#linuxcnc-devel | Logs for 2015-07-15

[11:10:46] <mozmck> PCW: you around?
[11:18:25] <PCW> sort of
[11:38:22] <jepler> RX bytes:68654384538 (63.9 GiB) TX bytes:111132015028 (103.4 GiB)
[11:39:02] <jepler> it's running reliably now that I switched back to rtapi_delay in the recv loop. no dropped packets, no collisions
[11:39:08] <jepler> 1.25kHz I think
[11:39:55] <PCW> Not bad
[11:40:59] <PCW> I'll have to try on a faster machine (assuming I can create some time somehow)
[11:41:24] <jepler> I think that rate is just shy of 1 billion packets per day
[11:41:27] <jepler> creating time is a good trick
[11:46:58] <PCW> There have been several attempts to remove the delay :-(
[11:56:08] <jepler> I did succeed in moving the delay "down", until after the first recv returns -1
[11:56:43] <jepler> without doing that, the delay is taken at least once per card, but probably only one delay is necessary
[11:57:03] <jepler> the maximum timeout of 100ms (I think) is way too big...
[11:58:19] <PCW> timeout should probably default to 7/10 of the servo period or some such
[11:58:33] <jepler> one problem is, I don't think the thread period is known way down at that layer
[11:58:43] <jepler> at least it's not a function parameter, it might be somewhere
[12:08:56] <PCW> error handling is still rather "undeveloped"
[12:20:42] <PCW> maybe I should add a bug mode to the 7I80 that randomly drops packets...
[12:52:16] <seb_kuzminsky> cmorley: do you have any insight into Rick Lair's problem on emc-users? he's using your python fanuc T-work remapping script, but i dont know if that's related to his problem
[12:54:06] <skunkworks> seb_kuzminsky, did you mean 75000 line gcode?
[12:54:47] <seb_kuzminsky> did i?
[12:54:48] <seb_kuzminsky> err
[12:55:30] <skunkworks> 7500 didn't seem like 140kb worth.. :)
[12:55:54] <seb_kuzminsky> 20 bytes per line
[12:55:56] <seb_kuzminsky> seems about right
[12:56:16] <skunkworks> maybe I don't have a feel for that...
[12:56:19] <skunkworks> :)
[12:56:38] * skunkworks crawls back under a rock
[12:56:46] <seb_kuzminsky> me neither, but i have a calculator ;-)
[12:57:52] <mozmck> what is a Call Stack Under run?
[12:58:15] <seb_kuzminsky> it's a bug in the interpreter's handing of o-word returns
[12:58:56] <seb_kuzminsky> it's in interp_o_word.cc, in the leave_context() function
[12:59:36] <seb_kuzminsky> i bet it's got something to do with the python remapping of T that he's using, but i dont know that code at all...
[13:00:13] <seb_kuzminsky> well, back to $DAYJOB for me
[13:21:47] <skunkworks> PCW, now your out of 7i92's
[13:22:15] <skunkworks> seems like an endless battle
[13:22:17] <skunkworks> :)
[13:23:01] <PCW> building more now but will be around 3-4 weeks
[13:23:45] <skunkworks> I ordered 2 a few days ago - I don't know if I was a lucky one though
[13:23:48] <PCW> first lot sat on the shelf for ~6 months
[13:23:53] <skunkworks> (from the web)
[13:23:59] <PCW> yeah yours is boxed
[13:24:22] <skunkworks> awesome!
[13:24:31] <PCW> I think moz got the last ones
[13:24:34] <skunkworks> heh
[13:24:40] <skunkworks> mozmck!!
[13:24:45] <mozmck> There were 3 this morning, then 2 and I ordered them :)
[13:25:00] <mozmck> I wanted five :(
[13:25:05] <skunkworks> yah - I saw there wasn;t many left
[13:35:05] <skunkworks> wow. mach doesn't reliebly turn the spindle on?
[13:36:46] <mozmck> Where do you see that?
[13:36:53] <skunkworks> cool... http://www.cnczone.com/forums/tormach-pathpilot-/276208-tormach-software-forum-posts.html
[13:37:28] <skunkworks> there is a large discussion on the yahoo group about detecting the spindle not being on - and how to stop mach
[13:38:37] <mozmck> huh, that's generally reliable. We have seen some odd bugs where M3 would do nothing, and you had to restart Mach for it to work again, but haven't seen that in a while.
[13:38:42] <skunkworks> and then this comment http://www.machsupport.com/forum/index.php/topic,30471.msg211593.html#msg211593
[13:39:13] <mozmck> Could be the same bug we've seen at times.
[13:39:52] <mozmck> I think we fixed it by removing bogus extra characters from the end of the M3 macro. It is an external macro in Mach.
[13:50:26] <jepler> weird, this motherboard (ASRock H97 "Anniversary" LGA 1150) has a PATA-era power 4-pin connector on the motherboard
[13:52:34] <jepler> PCIE_PWR1"
[13:53:15] <skunkworks> most motherboards still have the extra 4 pin motherboard connector.
[13:53:16] <seb_kuzminsky> for running those nvidia cards you have
[13:53:30] <jepler> yeah that seems to be the point
[13:53:46] <jepler> I never have that kind of video card, thank goodness
[13:53:58] <jepler> Connect "when more than three graphics cards are installed"
[13:54:47] <skunkworks> wow
[13:55:14] <skunkworks> oh - I see what you are saying. the molex inline 4 pin connector
[13:55:21] <jepler> right
[13:55:36] <skunkworks> ah - I was thinking the 4 pin square motherboard connector
[13:56:47] <jepler> in addition to the molex power connector, this has ATX 24-pin and 8-pin connectors
[14:14:51] <skunkworks> hmm I would think you would not want to use any halui stuff for realtime control...
[14:16:04] <seb_kuzminsky> +1
[14:16:15] <seb_kuzminsky> halui is not realtime
[14:16:24] <seb_kuzminsky> doesn't motion have a pause input for this reason?
[14:16:36] <cradek> there is a realtime feedhold motion pin
[14:16:50] <skunkworks> right
[14:16:56] <seb_kuzminsky> motion.feed-inhibit
[14:17:11] <skunkworks> 1 second seems long though.. unless the machine is really slow
[14:17:12] <seb_kuzminsky> there's also motion.feed-hold, not sure what the diff is
[14:17:35] <cradek> fiik
[14:36:04] <skunkworks> halui: don't forget the Task mode when queueing MDI commands
[14:36:28] <seb_kuzminsky> i looked at that one, i think it's ok
[14:36:28] <skunkworks> otherwise - if this is a gmocappy issue - there was a lot of changes between 2.5 and 2.6
[14:36:45] <seb_kuzminsky> he retracted that, he's now saying it's between 2.6.5 and 2.6.6
[14:37:18] <seb_kuzminsky> i bet motion.feed-hold (and -inhibit) won't pause traverse moves
[14:37:23] <seb_kuzminsky> so it's not quite what rick wants
[14:37:33] <skunkworks> oh - that was a change.
[14:37:38] <cradek> sure it will!
[14:37:44] <cradek> at least one will
[14:37:47] <cradek> I hope
[14:37:56] <seb_kuzminsky> both say feed on them
[14:38:03] <skunkworks> I read it that 2.6.5 worked 2.6.6 didn't
[14:38:10] <seb_kuzminsky> someone... who's not at work... or who works on linuxcnc... should try it
[14:38:13] <seb_kuzminsky> maybe rick
[14:38:40] <mozmck> by traverse do you mean rapid or jog?
[14:38:48] <seb_kuzminsky> i meant traverse
[14:38:52] <seb_kuzminsky> err
[14:38:53] <seb_kuzminsky> rapid
[14:38:55] <seb_kuzminsky> duh
[14:38:56] <seb_kuzminsky> hah
[14:38:57] <cradek> he means G0 gcode
[14:39:09] <cradek> people call it all sorts of things
[14:39:10] <seb_kuzminsky> what he said ^^^
[14:39:23] <mozmck> ok. feed-inhibit pauses all moves.
[14:39:32] <seb_kuzminsky> well that's great
[14:39:41] <mozmck> including jog. at least is does on 2.7 because I've tested it.
[14:39:42] <cradek> what's feed-hold do? (he asked, hesitantly)
[14:39:47] <seb_kuzminsky> we should update the manpage to say that too
[14:39:48] <mozmck> I don't know!
[14:39:51] <seb_kuzminsky> haha
[14:39:56] <cradek> "we"
[14:40:00] <seb_kuzminsky> we're like the monkey men in 2001
[14:40:06] <cradek> ook ook
[14:40:13] <cradek> I hear I'm the mild-mannered one
[14:40:15] <mozmck> The docs made it look like I wanted feed-inhibit so I used that.
[14:40:19] * seb_kuzminsky hits linuxcnc with a stick
[14:40:48] <cradek> I've been accused of hitting it with a knobkerrie
[14:41:18] <seb_kuzminsky> i dont know what that is but it sounds like an awful scottish intestine-based food
[14:41:33] <cradek> eww
[14:41:34] <mozmck> hotdogs?
[14:41:39] <cradek> eww
[14:41:41] <seb_kuzminsky> bbl lunch
[14:41:44] <cradek> ha
[15:03:16] <mozmck> hmm, hal_gremlin emits a signal line-clicked, but not if gremlin is clicked outside of a line.
[15:04:22] <mozmck> so you can highlight the gcode line if a line is clicked in gremlin, but I don't see a way to clear it if it is cleared in gremlin...
[15:51:07] <mozmck> Here's a fix, but will need testing to see if it breaks anything in Gmoccapy or Gscreen
[15:51:08] <mozmck> http://pastie.org/10295277
[15:53:09] <mozmck> <grumble> Gmoccapy *crashes* if you have a non-theme file in ~/.themes
[15:56:11] <mozmck> In my case it was a .tar.bz2 of a theme
[16:00:18] <mozmck> Looks like my hal_gremlin fix does not break gmoccapy or gscreen, but it does fix the same problem in gmoccapy at least.
[16:20:30] <mozmck> seb_kuzminsky: if you think my patch above is non-disruptive enough I'll push it to 2.7 tonight.
[17:02:22] <seb_kuzminsky> mozmck: are those the only users of line-clicked?
[17:05:24] <seb_kuzminsky> hmm, i see hal_gremlin emit that signal, but i dont see anyone catching it. what don't i understand about gsignals?
[19:16:42] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 4ef1626 06linuxcnc 10docs/man/man9/motion.9 docs: document motion.feed-inhibit better * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4ef1626
[19:20:58] <seb_kuzminsky> i should verify that spindle-sync motions are not inhibited, like they shouldn't be
[19:47:51] <seb_kuzminsky> yeah, it totally does the right thing in 2.6
[19:48:09] <seb_kuzminsky> linuxcnc, i forgive all your annoying quirks, because you got it where it counts
[19:48:21] <cradek> haha
[19:48:38] <cradek> ok now tell me you'll test 2.7 too
[19:48:50] * seb_kuzminsky hides
[19:49:20] <seb_kuzminsky> elsewhere in the news, i ran g33.1 for the first time today
[19:49:23] <seb_kuzminsky> i feel like a big boy
[19:49:26] <cradek> yay!
[19:49:30] <seb_kuzminsky> (it was only in simulation)
[19:49:32] <cradek> oh
[19:49:34] <cradek> boo
[19:49:36] <seb_kuzminsky> heh
[19:50:05] <cradek> you must not really need tapping on your machine
[19:50:12] <seb_kuzminsky> yeah, never
[19:50:37] <seb_kuzminsky> not even when i made a 5x9 inch fixture plate with tapped 8-32 holes on 1" centers
[19:50:49] <cradek> eek
[19:51:13] <cradek> I did one with 10-32 for jepler's machine
[19:51:27] <cradek> I used a tapmatic in the drill press, and it actually worked ok
[19:51:34] <seb_kuzminsky> cool
[19:51:41] <cradek> don't you have that tapping head?
[19:51:54] <seb_kuzminsky> yeah but i lack the balls to spin it up
[19:52:01] <cradek> ha
[19:52:08] <cradek> I actually used that on the bridgeport quite a bit
[19:52:17] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 0e98fc8 06linuxcnc 10docs/src/gcode/gcode.txt docs: update G33.1 example to include S100 M3 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0e98fc8
[19:52:24] <seb_kuzminsky> your balls must be bigger than mine
[19:52:33] <cradek> I sure won't comment on that
[19:53:10] <seb_kuzminsky> i should finish the spindle encoder
[19:53:16] <cradek> and I'm still trying to figure out how many holes would be in a 5x9 plate
[19:53:20] <cradek> is it 32?
[19:53:36] <cradek> yeah probably
[19:53:44] <cradek> that's a lot to do by hand
[19:54:54] <skunkworks> seb is the one with salty nuts
[19:55:15] <seb_kuzminsky> well it's 5ish by 9ish, whatever the first piece of scrap i found was
[19:55:26] <seb_kuzminsky> it's got, uh, 40 holes
[19:55:33] <seb_kuzminsky> so maybe by 10ish
[19:56:28] <cradek> ooh https://www.youtube.com/watch?v=9HLKXeWqTF0
[19:56:29] * seb_kuzminsky runs out to the garage
[19:57:17] <Tom_itx> i'm wanting to test the rigid tapping soon as i'm able... have the machine all set for it with a test program
[19:57:19] <seb_kuzminsky> http://highlab.com/~seb/linuxcnc/fixture-plate.jpg
[19:57:26] <cradek> about 8 seconds each
[19:57:47] <seb_kuzminsky> 8 seconds/hole, that's about what it took me to do it by hand
[19:57:51] <seb_kuzminsky> not
[19:58:08] <cradek> yeah I bet!
[19:58:28] <cradek> maybe you could have driven to my place and had it done faster
[19:58:32] <seb_kuzminsky> haha
[19:58:36] <seb_kuzminsky> possibly!
[19:58:56] <Tom_itx> you're going clear thru on those cradek?
[19:59:05] <Tom_itx> looks plenty deep
[19:59:07] <cradek> yeah, they go through
[19:59:19] <Tom_itx> is that a straight flute tap?
[19:59:36] <cradek> surely it was spiral point
[19:59:42] <Tom_itx> i'd hope
[20:00:01] <cradek> I think spiral points have totally straight flutes
[20:00:22] <Tom_itx> err i meant high helix
[20:01:03] <Tom_itx> they discharge chips better than even a spiral point
[20:01:10] <cradek> http://davethetools.com/ebay/var/resizes/Taps/spiral-point-gun-tap.jpg
[20:01:21] <cradek> I use the ones that look like this for that kind of hole
[20:01:27] <Tom_itx> right
[20:01:39] <cradek> they're magical
[20:02:19] <Tom_itx> wrong size but this is what i meant: http://www.ebay.com/itm/like/360240029445?lpid=82&chn=ps
[20:03:06] <cradek> I've used similar in blind holes
[20:03:13] <cradek> the chips come up?
[20:03:19] <Tom_itx> pretty sure
[20:03:30] <Tom_itx> they work best on cnc
[20:04:04] <Tom_itx> the spiral point do ok though
[20:04:52] <seb_kuzminsky> hmm, halscope failed to load for me in 2.7 just now
[20:05:04] <seb_kuzminsky> i hope i'm doing something dumb
[20:07:29] <cradek> well I'm sure not getting done the things I should be doing
[20:07:32] <cradek> bbl
[20:07:42] <seb_kuzminsky> seeya
[20:08:24] <jepler> PCW: no room in idrom for connector silkscreen names?
[20:08:53] <jepler> or in one of the other address spaces supported by the ethernet boards
[20:09:50] <PCW> probably is room by adding some non-hardware-module tags
[20:10:21] <seb_kuzminsky> jepler: dont we need mods to the llio driver for each board anyway? might as well put the connector names there, no?
[20:11:14] <PCW> a compatible daughter-board hint tag would be valuable
[20:12:03] <jepler> seb_kuzminsky: well I wrote this http://paste.debian.net/282767/ disabled the existing case to handle the 7i90, and everything seems to work but printing the correct pin names
[20:12:27] <jepler> hm2_eth: Unrecognized ethernet board found: 7I92
[20:12:35] <jepler> hm2/hm2_7i92.0: 34 I/O Pins used:
[20:12:35] <jepler> hm2/hm2_7i92.0: IO Pin 000 (??-01): IOPort
[20:12:36] <jepler> hm2/hm2_7i92.0: IO Pin 001 (??-14): IOPort
[20:12:51] <seb_kuzminsky> ah
[20:13:44] <jepler> or maybe 01 and 14 are actually right?
[20:13:58] <PCW> They are
[20:14:20] <jepler> (er and I meant 7i92 where i said 7i90 above)
[20:15:03] <jepler> I also don't bother retrieving the FPGA part number though it might be available in one of those other memory spaces
[20:15:34] <PCW> the connector reference designator stuff is currently on the driver but you are right, it should be in the IDROM
[20:15:41] <PCW> in the driver
[20:17:20] <seb_kuzminsky> hmm, homing ignores motion.feed-inhibit
[20:18:37] <jepler> PCW: I have to hard-code a limit for the number of ethernet boards. 4? 8?
[20:19:37] <PCW> 7I92 (and all parallel port pinout cards) have somewhat funny DB-25 pinout (its in physical order so DB25 pin1, 14,2,15 etc)
[20:20:19] <Tom_itx> i noticed that when i did a bitfile for one
[20:20:21] <seb_kuzminsky> yeah, motion.feed-inhibit works as expected with jogs, feed moves, rapid moves, and spindle-synched motions in 2.7
[20:20:24] <seb_kuzminsky> yay
[20:20:41] <PCW> 4 is probably fine
[20:21:03] <jepler> should be just one site to increase an integer constant if anybody needs it
[20:21:19] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 7679b83 06linuxcnc 03docs/src/gcode/g-code.txt Merge remote-tracking branch 'origin/2.6' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7679b83
[20:22:20] <jepler> ah I see that the numbering arrangement is hard-coded in the driver
[20:22:34] <jepler> case 17: /* 25 pin 17 I/O parallel port type cards funny DB25 order */
[20:22:38] <jepler> pin->port_pin = DB25[i % 17];
[20:24:15] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 16c07b7 06linuxcnc 10src/hal/utils/scope.c 10src/hal/utils/scope_rt.c halscope: report shm key when rtapi_shmem_new() fails * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=16c07b7
[20:24:20] <seb_kuzminsky> ahem
[20:27:28] <PCW> There are some other problems in the current IDROM version (notably the inability to specify more than 2 strides)
[20:27:30] <PCW> At some point the version should be bumped to 4 and some of these limitations removed
[20:28:49] <PCW> first thing would be to get rid of the silly indirect stride specifier and just use 2**N in the module ID record
[20:30:07] <PCW> then every module can have its own stride
[20:32:21] <KGB-linuxcnc> 03Jeff Epler 052.7 8315fe8 06linuxcnc 10src/rtapi/uspace_rtapi_app.cc uspace: fix uninitialized bytes in syscall sigaction * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=8315fe8
[20:32:21] <KGB-linuxcnc> 03Jeff Epler 052.7 641e98a 06linuxcnc 10src/rtapi/uspace_rtapi_app.cc uspace: must advise user to set RTAPI_FIFO_PATH * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=641e98a
[20:33:21] <skunkworks> wow - linuxcnc is on fire!
[20:33:25] <skunkworks> ;)
[20:34:19] <KGB-linuxcnc> 03Jeff Epler 05master 66ec566 06linuxcnc Merge remote-tracking branch 'origin/2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=66ec566
[20:41:14] <KGB-linuxcnc> 03Jeff Epler 05jepler/hm2-eth-multiple 6eb5855 06linuxcnc 10docs/man/man9/hm2_eth.9 10src/hal/drivers/mesa-hostmot2/hm2_eth.c 10src/hal/drivers/mesa-hostmot2/hm2_eth.h hm2_eth: allow multiple instances (up to 4) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6eb5855
[20:41:15] <KGB-linuxcnc> 03Jeff Epler 05jepler/hm2-eth-multiple 491e63f 06linuxcnc 10(6 files in 2 dirs) hostmot2: support split reads * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=491e63f
[20:41:15] <KGB-linuxcnc> 03Jeff Epler 05jepler/hm2-eth-multiple 81ff3fb 06linuxcnc 10src/hal/drivers/mesa-hostmot2/hm2_eth.c hm2_eth: in case of failed recv(), show an error * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=81ff3fb
[20:41:17] <KGB-linuxcnc> 03Jeff Epler 05jepler/hm2-eth-multiple 7be45d7 06linuxcnc 10src/hal/drivers/mesa-hostmot2/hm2_eth.c hm2_eth: make unrecognized boards work * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7be45d7
[20:47:59] <skunkworks> why or how do you set the RTAPI_FIFO_PATH path
[21:01:44] <jepler> skunkworks: if you want to run rtapi_app with access to hardware but under a debugger, you have to do tat
[21:01:49] <jepler> $ sudo bin/rtapi_app
[21:01:49] <jepler> [sudo] password for jepler:
[21:01:49] <jepler> Refusing to run as root without fallback UID specified
[21:01:49] <jepler> To run under a debugger with I/O, use e.g.,
[21:01:49] <jepler> sudo env RTAPI_UID=`id -u` RTAPI_FIFO_PATH=$HOME/.rtapi_fifo gdb rtapi_app
[21:02:53] <jepler> not important for your average user
[21:03:40] <Tom_itx> linuxcnc users are average?
[21:04:27] <jepler> some of them, yes.
[21:31:18] <seb_kuzminsky> sad news:
[21:31:21] <seb_kuzminsky> http://blog.cnccookbook.com/2015/07/15/greg-jackson-has-passed-away/
[21:34:46] <cradek> that really sucks.
[21:35:47] <seb_kuzminsky> yeah :-(
[21:40:09] <skunksleep> Wow rip
[21:49:52] <skunksleep> http://journaltimes.com/news/local/lake-michigan-drowning-tragic-accident/article_d2aa217d-cd0e-5e4f-aa7b-18615469fe1e.html
[22:00:58] <cradek> crap, I hope his friend is ok.
[22:22:32] <cradek> I wonder how iceweasel knows when flashplayer is worse than usual and disables it
[22:32:30] <mozmck> I think when iceweasel gets updated it gets new information or something.
[22:32:47] <cradek> it knows more frequently than it updates
[22:32:52] <mozmck> usually there is a flashplayer update soon on my system and that fixes the warnings.
[22:33:05] <cradek> it's done it off and on several times in the last week or two
[22:33:19] <cradek> yeah - I just wonder how it knows.
[22:33:24] <mozmck> I don't know then, maybe it checks something on the web?
[22:33:56] <cradek> yeah, I guess it must
[22:34:07] <mozmck> I use linux mint, and my firefox was fussing about it, and today there was a flashplayer update that fixed it.
[22:34:39] <cradek> I hope flash is over soon
[22:34:44] <mozmck> seb_kuzminsky: you can connect signals in glade.
[22:34:58] <mozmck> heh, I never liked flash for gui stuff at all.
[22:35:13] <cradek> flash has new bugs again today and there's no fix yet
[22:35:56] <mozmck> Adobe quit supporting flash for linux a while back, so all we get are updates to keep iceweasel/firefox from complaining.
[22:36:49] <mozmck> If you need the latest flash (which some webinar sites require :( ), you have to install the google virus.
[22:37:26] <mozmck> You can install chromium, and then there is a way to get the flash plugin out of chrome without installing it.
[22:42:10] <cradek> I remember seeing a package that did that, but I can't seem to spot it right now
[22:43:17] <cradek> oh well, I like that it doesn't auto-play crap right now
[22:47:18] <mozmck> on my system it is pepperflashplugin-nonfree
[22:48:40] <cradek> that sounds right! thanks
[22:49:00] <cradek> it must not be in my sources because it didn't show up in apt-cache search.
[22:49:32] <mozmck> I might have gotten it in a ppa? not sure.
[22:50:08] <cradek> https://wiki.debian.org/PepperFlashPlayer/Installing
[22:50:15] <cradek> it might come from backports
[22:50:20] <mozmck> could be.
[22:50:28] <cradek> I'm not going to try it here, but if people whine enough at work I'll try it there.
[22:50:52] <cradek> it's not my fault, but our own website uses flash
[22:50:58] <cradek> so they sure might whine
[22:50:59] <mozmck> ah, yes. I used it and it worked well.
[22:51:26] <cradek> goodnight!
[22:51:41] <mozmck> I "attended" the tormach webinar, and the service they used wanted a newer flash, so I installed chromium and used that package.
[22:52:00] <mozmck> good night, need to go to bed myself.
[22:53:10] <mozmck> seb_kuzminsky: looks like that signal is used only in the gaxis gscreen skin, and in gmoccapy, and they are both connected in the glade files. search for 'signal name="line_clicked"'
[22:58:28] <seb_kuzminsky> the - became a _?
[23:00:01] <seb_kuzminsky> wow, it did
[23:00:10] <seb_kuzminsky> ok
[23:05:24] <mozmck> When you click on a line in gremlin, it highlights it. When you click in empty space, it removes the highlight. Before my fix, the line of gcode would be highlighted with the line in gremlin, but would not be cleared like the line in gremlin.
[23:20:23] <seb_kuzminsky> i'll take your word for it
[23:38:17] <mozmck> :) I think I'll push it.
[23:38:40] <mozmck> I can't imagine it will cause any problem.
[23:39:18] <seb_kuzminsky> i trust your udgement
[23:39:20] <seb_kuzminsky> *j
[23:44:21] <KGB-linuxcnc> 03Moses McKnight 052.7 ab635c7 06linuxcnc 10lib/python/gladevcp/hal_gremlin.py Make hal_gremlin send 'line-clicked' signal even if line == None * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ab635c7
[23:45:57] <seb_kuzminsky> thx