#emc-devel | Logs for 2010-09-09

[00:21:22] <MarkPictor> jepler: src/emc/rs274ngc/tool_parse.cc and .hh don't have a license comment. they're gpl v2, right?
[00:22:25] <cradek> I think that code was directly split out of an existing file - check git history to be sure - if so they have the same license as the original file
[00:23:04] <MarkPictor> of course! I knew they did... I just didn't think about it
[00:23:54] <MarkPictor> for some reason, the comment that wend along with loadToolTable() got lost when the function was put in a separate file
[00:24:22] <cradek> MarkPictor: I think jepler did the split - he will see this and put the license comments in, thanks for pointing it out
[00:24:22] <MarkPictor> It doesn't mention the license, but someone who can commit might want to put it in the new file
[00:24:28] <MarkPictor> ok
[00:24:46] <MarkPictor> the missing comment can be found in http://git.linuxcnc.org/gitweb?p=emc2.git;a=blobdiff;f=src/emc/iotask/ioControl.cc;h=d06c91ce54db36f9a5a6b2149ff83cf39da710ab;hp=c9730c95ff61a82ab07eb5e64f518fcd6c876b43;hb=9e03062a872aaf7f50019b8ea37a5efb6ba71408;hpb=81d7f2a54260347e0a251d7040242308802e0ac5
[00:25:09] <MarkPictor> the new tool table format is pretty stable, isn't it?
[00:25:19] <MarkPictor> I'm thinking of borrowing that code
[01:24:14] <KimK> Re the link to the Honeywell Kitchen Computer (1969) posted earlier by jepler http://en.wikipedia.org/wiki/Honeywell_316#Kitchen_Computer , don't miss this from the footnotes: http://valerieaurora.org/kitchen.html
[03:13:17] <CIA-5> EMC: 03jepler 07v2.4_branch * r2c7d59c009f1 10/src/emc/rs274ngc/ (tool_parse.cc tool_parse.h): add missing copyright notices
[19:45:35] <cradek> cool, matt shaver found a copy of "NCMS"
[20:04:26] <KimK> NCMS?
[20:04:49] <cradek> the ngc interpreter was based partly on it
[20:05:13] <KimK> Ah, OK, thanks
[20:05:48] <cradek> it's actually a hideous mess - interesting to see. :-)
[20:06:02] <Jymmm> KimK: That's ok, I had no clue what it meant either, doubt it stood for what google said =)
[20:07:02] <cradek> comments all over the interpreter talk about [NCMS], but I had never seen a copy
[20:27:38] <KimK> Maybe a copy should be put on the wiki? There should be a place to archive obsolete stuff.
[20:28:43] <KimK> I wouldn't mind looking at it briefly. (Probably very briefly, lol.)
[20:29:00] <cradek> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?RecentChanges
[20:34:29] <jepler> hm, there's an option to "square off" the corners in cutter comp. page number 109 TTRC Generated Blocks (G39 vs G39.1)
[20:36:59] <jepler> G33 uses F (or E?) to specify thread pitch, and includes Q to specify starting angle and K to specify change in thread pitch per revolution
[20:37:06] <jepler> (page number 120)
[20:37:36] <cradek> g28 actually does homing in ncms
[20:38:07] <jepler> er, G34 takes K; the table with G33 wrongly includes K if you believe the prose
[20:38:25] <cradek> the whole document is a frightful mess
[20:38:42] <jepler> we should send a copy to the fellow who was so disturbed by the state of emc documentation yesterday
[20:38:49] <cradek> haha
[20:39:48] <jepler> and what document would be complete without a table of 7-bit ascii. I should add one to our documentation..
[20:40:55] <cradek> "The previous section to the cutter compensation and TTRC."
[20:41:15] <cradek> ^ circled and a "?" written nearby
[20:41:26] <jepler> there's something about random toolchangers on page 75
[20:42:05] <jepler> G10.1 L20 being used to program tool table entries
[20:42:46] <jepler> tool length offset (plus), tool length offset (minus) (page 72)
[20:43:49] <cradek> heh, in their polar mode, X is radius and Y is angle
[20:45:00] <jepler> lots more parameters with defined meanings, page 54ff
[20:45:40] <andypugh> I just joined, are you talking about what I think you are?
[20:45:53] <cradek> that depends
[20:45:56] <jepler> For example, programming #3006=.1 (Install Tool Number 6); will cause program execution to stop at the beginning of this block and display the message telling the operator to read the comment in the block.
[20:46:14] <cradek> uh
[20:46:23] <andypugh> No, clearly not.
[20:46:38] <jepler> andypugh: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?NCMS
[20:47:01] <andypugh> SO I won't campaign for #5501 (or something unused) to report the current value of X (for example)
[20:48:41] <jepler> andypugh: that's page 57
[20:48:57] <andypugh> Ah, yes I will. 5001 - 5006 could be jolly handy at times.
[20:49:02] <jepler> #5001 to 5006 Coordates of End Point, #5021 to 5026 Coordinates of Commanded Position
[20:49:17] <andypugh> Would save frigging it with G92 then G92.2
[20:49:47] <alex_jon1> august 1994, it's not even _that_ old
[20:50:53] <jepler> huh, G38 is hole probing cycle and G38.1 is parallel probing cycle ("paramacro"?)
[20:52:48] <jepler> exponential acceleration? (page 38)
[20:53:12] <cradek> #3002 System Clock - The maximum value for this parameter is 1 year. ... It is incremented by 000005556 every 20 ms.
[20:53:18] <andypugh> For induction motors maybe?
[20:54:02] <jepler> ah, what we call probing is G31-G31.4 "External Skip" and/or G37-G37.4 "Tool Gauging"
[20:56:05] <jepler> there are 4 different probe inputs, G31 and G31.1 selecting the first, G31.2 selecting the second, and so on
[20:56:48] <jepler> and probe tip radius compensation is included
[20:57:00] <cradek> I think 5021... and 5041... are offset current position and unoffset current position
[20:58:05] <jepler> I don't see how that works -- even if your motion is strictly in Z that doesn't mean the point of contact between work and probe tip is exactly in the line of the spindle
[21:01:09] <alex_jon1> lol @ 000005556 every 20 ms.
[21:01:58] <andypugh> Hmm, with a fresh LiveCD install the SIM-5AXIS won't go into world mode, or run an MDI command (because of that)
[21:02:22] <skunkworks> don't you have to home first?
[21:02:35] <andypugh> Yes.
[21:03:17] <andypugh> Ah, OK. I restarted it and it works. It seems to get stuck if you try to go to MDI mode before homing it.
[21:05:12] <andypugh> I think this might be related to Viesturs problem with dropping into joint mode and racking his gantry.
[21:09:00] <jepler> bbl
[22:19:01] <alex_jon1> alex_jon1 is now known as alex_joni
[22:19:34] <moopy> should submit a patch via git for the genserkins hacks i made?
[22:20:53] <moopy> it seems to work with linear joints now
[22:23:05] <andypugh> Do you have commit access to the repository?
[22:23:27] <andypugh> You probably need to make a git diff, then send it to the dev mailing list.
[22:23:54] <moopy> does anyone want to test and submit it for me?
[22:24:29] <andypugh> email the diff to the mailing list, then anyone who wants to can test it.
[22:24:45] <moopy> i dunno what access i have probably non, i dont know anything about git
[22:25:28] <andypugh> I would anticipate "none" too. The stuff I have done has got in through the mailing list.
[22:25:39] <moopy> want to test it andypugh?
[22:26:10] <andypugh> I don't think I know enough about the current Genserkins to usefully test it.
[22:26:36] <moopy> http://soundproofingforum.co.uk/halitosis/
[22:26:53] <moopy> it should all be autoconfigured
[22:27:41] <moopy> i left a couple of demo files in the tar
[22:31:39] <andypugh> Looks good, but I am rather busy at the moment (project overload)
[22:32:59] <seb_kuzminsky> holy crap, now they're talking about evacuating the west side of boulder city... 50 mph winds forecast for this evening
[22:33:29] <alex_joni> seb_kuzminsky: ouch
[22:33:35] <alex_joni> you're east, right?
[22:34:09] <seb_kuzminsky> i'm 3 blocks east of the area that's being told to prepare to evacuate
[22:34:29] <seb_kuzminsky> i strategically bought a house downwind from a really expensive neighborhood, so my house will be well protected ;-)
[22:34:33] <moopy> 50mph is nothing
[22:34:49] <seb_kuzminsky> it is if there's a dry forest on fire upwind!
[22:35:14] <moopy> not if you have an asbestos house
[22:35:20] <seb_kuzminsky> heh
[22:53:31] <alex_joni> .away sleep
[22:53:48] <alex_joni> meh, typos ftw
[22:54:22] <Jymmm> alex_joni: I speak typo