek> gate outputs: positive gates from both time bases (~ 5V)
[20:34:22] <cradek> ch1 vertical signal
[20:34:30] <cradek> Z in
[20:34:49] <cradek> so not quite what I'd like
[21:30:24] <seb_kuzminsky> you guys
[21:30:44] <seb_kuzminsky> my library finally sent me my copy of seveneves
[21:30:48] <seb_kuzminsky> see you in 2 weeks
[21:31:39] <seb_kuzminsky> "The moon blew up without warning and for no apparent reason. It was waxing, only one day short of full."
[21:32:08] <jepler> wasn't that the plot of space: 1999?
[23:52:40] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/debs_wodocs d4bbfa9 06linuxcnc 10(7 files) debian: support deb build without doc packages * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d4bbfa9
