#emc-devel | Logs for 2011-01-27

[00:04:35] -!- theorbtwo has quit [Ping timeout: 260 seconds]
[00:04:39] theorb is now known as theorbtwo
[00:20:17] -!- mhaberler has quit [Ping timeout: 265 seconds]
[00:25:33] -!- mhaberler [mhaberler!~chatzilla@62-47-138-68.adsl.highway.telekom.at] has joined #emc-devel
[00:27:36] -!- andypugh has quit [Quit: andypugh]
[00:29:49] -!- toastydeath has quit [Quit: Leaving]
[00:38:52] -!- toastydeath has quit [Ping timeout: 255 seconds]
[00:46:53] -!- mhaberler has quit [Ping timeout: 255 seconds]
[00:56:29] -!- adb has quit [Ping timeout: 264 seconds]
[00:58:39] -!- skunkworks has quit [Ping timeout: 240 seconds]
[00:58:48] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[00:59:11] -!- skunkworks [skunkworks!~chatzilla@str-bb-cable-south2-static-6-412.dsl.airstreamcomm.net] has joined #emc-devel
[01:06:22] -!- dgarr has quit [Quit: Leaving.]
[01:12:05] -!- Valen has quit [Quit: Leaving.]
[01:13:14] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc-devel
[01:25:03] -!- robh__ has quit [Ping timeout: 240 seconds]
[02:08:15] -!- Tom_itx has quit []
[02:22:12] -!- tlab has quit [Quit: Leaving]
[02:25:04] -!- fatpandas has quit [Quit: leaving]
[02:28:45] -!- elmo40 has quit [Read error: Connection reset by peer]
[02:37:13] -!- WalterN has quit [Ping timeout: 255 seconds]
[02:46:21] -!- ries has quit [Quit: ries]
[03:01:58] -!- toastydeath has quit [Ping timeout: 255 seconds]
[03:15:41] -!- skunkworks has quit [Read error: Operation timed out]
[03:35:44] -!- frank222 has quit [Quit: Visitor from www.linuxcnc.org]
[03:35:50] -!- Tom_L has quit []
[03:49:40] <CIA-2> EMC: 03cradek 07master * r8a325a06fd42 10/src/hal/components/bldc.comp: Add bldc.comp motor interface component
[03:49:48] <CIA-2> EMC: 03cradek 07master * redd9e8d5443a 10/ (7 files in 3 dirs): Add support for Mesa 8i20 and 7i64
[03:53:24] <cradek> seb_kuzminsky: you can *receive* dcc transfers behind nat, since your machine initiates the transfer
[03:53:38] <cradek> fwiw.
[03:54:01] <cradek> /dcc send is "I have a file for you. connect to host/port to get it"
[03:59:21] -!- toastyde1th has quit [Ping timeout: 240 seconds]
[04:40:58] -!- Emcrules_Laptop has quit [Ping timeout: 255 seconds]
[05:07:05] -!- EDocToor [EDocToor!~chatzilla@69-196-183-9.dsl.teksavvy.com] has joined #emc-devel
[05:19:07] -!- Dannyboy has quit [Remote host closed the connection]
[05:24:05] -!- rooks has quit [Ping timeout: 264 seconds]
[05:29:22] -!- KimK [KimK!~Kim__@ip174-71-95-176.om.om.cox.net] has joined #emc-devel
[05:52:04] -!- toastydeath has quit [Ping timeout: 255 seconds]
[05:55:13] -!- toastyde1th has quit [Ping timeout: 255 seconds]
[06:15:55] -!- Connor has quit [Ping timeout: 255 seconds]
[06:15:56] -!- vezza has quit [Quit: Sto andando via]
[06:16:12] -!- Connor1 has quit [Ping timeout: 240 seconds]
[06:59:52] -!- LawrenceG has quit [Remote host closed the connection]
[07:13:18] -!- mhaberler [mhaberler!~chatzilla@62-47-143-22.adsl.highway.telekom.at] has joined #emc-devel
[07:26:34] -!- drill has quit [Ping timeout: 255 seconds]
[07:35:30] -!- Dannyboy has quit [Ping timeout: 260 seconds]
[07:51:58] -!- pingufan has quit [Remote host closed the connection]
[08:11:07] -!- psha [psha!~psha@] has joined #emc-devel
[08:44:25] -!- adb has quit [Ping timeout: 255 seconds]
[08:49:14] -!- WalterN has quit [Read error: Connection reset by peer]
[08:50:00] -!- micges_work [micges_work!~x@aejb205.neoplus.adsl.tpnet.pl] has joined #emc-devel
[09:27:05] -!- mhaberler has quit [Ping timeout: 255 seconds]
[09:33:27] -!- mhaberler [mhaberler!~chatzilla@62-47-143-221.adsl.highway.telekom.at] has joined #emc-devel
[09:44:00] -!- mhaberler has quit [Read error: Operation timed out]
[09:46:08] -!- KimK has quit [Quit: Leaving]
[09:51:54] -!- maximilian_h [maximilian_h!~bonsai@ulmg-5d84dd8f.pool.mediaWays.net] has joined #emc-devel
[09:52:21] -!- robh__ [robh__!~robert@5ace70af.bb.sky.com] has joined #emc-devel
[10:20:10] -!- robh__ has quit [Quit: Leaving]
[10:25:06] -!- awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc-devel
[10:25:54] -!- maximilian_h has quit [Ping timeout: 265 seconds]
[10:43:20] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc-devel
[11:05:13] -!- Techrat has quit [Quit: Leaving]
[11:28:27] -!- mhaberler [mhaberler!~chatzilla@62-47-129-108.adsl.highway.telekom.at] has joined #emc-devel
[11:30:07] <mhaberler> hi all - anybody out here willing to discuss interpreter internals?
[11:30:47] <archivist> just throw the question in then a lurker can answer
[11:33:10] <psha> mhaberler: you'd better wait till 13 GMT when jeff will be up
[11:52:37] -!- psha has quit [Quit: Lost terminal]
[11:52:39] <mhaberler> ok, so.. subject is interpreter instantiation and state
[11:52:41] <mhaberler> In theory the Interpreter is an instance object
[11:52:42] <mhaberler> however, state is really global, and has the following components:
[11:52:44] <mhaberler> 1. the 'world model' (machine actual state?), in the _setup static member struct
[11:52:45] <mhaberler> 2. the interpreter execution state (current block, stack, filedescriptors, o-word handling auxiliary structures etc)
[11:52:47] <mhaberler> 3. the interpreter list of generated NML messages
[11:52:48] <mhaberler> 1+2 are lumped together in setup_struct
[11:52:50] <mhaberler> 3 - there's really just a single list on which task operates.
[11:52:51] <mhaberler> Let's assume I'd like to create a new interp instance to execute something in the middle of a block, like replacing a Tx or M6 by an oword sub call
[11:52:53] <mhaberler> I want it to inherit the world model, execute a subroutine, which might modify the world model, and terminate, but I would NOT want it to modify otrher interp instances execution state
[11:52:55] <mhaberler> now to actually make that usable, the new instance needs to inherit the current world model (1) (or synch with it), AND get a disjunct execution state descriptor as to not trample on the other interpreter's internal state
[11:52:56] <mhaberler> that is currently impossible because world model and interp state are in a single structure - read as: yes, it's a class, but no, another instance doesnt give you an independent execution context
[11:52:58] <mhaberler> so world model and interp state need to be separated into different structs; world model=static member, state = instance attribute; this might mean extra parameters to many functions
[11:53:00] -!- HDB10 has quit [Quit: Ex-Chat]
[11:55:55] -!- ries [ries!~ries@] has joined #emc-devel
[11:58:41] <mhaberler> anywhere making sense yet ;-?
[12:07:16] <archivist> I think so and that seems a traumatic change
[12:10:09] <archivist> are you trying to add separate interpreters for n axes
[12:11:25] <archivist> or a detect worn tool stop pull out change return to previous state
[12:11:30] <mhaberler> not on axes; I want to startup a pristine execution context in the middle of a block execution - goal is to replace arbitrary codes by o-word sub calls but for now I limit myself to replacing Tx and M6
[12:12:45] <archivist> replace arbitrary codes by o-word sub calls seems a text replacement to the raw gcode is all thats needed
[12:17:25] -!- Dallur has quit [Quit: Leaving.]
[12:26:15] <mhaberler> that's what I thought, too, unfortunately it aint so
[12:26:17] <mhaberler> a 'block' is parsed into a block structure, and then that stuff is executed in a given order
[12:26:19] <mhaberler> that means that you need to be able to call a sub during a block execution
[12:26:20] <mhaberler> and that tramples on the state of the current block
[12:26:22] <mhaberler> which is why I want to separate out the state
[12:28:55] <mhaberler> the other show stopper here is that there's a limitation of one oword call per block
[12:29:16] <mhaberler> that I can get rid of by using a new interp instance
[12:30:44] <archivist> if you text replace, then you are adding blocks not clobbering the current one, still not sure the real application yet
[12:32:35] <archivist> yes some modal setting will be clobbered though
[12:34:13] -!- Valen has quit [Quit: Leaving.]
[12:50:02] -!- psha [psha!~psha@] has joined #emc-devel
[13:15:04] <mhaberler> that's the issue
[13:15:06] <mhaberler> modal settings are interp state, not part of the world model
[13:16:08] <mhaberler> the goal is to optionally replace the toolchange support (Tx,M6,M61) by o-word subroutines in g-code
[13:35:23] -!- skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.] has joined #emc-devel
[13:43:30] <jepler> mhaberler: I have no idea why _setup is static
[13:44:11] <mhaberler> well it works just fine with making it an instance variable
[13:45:01] <mhaberler> that's sort of the cheap solution
[13:45:03] <mhaberler> the clean solution would be to separate world model and interpreter state
[13:45:49] <mhaberler> it is major lobotomy though
[13:46:38] <jepler> Interp::synch is supposed to update anything that is external as opposed to interpreter state
[13:47:44] <jepler> unfortunately, I don't think you could call anyone on the project an expert on the interpreter structure :-/
[13:48:40] <mhaberler> re structure.. its a bit arcane, yes
[13:48:42] <mhaberler> re sync..right.. well either I instantiate a new interp with the current :_setup as parameter or interp::synch it, should be pretty much the same effect I guess
[13:49:26] <mhaberler> "pretty much the same" like "a bit pregnant"
[13:50:00] -!- elmo40 has quit [Ping timeout: 276 seconds]
[13:50:48] <mhaberler> cloning the setup would alleviate going through inifile, param restore, initial gcodes etc - basically just an independent throwaway copy of the previous interp
[13:52:05] -!- izua has quit [Ping timeout: 276 seconds]
[13:52:13] <psha> mhaberler: most of structures in _setup are defined 'in place' and not with pointers
[13:52:21] <psha> so plain 'memcpy' will move lot of stuff
[13:52:34] <jepler> use operator= to copy structures
[13:52:40] <psha> and you'll need to adjust only stack and friends
[13:52:51] <psha> ah, forgot that it's C++ :)
[13:53:07] <jepler> setup saved_state = _setup; ....; _setup = saved_state
[13:53:25] <jepler> if it becomes useful later to actually define an operator= that does something smarter than a bytewise copy, you can do it..
[13:54:10] <mhaberler> ah, good idea
[13:54:12] <mhaberler> stack: good point, I missed that one
[13:54:13] <mhaberler> I though I saw something about interpreter regression ..
[13:54:33] <jepler> there are interpreter tests
[13:54:37] <mhaberler> regression testing - any pointers? I think the saicanon stuff serves that, right?
[13:54:41] <jepler> you can run the bundled tests this way: . scripts/emc-environment; runtests
[13:54:49] <mhaberler> ah, thanks
[13:55:08] <jepler> yes, the interpreter tests use the 'rs274' program which is from src/emc/sai
[13:55:46] <jepler> look at an existing test like tests/interp/crazy-paths if you want to create your own that exercise toolchange
[13:56:15] <mhaberler> ok, I'll see wether I can get by making _setup an instance attribute (is this proper c++ speak? is it a 'member variable'? dunno) and save/restore that, that should be most of the milage
[13:56:36] <mhaberler> actually I'm pretty sure that would work
[13:56:38] <psha> mhaberler: no difference how to call it if others understand you ;)
[13:56:53] <jepler> I see one unfortunate thing about doing that: it seems to mean you have to include interp_internal.hh in rs274ngc_interp.hh
[13:57:02] <mhaberler> yes, did that already
[13:57:05] <jepler> because the definition of the setup struct has to be available then
[13:57:25] <mhaberler> or that goes to a separate include just for that purpose
[13:57:52] <mhaberler> it could be private anyway
[13:58:33] <psha> jepler: or use forward declaration
[13:59:00] -!- micges_work [micges_work!~x@aejb205.neoplus.adsl.tpnet.pl] has parted #emc-devel
[13:59:04] <mhaberler> doesnt have the size to be known?
[13:59:15] <jepler> 1 struct incomplete;
[13:59:15] <jepler> 2 struct wont_compile {
[13:59:15] <jepler> 3 incomplete i;
[13:59:15] <jepler> 4 };
[13:59:19] <jepler> incomplete.cc:3: error: field ‘i’ has incomplete type
[13:59:26] <psha> incomplete &i; incomplete *i;
[13:59:31] <jepler> you can't declare a struct or class member that is an incomplete type
[13:59:41] <mhaberler> been there..
[13:59:45] <jepler> you can declare a static class member that is incomplete..
[14:00:13] <mhaberler> aja, that's why that worked, and maybe that's the reason why it was declared static in the first place
[14:00:18] <jepler> well, it is not a showstopper to add the required include there
[14:00:33] <jepler> it just makes me a bit sad
[14:01:37] <mhaberler> well, lets see wether it works and then there will be some upside to counter them tears ;-)
[14:01:57] <jepler> it looks like you can't specify an inifile with the standalone rs274
[14:02:07] <jepler> that'll make it tough to test this feature from the testsuite :-/
[14:03:42] <mhaberler> hm, well I need to explore the regression test first
[14:03:43] <mhaberler> anyway, I gotta run for some unfortunate real-life stuff - thanks very much for the hints, it's fairly clear for me how to proceed!
[14:03:51] <jepler> best of luck
[14:04:01] <jepler> with emc and real life both
[14:04:11] <mhaberler> hm, I understand now why 'cradek let me do it'
[14:04:36] <mhaberler> I'm now at throwaway branch #5;-)
[14:04:51] <jepler> beats doing it in cvs
[14:05:06] <jepler> where you can never organize your thoughts without letting the central authority know
[14:05:10] * jepler bbls too
[14:05:18] <mhaberler> anytime, still its more meandering like software engineering
[14:05:36] <mhaberler> cu!
[14:05:37] <psha> jepler: i've send mail to dev list with a short summary of 'research' on docs format and it seem that all mentioned requirements may be done with asciidoc
[14:05:56] <jepler> psha: that's great, I've secretly wanted asciidoc all along
[14:06:07] <mhaberler> psha: thanks for doing that..
[14:06:12] <cradek> me too
[14:08:41] <psha> so i guess if migration is approved i'll start working on convertor
[14:10:28] -!- mhaberler has quit [Ping timeout: 246 seconds]
[14:23:50] <archivist> psha, and then... if I can get a list of anchors for the html !man g76 becomes possible from the bot
[14:24:20] <archivist> !man25 g76 to go to a particular version
[14:25:06] <archivist> it carries a root url to base a versions docs
[14:25:18] <archivist> table or root urls
[14:26:53] <psha> what exactly do you need? list of section names with URL's?
[14:28:20] <archivist> for mysql I read the xml and pull anchors and page titles and index terms, and give those a priority
[14:29:57] -!- drill has quit [Ping timeout: 240 seconds]
[14:30:32] <archivist> I then keep tables to create the replies, I keep the http://linuxcnc.orc/docs/2.5 separate from title , pagehtmlname
[14:31:45] <archivist> title being the term we are searching for, the taught items are for wiki type pages which dont have structure
[14:35:04] <archivist> so section header names as anchors, page titles, and any special index terms that should go to a particular page
[14:35:49] -!- ries has quit [Quit: ries]
[14:36:03] -!- ries [ries!~ries@] has joined #emc-devel
[14:36:55] <archivist> there is a bit of work I need to do to fix the !man system to become channel specific as its mysql only at the moment
[14:40:44] <psha> maybe docbook sources will help you?
[14:41:51] -!- nullie has quit [Read error: Connection timed out]
[14:41:54] <archivist> are you converting lyx to docbook or asciidoc to docbook as an output
[14:42:09] <psha> asciidoc to docbook as an intermidiate stage
[14:43:58] <archivist> I should learn xslt really
[14:44:09] <archivist> black art though
[14:46:04] <psha> but really it's possible to write small asciidoc config for custom output
[14:46:15] <psha> which would ignore everything except sections
[14:46:20] <jepler> hmmm ... make[2]: *** No rule to make target `/home/buildslave/BuildBot/slave/lucid-rtai-i386/lucid-i386-realtime-rip/build/src/hal/drivers/mesa-hostmot2/sserial.o', needed by `/home/buildslave/BuildBot/slave/lucid-rtai-i386/lucid-i386-realtime-rip/build/src/abs.o'. Stop.
[14:46:46] <psha> jepler: lyxtree converts LyX to XML 1 to 1?
[14:49:35] <jepler> I suspect there are some things lyxtree ignores or skips over.. so I would hesitate to call it 1 to 1
[14:53:53] <jepler> cradek, andypugh: it looks like the hostmot2 additions were incomplete.. sserial.c seems to be missing
[14:54:04] <jepler> buildbot encounters an error and so do I
[14:54:12] <seb_kuzminsky> yeah, the build failed, but the failure email didn't make it onto the list
[14:54:28] <seb_kuzminsky> the new buildmaster is not subscribed to the emc-commit list
[14:54:34] <seb_kuzminsky> i'll look at that later today
[14:54:36] <seb_kuzminsky> bbl
[14:55:32] <psha> jepler: but for conversion tasks it would be fine i guess?
[14:56:25] <jepler> psha: it obviously doesn't miss any feature you observe in the html documentation..
[14:57:51] <jepler> seb_kuzminsky: what address should I subscribe to the list? I got the list administration password since the last time we talked.
[14:58:18] <psha> jepler: yes, that's why i guess it's ok ;)
[14:59:11] -!- seb_kuzminsky has quit [Ping timeout: 255 seconds]
[15:00:08] -!- mozmck has quit [Quit: Leaving.]
[15:02:11] -!- mozmck [mozmck!~moses@client-] has joined #emc-devel
[15:08:50] <cradek> jepler: arggh
[15:10:19] <cradek> is just the rt build broken?
[15:10:26] <jepler> cradek: yes I think so
[15:10:30] <cradek> fff
[15:10:40] <jepler> that's hostmot2 and that only builds for rt
[15:10:55] <cradek> I don't have what it takes to fix it - should I revert?
[15:11:15] <cradek> or should we wait for andy?
[15:13:23] <jepler> I'd give him a bit of time to pop up and fix the problem
[15:18:49] <cradek> I need to set up realtime build on my laptop, even if it doesn't run.
[15:19:09] <jepler> yeah, you can build the realtime version without booting the realtime kernel
[15:19:14] <cradek> sure
[15:35:45] <psha> jepler: lyx holds everything in 'layout' blocks?
[15:36:01] <jepler> psha: yes, everything's a layout of one type or another
[15:36:14] <psha> thx
[15:38:19] -!- psha has quit [Read error: Connection reset by peer]
[15:40:25] -!- psha [psha!~psha@] has joined #emc-devel
[15:54:53] -!- mhaberler [mhaberler!~chatzilla@62-47-140-143.adsl.highway.telekom.at] has joined #emc-devel
[16:09:47] <psha> jepler: your converter instead of 'layout Description' generates '<descr>'
[16:09:50] <psha> but that's fine i guess
[16:14:24] -!- nullie has quit [Read error: Connection timed out]
[16:17:08] -!- dgarr [dgarr!~dgarrett@adsl-76-204-28-139.dsl.pltn13.sbcglobal.net] has joined #emc-devel
[16:20:16] <dgarr> small patch for twopass handling of comments: http://www.panix.com/~dgarrett/stuff/0001-twopass.tcl-ignore-hal-trailing-comments.patch
[16:22:10] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has joined #emc-devel
[16:29:29] <CIA-2> EMC: 03jepler 07master * r7a09df425e0d 10/tcl/twopass.tcl: twopass.tcl: ignore hal trailing comments
[16:30:04] -!- seb_kuzminsky [seb_kuzminsky!~seb@ucb-np2-254.colorado.edu] has joined #emc-devel
[16:33:11] <dgarr> thanks:)
[16:33:16] <jepler> welcome
[16:35:25] <psha> jepler: you lyxtree simplifies conversion a lot ;)
[16:39:13] -!- micges [micges!~micges@enw236.neoplus.adsl.tpnet.pl] has joined #emc-devel
[16:40:01] <psha> heh, there are _lot_ of places where comma is inside quotes :)
[16:40:06] <psha> but i guess it's ok
[16:41:04] <micges> logger[psha]: czesc
[16:44:29] -!- mhaberler has quit [Ping timeout: 255 seconds]
[16:44:47] -!- milestone_ has quit [Client Quit]
[16:47:14] -!- seb_kuzminsky has quit [Ping timeout: 240 seconds]
[16:52:07] -!- wobblybootie has quit [Client Quit]
[16:57:28] <psha> jepler: http://psha.org.ru/tmp/src/machining_center.html
[16:57:32] <psha> without manual intervention
[17:07:14] <psha> http://psha.org.ru/tmp/src/lathe-user.html
[17:07:24] <psha> with itemize and enumerations
[17:08:55] -!- awallin_ has quit [Ping timeout: 260 seconds]
[17:10:42] -!- dgarr has quit [Quit: Leaving.]
[17:16:55] -!- isssy has quit [Quit: Visitor from www.linuxcnc.org]
[17:18:40] -!- nullie has quit [Quit: Ex-Chat]
[17:22:34] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has parted #emc-devel
[17:28:48] <jepler> psha: you wrote the conversion from lyxtree xml to docbook that quickly!?
[17:28:54] <jepler> er, to asciidoc
[17:33:04] <psha> yes
[17:33:14] <psha> thanks to asciidoc simplicity ;)
[17:34:01] <jepler> I tried to view the .txt to see how natural or unnatural it looks, but it's not there
[17:34:14] <psha> wait a bit - i'm uploading whole conversion
[17:34:18] <jepler> ok
[17:37:09] <psha> http://psha.org.ru/tmp/src/
[17:37:16] <psha> mirror of docs/src/
[17:37:55] <jepler> can you wrap all the paragraphs at 72 characters or so?
[17:41:14] <psha> hm, i'll see
[17:41:41] <psha> i guess with a small risk of breaking anchors, xrefs and friends it would be simple
[18:03:06] <psha> done
[18:03:17] <psha> hope without errors ;)
[18:03:39] <psha> rebuilding html docs to check that
[18:09:21] <psha> http://psha.org.ru/tmp/src/gcode/main.html#_g33_spindle_synchronized_motion_g33_g33_1_spindle_synchronized
[18:09:35] <psha> at least one place where 72-width split is broken
[18:19:23] -!- pcw_home has quit [Remote host closed the connection]
[18:21:17] <psha> fixed code
[18:30:03] <psha> fixed inline images
[18:30:58] <psha> hm
[18:32:05] <psha> http://www.linuxcnc.org/docs/devel/html/gui_axis.html#r1_7
[18:32:23] <psha> axis-remote is really using three dashes for params? (e.g ---ping)
[18:47:58] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc-devel
[18:53:06] -!- nullie has quit [Quit: Ex-Chat]
[18:57:27] -!- adb has quit [Ping timeout: 240 seconds]
[18:58:44] <psha> jepler: what's lyx 'shape' tag?
[19:01:05] -!- emcrules_mobile has quit [Quit: Chatmosphere - IRC for Blackberry]
[19:01:34] -!- adb [adb!~Moldovean@] has joined #emc-devel
[19:03:36] <jepler> psha: with values like default, smallcaps, italic I think it sets font variants
[19:03:46] <jepler> not sure what "\shape up" is
[19:05:12] <psha> hm, tx
[19:06:32] -!- ries has quit [Quit: ries]
[19:09:10] <andypugh> To save any trouble with incompatibility with the "old" gearchange.comp, how about calling the "new" one just "gear"?
[19:09:40] <andypugh> (I am generally in favour of anything to reduce typing in HAL files)
[19:15:03] <jepler> andypugh: after your recent hostmot2 changes, master doesn't build. can you look into it and get another patch to apply on top of all the rest to get things working again?
[19:15:39] <andypugh> Oh.
[19:15:44] <jepler> make[2]: *** No rule to make target `/home/buildslave/BuildBot/slave/lucid-rtai-i386/lucid-i386-realtime-rip/build/src/hal/drivers/mesa-hostmot2/sserial.o',
[19:17:10] <andypugh> I think I see the problem
[19:17:30] <andypugh> sserial.c isn't in the patch file.
[19:17:39] <jepler> yeah that would do it
[19:18:19] <jepler> psha: no, axis-remote uses two dashes just like any sane gnuish program
[19:18:57] <jepler> looks like the LyX source is wrong
[19:19:04] <jepler> making AXIS exit (
[19:19:04] <jepler> \family typewriter
[19:19:04] <jepler> -\SpecialChar \-
[19:19:04] <jepler> -quit
[19:19:04] <jepler> \family default
[19:19:07] <jepler> ).
[19:31:07] <psha> hm, i've found at least one place where asciidoc is inferiour to lyx...
[19:31:40] <jepler> andypugh: when do you think you'll be able to supply a patch that adds sserial.c?
[19:31:41] <psha> http://www.linuxcnc.org/docs/devel/html/hal_comp.html#r1_5
[19:31:51] <psha> description of DOC
[19:32:57] <jepler> I see that the asciidoc html is cut off...
[19:33:09] <psha> update
[19:34:40] <jepler> so you can't nest certain things inside a definition list or whatever asciidoc calls them?
[19:35:14] <psha> yes
[19:35:27] <jepler> :(
[19:35:27] <psha> some thinks (like enums, itemizations) are nestable
[19:35:40] <psha> others - like defs list are not
[19:35:47] <psha> that's pity...
[19:36:28] -!- bootnecklad has quit [Ping timeout: 255 seconds]
[19:37:12] <psha> i'll dig a bit more - maybe i've missed something
[19:38:30] <psha> ah
[19:38:36] <psha> it's nestable
[19:39:22] -!- ries [ries!~ries@] has joined #emc-devel
[19:42:21] <psha> http://psha.org.ru/tmp/src/hal/comp.html#_syntax
[19:43:00] <psha> scroll down to OPT, VALUE
[19:44:07] -!- pjm has quit [Ping timeout: 255 seconds]
[19:48:02] <andypugh> jepler: I am working on it. I need to persuade git to be nice to me first
[19:51:28] <andypugh> OK, should I send it to the list, or do you want it?
[19:51:36] <jepler> privately is fine
[19:52:40] <andypugh> Trying to send it now, unless you prefer email?
[19:52:58] <psha> hal/comp.lyx looks like excersises in different font options...
[19:54:05] <jepler> it's been a long while since I used dcc
[19:54:07] <andypugh> jepler: Can you see if it fixes things locally first? My confidence in my ability to prodice working patch files is very, very low.
[19:54:30] <jepler> the build runs to completion
[19:54:54] <andypugh> That's a start, at least
[19:55:11] <andypugh> I will pull a new master and check for function
[19:55:30] <jepler> thanks
[19:55:42] <CIA-2> EMC: 03jepler 07master * r50bd2c1ce86c 10/src/hal/drivers/mesa-hostmot2/sserial.c: missing file: sserial.c
[19:55:44] <jepler> …pushed
[19:57:47] <jepler> if it makes you feel any better, at my day job we use git and I recently broke the development tree by making the same error (not committing a newly-created file)
[20:04:15] <andypugh> The odd thing is that cradek noticed a problem with the bldc.comp patch, but not with the main one, before he pushed them for me.
[20:04:36] <jepler> he built on sim only
[20:04:47] <jepler> so a hardware driver like hostmot2 wasn't built at all
[20:04:52] <andypugh> Ah, yes, that would do it.
[20:05:46] <cradek> sigh
[20:09:42] <andypugh> How can I get back to a normal, clean, master branch? I can't pull at the moment because of:
[20:09:43] <andypugh> error: Untracked working tree file 'configs/common/shuttlexpress.hal' would be overwritten by merge. Aborting
[20:10:07] <andypugh> I am tempted to just delete my whole git repo and start again. It's a mess
[20:10:55] <andypugh> (Applying patches in the wrong level of the tree hasn't helped at all)
[20:11:29] -!- pjm__ has quit [Ping timeout: 264 seconds]
[20:15:01] -!- Dallur has quit [Quit: Leaving.]
[20:22:16] <psha> git reset --hard
[20:22:18] <psha> git clean -fd
[20:22:28] <psha> but that'll kill _everything_
[20:22:35] <andypugh> That's OK
[20:22:56] <psha> first resets all known files to HEAD
[20:23:04] <psha> second removes all unknown files and dirs
[20:23:19] <andypugh> Perfect
[20:23:29] <andypugh> Now to delete all my branches
[20:23:41] <psha> git branch -d branch-name
[20:23:47] <psha> or -D if it's not merged
[20:24:05] <andypugh> Ah, that last bit is what I have been missing
[20:26:41] <andypugh> Still can't pull master.
[20:26:54] <andypugh> error: Untracked working tree file 'lib/python/gladevcp/hal_mdihistory.py' would be overwritten by merge. Aborting
[20:28:16] <psha> ah
[20:28:20] <psha> it don't respect .gitignores
[20:28:39] <andypugh> I am sure that git is briliant, if you can get your head around how it works. It doesn't appear that I can.
[20:28:39] <psha> you have to kill that file by hand
[20:28:54] <andypugh> I wonder how many others?
[20:29:07] <psha> you applied some patches (but not commited them)
[20:29:17] <psha> but files from them are ignored (by gitignore)
[20:29:31] <psha> git clean -fdx
[20:29:48] <psha> -x remove ignored files, too
[20:32:02] <psha> jepler: i've managed to fix nearly everything except nested verbatim blocks
[20:32:10] <andypugh> OK, so this is my own fault at least:
[20:32:11] <andypugh> Auto-merging src/hal/components/bldc.comp
[20:32:11] <andypugh> CONFLICT (add/add): Merge conflict in src/hal/components/bldc.comp
[20:32:11] <andypugh> Automatic merge failed; fix conflicts and then commit the result.
[20:33:24] <psha> you've got two copies of your patches
[20:33:27] <psha> one sitting in your repo
[20:33:36] <psha> second applied by criss&jeff
[20:33:50] <andypugh> Probably.
[20:33:52] <psha> one possible way is to use 'rebase' instaead of merge
[20:34:01] <psha> git rebase -i origin/master
[20:34:13] <andypugh> Or I could just re-clone the repo
[20:34:16] <psha> then mark patches you don't need
[20:34:21] <andypugh> Or nuke the site from orbit
[20:34:24] <psha> or just drop this branch ;)
[20:34:35] <psha> git branch -M tmp
[20:34:40] <psha> git checkout -b master origin/master
[20:34:43] <psha> git branch -D tmp
[20:35:49] <andypugh> Aha!
[20:36:13] <andypugh> I even understood what you were doing there
[20:36:44] <psha> that's how i do my gladevcp branches
[20:37:07] -!- toastydeath has quit [Quit: Leaving]
[20:37:09] <psha> prepare lot of feature branches, then merge them in one, ask jeff to pick it and then drop
[20:39:12] <andypugh> My main problem is a tendency to fiddle around with something to see if it works, then realise that I didn't create a banch for it, so it is all mixed up with something else.
[20:42:08] <psha> you may branch whenever you want
[20:42:15] <psha> it's not mercurial with commit-bounded branches
[20:42:26] -!- ries has quit [Ping timeout: 240 seconds]
[20:42:48] <psha> for example if you are in situation where you changed two files (with two different features) and want to branch only one
[20:42:53] <psha> run git stash -p
[20:43:13] <psha> it will ask you what changes you want to stash - select features you don't want to branch
[20:43:24] <psha> then git checkout -b cool-feature
[20:43:28] <psha> commit there
[20:43:31] <psha> checkout back
[20:43:33] <psha> and git stash pop
[20:43:59] <psha> so you'll split you large change into smaller and in different branches
[20:44:28] <andypugh> Clever. And baffling.
[20:45:15] <psha> like any tool - like mobile phone, hammer or even spoon, you have to explore it's possiblities first...
[20:45:38] <psha> for example i've used my spoon several years ago to dig to the water ;)
[20:46:57] <andypugh> I am about to install Autodesk Inventor on my Mac in a Virtual Machine. I can see that working poorly :-)
[20:50:45] <psha> jepler: what's "ert" tag?
[20:51:13] <jepler> psha: "The acronym ERT (Evil Red Text) refers to raw LaTeX commands inserted into the body of your LyX document"
[20:54:23] <jepler> for instance, \clearpage (go to next page) is an ERT, and it is just omitted in the html translation
[20:54:54] <jepler> <xsl:template match="ert"/>
[20:55:06] <jepler> yeah, that's the xsl rule for ERTs .. they are just ignored by the html conversion
[20:56:31] <psha> :D
[20:56:39] <psha> so i'll do the same...
[20:56:42] <psha> another question
[20:56:47] <psha> http://www.linuxcnc.org/docs/devel/html/hal_components.html#sec:Commands-and-Userspace-Components
[20:56:54] <psha> here is strange itemize
[20:56:57] <psha> with bullets and =
[20:57:06] <psha> in LyX xml it looks like
[20:57:11] <psha> <item><labelwidthstring data="00.00.0000"/>= AXIS Remote Interface</item>
[20:57:19] <psha> i guess it's safe to ignore labelwidthstring
[20:57:26] <psha> since it's all zeroes ;)
[20:58:11] <jepler> huh
[20:59:10] <psha> hm
[20:59:14] <psha> another dumb
[20:59:15] <psha> one
[20:59:24] <psha> i've seen several 'pagebreak_bottom' tags
[20:59:39] <psha> but i guess they are not really intended to be 'hard page breaks'
[20:59:48] <psha> <layout class="Standard">Make sure the loadusr line is before any net's that make use of the pyVCP<pagebreak_bottom/> pins.</layout>
[21:00:03] <psha> forcing page break with one word left....
[21:00:04] <psha> ;)
[21:01:33] <jepler> huh, I have no idea what's gong on with that stuff in components.lyx
[21:04:34] <jepler> well, it's function ListFixer in lyxtree.py that loses the first word
[21:05:42] <jepler> \layout List
[21:05:42] <jepler> \labelwidthstring 00.00.0000
[21:05:42] <jepler> 1.
[21:05:42] <jepler> Widgets de signalisation, signaux BIT: led, rectled
[21:06:04] <jepler> in the files I tested, I think that the first word was the index in the list, which you don't want to include in the output...
[21:06:06] <psha> i guess it's safe to ignore it
[21:07:46] -!- ries [ries!~ries@] has joined #emc-devel
[21:07:57] <jepler> probably comment out the call to ListFixer for the purposes of confusion
[21:08:06] <jepler> er, for the purposes of *conversion*
[21:10:42] <psha> another one
[21:10:44] <psha> http://www.linuxcnc.org/docs/devel/html/gui_mini.html#r1_4_1
[21:10:52] <psha> here are many 'Quotations'
[21:11:05] <psha> indented blocks with Quotation layout class
[21:11:12] <psha> how to deal with them?
[21:11:22] <psha> treat them as asciidoc quotations?
[21:11:24] <jepler> in this case? it looks like it's not even desired
[21:12:05] <psha> there is footnote
[21:12:06] <psha> 1 Much of this chapter quotes from a chapater of the Sherline CNC operators manual. back
[21:12:10] <psha> so maybe it's really quotation
[21:13:43] <jepler> If the text is free to incorporate into our documentation because the sherline manual is also GFDL then we should just incorporate it.
[21:14:02] <jepler> if we're not free to incorporate it willy-nilly, I'd rather the text jettisoned and something original written in its place
[21:14:39] <psha> ok, so i'd drop quotation after conversion
[21:15:51] -!- ries has quit [Ping timeout: 240 seconds]
[21:16:24] <psha> http://www.linuxcnc.org/docs/devel/html/motion_tweaking_steppers.html#r1_1_2
[21:16:25] <psha> same here
[21:17:10] <jepler> here I think using a block quotation is sensible
[21:17:30] <psha> there are 4 files with Quotation layouts
[21:17:32] <jepler> (well, except that "For exmaple, the Gecko G202 manual says this:" shouldn't be inside the block)
[21:17:43] <psha> let me convert them to quotes and then do manual cleanup
[21:17:45] -!- ries [ries!~ries@] has joined #emc-devel
[21:20:07] <psha> so now i've covered all tags (at least converter don't complain about something it don't know)
[21:20:28] <psha> i'm running rebuild and uploading in a minutes
[21:21:29] <andypugh> jepler: A fresh pull and make seems to run the 8i20 and 7i64 as expected.
[21:22:40] -!- ries has quit [Ping timeout: 260 seconds]
[21:23:59] -!- ries [ries!~ries@] has joined #emc-devel
[21:37:27] -!- ries has quit [Read error: Connection reset by peer]
[21:39:01] -!- ries [ries!~ries@] has joined #emc-devel
[21:57:15] -!- seb_kuzminsky [seb_kuzminsky!~seb@174-16-125-123.hlrn.qwest.net] has joined #emc-devel
[22:00:21] -!- micges has quit [Quit: Ex-Chat]
[22:03:16] -!- LawrenceG [LawrenceG!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc-devel
[22:03:55] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc-devel
[22:04:11] -!- LawrenceG [LawrenceG!~LawrenceG@S0106009027972e37.pk.shawcable.net] has parted #emc-devel
[22:05:58] -!- psha has quit [Ping timeout: 272 seconds]
[22:06:15] -!- psha [psha!~psha@] has joined #emc-devel
[22:06:28] -!- ve7it has quit [Client Quit]
[22:06:45] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc-devel
[22:09:55] ve7it is now known as gonso
[22:12:31] gonso is now known as ve7it
[22:12:49] <psha> is there Master_User or Master_Integrator HTML docs?
[22:12:50] <psha> full?
[22:13:32] <andypugh> I have a cleaned-up version of gearchange. Now I have decided to scrap compatibility, i have changed a few pin names too.
[22:14:21] <andypugh> The only remaining question is whether to keep the name, or rename to "gear" so that it can coexist with the old version for a while.
[22:19:53] -!- odiug has quit [Quit: Leaving.]
[22:20:35] <psha> hm
[22:20:39] <psha> i think i'm done
[22:21:45] <skunkworks> psha: Thanks for working on this stuff!
[22:22:27] <psha> so now i need your eyes :)
[22:22:29] <psha> http://psha.org.ru/tmp/src/Master_Integrator.html
[22:22:33] <psha> http://psha.org.ru/tmp/src/Master_HAL.html
[22:22:33] <andypugh> Yes, indeed.
[22:22:37] <psha> http://psha.org.ru/tmp/src/Master_User.html
[22:23:07] <psha> theese are converted docuements for Integrator manual, User manual and HAL
[22:23:18] <psha> there are separate pages living down in that tree
[22:24:32] <psha> i've read them through to verify that at least there are no major bugs
[22:24:44] <psha> but that's not enought
[22:24:51] <psha> archivist: re index terms
[22:24:54] <psha> archivist: http://psha.org.ru/tmp/src/Master_User.html#_index
[22:24:59] <psha> this would be enought?
[22:27:49] <psha> there are bunch of formatting errors realted to indentation of first paragraph lines
[22:29:42] <andypugh> These are converted from existing docs?
[22:29:58] <psha> yes
[22:30:07] <psha> with minor manual intervention
[22:30:17] <psha> since it complains about empty sections
[22:30:20] <andypugh> Hm, 23.2.1 seems to consistently use "to" for "too"
[22:31:26] <psha> maybe it's already fixed but since this conversion is based on ~2-3 weeks ago checkout it's not legal to fix it now ;)
[22:32:50] -!- pjm has quit [Ping timeout: 240 seconds]
[22:34:36] elmo401 is now known as elmo40
[22:35:34] -!- ve7it has quit [Quit: Leaving]
[22:35:47] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc-devel
[22:36:25] -!- LawrenceG [LawrenceG!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc-devel
[22:36:45] -!- ve7it has quit [Client Quit]
[22:36:46] -!- LawrenceG has quit [Client Quit]
[22:37:00] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc-devel
[22:37:15] -!- garage_seb [garage_seb!~seb@174-16-125-123.hlrn.qwest.net] has joined #emc-devel
[22:42:55] -!- skunkworks has quit [Ping timeout: 265 seconds]
[22:43:41] <garage_seb> basic wiring question from a helpless software geek...
[22:43:53] <garage_seb> i'm wiring up a bunch of buttons on the control panel on my mill
[22:44:12] <garage_seb> it's about a 6' cable run to the 7i37 isolated i/o board in the electronics cabinet
[22:44:25] <garage_seb> i'm using cat5 cables (utp) because that's what i have handy
[22:45:06] <garage_seb> question: is it ok to run just one ground line from the back to the front, then spider it out to all the buttons on the panel?
[22:45:41] <cradek> 24v?
[22:45:48] <garage_seb> cradek, yes
[22:45:56] <andypugh> You mean 0V or ground/shield?
[22:45:56] <garage_seb> the buttons are momentary, and i'm thinking each gets its own signal wire but a shared ground
[22:46:07] <garage_seb> it's logic ground
[22:46:15] <cradek> probably deosn't really matter...
[22:46:16] <garage_seb> not the wire shield
[22:46:37] <andypugh> It'll be fine, and if not, fix it in software :-)
[22:46:54] <garage_seb> yeah i was thinking it'd be safe to do, but i didn't have a lot of confidence in that thought :-)
[22:47:00] <garage_seb> ok next question
[22:47:18] <garage_seb> each wire consists of four twisted pairs
[22:47:21] <cradek> I'd probably run a pair to each button, to make it all nice though
[22:47:48] <garage_seb> is it ok to run different buttons on the two wires of a pair?
[22:48:01] <garage_seb> or should i just not try to optimize my wire count and run a pair to each button?
[22:48:03] -!- Valen has quit [Quit: Leaving.]
[22:48:07] <cradek> it's just buttons...
[22:48:09] <andypugh> It's a DC signal, so I would say yes.
[22:48:10] <cradek> brb
[22:48:27] <garage_seb> ok thanks guys :-)
[22:48:39] <psha> garage_seb: if you don't want to control stepper with button presses then what's problem?
[22:49:10] -!- ries has quit [Ping timeout: 240 seconds]
[22:53:03] -!- Fox_Muldr has quit [Ping timeout: 240 seconds]
[23:05:51] -!- e3m has quit [Quit: Do fish get thirsty?]
[23:08:10] -!- ries [ries!~ries@] has joined #emc-devel
[23:10:43] -!- skunkworks [skunkworks!~chatzilla@str-bb-cable-south2-static-6-412.dsl.airstreamcomm.net] has joined #emc-devel
[23:14:08] -!- andypugh has quit [Quit: andypugh]
[23:24:49] <psha> jepler: i've created branch for docs conversion - please take a look on it
[23:24:56] <psha> i've not hooked into Submakefile yet
[23:24:58] <psha> http://psha.org.ru/cgit/psha/emc2.git/log/?h=asciidoc
[23:31:47] <psha> amended last patch
[23:51:37] -!- mrsunshine_ has quit [Ping timeout: 265 seconds]
[23:53:03] <psha> bb all
[23:53:09] -!- psha has quit [Quit: leaving]
[23:54:42] -!- fatpandas has quit [Quit: leaving]