#linuxcnc-devel | Logs for 2012-02-02

[00:03:44] -!- sumpfralle has quit [Read error: Connection reset by peer]
[00:04:01] -!- theorbtwo has quit [Ping timeout: 252 seconds]
[00:04:17] theorb is now known as theorbtwo
[00:24:55] -!- Nick001-Shop has quit [Quit: ChatZilla 0.9.88 [Firefox 3.6.24/20111107173218]]
[00:29:05] -!- WalterN has quit [Quit: Leaving]
[00:35:22] -!- rob_h has quit [Quit: Leaving]
[00:36:15] -!- zlog has quit [Remote host closed the connection]
[00:36:27] -!- Tom_itx has quit []
[00:48:48] -!- Loetmichel has quit [Ping timeout: 244 seconds]
[00:55:56] -!- sumpfralle1 has quit [Ping timeout: 245 seconds]
[01:11:17] -!- FinboySlick has quit [Ping timeout: 248 seconds]
[01:18:31] <cmorley> mhaberier: sounds like you are talking of themes. Themes have a gtkrc file and you can change almost everything in them. May be a bit heavy handed though. GladeVCP already handles themes.
[01:40:53] -!- WalterN has quit [Read error: Connection reset by peer]
[01:52:04] <CIA-125> 03mhaberler 07master * r71441b0790d3 10/src/hal/user_comps/gladevcp.py: gladevcp: silence informational messages, use stderr for errors
[01:52:05] <CIA-125> 03mhaberler 07master * rf62708319375 10/ (docs/man/man1/gladevcp.1 src/hal/user_comps/gladevcp.py): gladevcp: handle ~/.gtkrc-2.0 files properly, enable custom gtkrc file
[01:53:09] <mhaberler> cmorley: yes, I found out.. a workaround is needed to make it honor widget and widget_class matches - it does now
[01:53:36] <mhaberler> the reason for the heavy hands was that bug, its not necessary anymore
[02:09:02] <jepler> has anybody used xlinuxcnc lately? I can't actually run a program with it, because the home key doesn't do anything and the home button is greyed out...
[02:09:39] <jepler> er, wait, the home button is fine when I'm not in estop :-/
[02:09:59] <cradek> dude
[02:10:45] <jepler> I wonder why the home key doesn't work
[02:11:17] <jepler> we should make a list of "I don't think anybody's maintaining this" and post a plea for maintainers
[02:11:17] <cradek> I'm sure it used to...
[02:11:26] <jepler> with the threat that we're going to remove it in the release after 2.5 if nobody does
[02:11:34] <cradek> does it even have abc axis support?
[02:11:47] <cradek> there's not maintained, and there's positively old
[02:12:03] <jepler> no idea
[02:12:29] <jepler> I think maybe you have to run yemc to get 6 axis support
[02:12:32] <jepler> or maybe that was just a nightmare
[02:13:04] <jepler> I was running it just to remind myself how well early-90s toolkit programs work over the internet
[02:13:04] <cradek> we should copy it a second time to zemc, and add uvw then
[02:13:30] <jepler> actually I was surprised at how badly the coordinate display flickered
[02:13:31] <cradek> oh yeah, they worked fine through dialup - I bet they're perfectly good now.
[02:13:40] <jepler> I was caught in yet another wayland thread
[02:13:57] <cradek> oh I forgot about that insanity
[02:14:02] <jepler> "they'll do network transparency some weekend. it's so easy. just screen scrape it, compress it with some video codec, and shove it in the internet"
[02:14:16] <jepler> "it's easy, stop clinging to your guns^WX server"
[02:14:45] <jepler> and they rightly point out that all the modern toolkits are just shoveling bitmaps over X anyway, since that's plenty fast locally
[02:16:22] <cradek> uh so they're going to rewrite vnc?
[02:17:48] -!- drizt [drizt!~drizt@n44.jabber.d6.hsdnsrv.net] has joined #linuxcnc-devel
[02:18:42] <jepler> it sounds like their schemes for network transparency are (pick all of the following): (0) run an X server AND a Wayland server; (B) shove bitmaps down the network; or (iv) everyone uses web apps these days, don't they?
[02:19:36] <drizt> hi guys. fix src/Makefile. use tabs instead of spaces. 969, 970 and 971 lines
[02:20:18] <jepler> fix has been committed to v2.5_branch. thank you for your report.
[02:20:18] <jepler> 2b0cff8 use tabs instead of spaces in makefiles
[02:20:46] <drizt> okey
[02:21:31] <drizt> I have looked the master branch.
[02:22:23] <drizt> sorry for reteating.
[02:22:35] <jepler> in our workflow, we only merge v2.5_branch into master branch from time to time, not after every commit. You can merge it yourself locally: git merge origin/v2.5_branch
[02:22:41] -!- ries has quit [Quit: ries]
[02:22:50] <jepler> at the moment this merge does not result in any conflicts, so it is very easy
[02:23:17] <cradek> after merging, looks like jepler will have to fix the rest of master's Makefile
[02:23:25] <jepler> oh is there even more in master?
[02:23:25] <jepler> sigh
[02:23:29] <cradek> loads
[02:25:09] <jepler> anybody know why gnu make on ubuntu and debian don't diagnose this error?
[02:27:09] -!- ries has quit [Ping timeout: 248 seconds]
[02:27:24] <cradek> does gnu make on some other architecture?
[02:27:43] <jepler> I think drizt reported that he's on redhat or fedora or something?
[02:27:51] <jepler> does that count as a different architecture from debian these days?
[02:27:53] <drizt> Fedora 15
[02:28:06] <cradek> wow, fedora has large integers now
[02:29:12] <drizt> cradek: Did you say about fedora version?
[02:30:49] <jepler> I'll merge v2.5_branch and fix the Makefile in master shortly
[02:30:56] -!- cmorley has quit [Ping timeout: 245 seconds]
[02:31:20] <drizt> jepler: i will use v2.5_branch
[02:32:47] <drizt> I want to ask why emc uses such workflow and non-standard building rules? First I was confused with it.
[02:34:44] <cradek> that kind of question generally has no answer. can you ask something more specific?
[02:35:23] -!- cmorley [cmorley!~chris@] has joined #linuxcnc-devel
[02:38:16] -!- atom1 has quit [Quit: Leaving]
[02:38:38] <drizt> ussually ./configure script is in project_folder not in project_folder/src. Also emc2 has ready Makefile but ussualy Makefile is generating with helping ./configure script.
[02:40:23] <cradek> ah yes. we don't use automake, because the person who set up the build system thinks automake sucks.
[02:40:50] <drizt> ))) very funny!
[02:41:42] <cradek> I think at the time it only knew how to generate horrifying recursive Makefiles
[02:41:46] <cradek> (hopefully it's fixed by now)
[02:43:07] <cradek> yes perhaps they fixed that around 2006?
[02:43:37] <jepler> to put it another way: we use the current system because it works for us. Yes, it may exhibit a high amount of NIH, but I wouldn't want to switch to something else without it having specific advantages (such as a faster build) and no signficant downsides.
[02:43:54] <jepler> (one incremental change we should make is to avoid generating dependency information for most files as a separate pass...)
[02:45:40] <jepler> neat, I'd never used 'git rebase -p' before. -p means "preserve merges". And it did.
[02:46:44] <CIA-125> 03jepler 07master * r57bcc9a41b4c 10/ (22 files in 13 dirs): Merge remote branch 'origin/v2.5_branch'
[02:46:45] <CIA-125> 03jepler 07master * r8fa134431bd3 10/src/Makefile: use tabs instead of spaces in makefiles
[02:47:07] <cradek> git still surprises me. that is sometimes good.
[02:49:48] -!- dgarr [dgarr!~dgarrett@adsl-75-61-74-108.dsl.pltn13.sbcglobal.net] has joined #linuxcnc-devel
[02:51:35] <dgarr> a change to haltcl (use -i not -ini to match halcmd) and manpage updates for halrun,halcmd.haltcl(new) for consideration (master):
[02:51:36] <dgarr> http://www.panix.com/~dgarrett/stuff/haltcl,halcmd,halrun_update.mbox
[02:52:53] <jepler> has haltcl been in any released version?
[02:53:13] <dgarr> i know it is in 2.5, not sure about 2.4
[02:53:18] <jepler> if so, you have to preserve compatiblity: you can add -i but you can't lose -ini
[02:53:21] <jepler> I'll check if it's in 2.4.. jas
[02:53:41] <jepler> yes, haltcl.in was added before v2.4.0-pre1
[02:54:01] <jepler> I think it's fairly easy to support both
[02:54:13] <dgarr> i'll revise
[02:54:17] <jepler> thanks!
[02:55:44] <jepler> and thanks especially for the doc work
[03:00:12] <jepler> the rest looks good .. let me know when the revised patch is up
[03:08:05] <dgarr> revised: http://www.panix.com/~dgarrett/stuff/v2_haltcl,halcmd,halrun_update.mbox
[03:10:41] <jepler> thanks
[03:10:44] <CIA-125> 03jepler 07master * rf10b9312aaf6 10/docs/man/man1/ (halcmd.1 halrun.1): halcmd,halrun man pages
[03:10:45] <CIA-125> 03jepler 07master * re5d1fdf044bd 10/docs/man/man1/haltcl.1: haltcl.1 man page (new)
[03:10:46] <CIA-125> 03jepler 07master * raa36da227ef5 10/scripts/ (halrun.in haltcl.in linuxcnc.in): haltcl: allow use of -i or -ini to be more like halcmd
[03:10:47] <CIA-125> 03jepler 07master * r72cb5b818f59 10/docs/man/man1/halcmd.1: halcmd.1 manpage update
[03:16:34] -!- drizt [drizt!~drizt@n44.jabber.d6.hsdnsrv.net] has parted #linuxcnc-devel
[03:17:18] -!- cmorley has quit [Ping timeout: 252 seconds]
[03:17:59] <dgarr> welcome
[03:18:01] <jepler> 'night folks
[03:35:07] -!- cmorley [cmorley!~chris@] has joined #linuxcnc-devel
[04:47:40] -!- WalterN has quit [Quit: Leaving]
[05:06:13] <mhaberler> jthornton: see forum
[05:06:21] <mhaberler> good night
[05:06:23] -!- mhaberler has quit [Quit: mhaberler]
[05:10:13] -!- ve7it has quit [Remote host closed the connection]
[05:12:27] -!- Valen has quit [Quit: Leaving.]
[05:41:15] -!- dgarr has quit [Remote host closed the connection]
[05:49:13] -!- cstop has quit [Quit: Leaving]
[06:19:03] -!- fatpandas has quit [Ping timeout: 255 seconds]
[07:04:21] -!- linuxcnc-build has quit [Ping timeout: 248 seconds]
[07:04:34] -!- hm2-buildmaster has quit [Ping timeout: 276 seconds]
[07:04:57] -!- seb_kuzminsky has quit [Ping timeout: 252 seconds]
[07:11:19] -!- vladimirek [vladimirek!~vladimire@bband-dyn226.178-40-36.t-com.sk] has joined #linuxcnc-devel
[07:39:24] -!- vladimirek has quit [Remote host closed the connection]
[07:46:12] -!- psha[work] [psha[work]!~psha@] has joined #linuxcnc-devel
[07:55:11] -!- x_cnc_x has quit [Quit: Bye Bye]
[08:10:49] -!- AndChat- [AndChat-!~Seb@] has joined #linuxcnc-devel
[08:11:04] -!- GoSebGo has quit [Ping timeout: 244 seconds]
[08:29:01] -!- seb_kuzminsky [seb_kuzminsky!~seb@97-122-122-77.hlrn.qwest.net] has joined #linuxcnc-devel
[08:30:07] -!- e-ndy [e-ndy!jkastner@nat/redhat/x-dlahtwjmczbxvzmk] has joined #linuxcnc-devel
[08:33:50] -!- WalterN has quit [Read error: Connection reset by peer]
[08:33:52] -!- seb_kuzminsky has quit [Ping timeout: 245 seconds]
[08:52:25] -!- seb_kuzminsky [seb_kuzminsky!~seb@97-122-119-19.hlrn.qwest.net] has joined #linuxcnc-devel
[09:30:15] Loetmichel_ is now known as Loetmichel
[09:36:49] -!- AndChat- has quit [Ping timeout: 240 seconds]
[09:38:10] -!- sumpfralle has quit [Read error: Operation timed out]
[10:04:42] -!- fragalot has quit [Ping timeout: 245 seconds]
[10:06:19] fragalot is now known as Guest3403
[10:17:51] -!- maximilian_h [maximilian_h!~bonsai@] has joined #linuxcnc-devel
[10:20:52] -!- davec_ has quit [Ping timeout: 276 seconds]
[10:24:18] -!- mhaberler [mhaberler!~mhaberler@] has joined #linuxcnc-devel
[10:52:46] -!- mk0 has quit [Ping timeout: 260 seconds]
[10:59:44] -!- Geissler has quit [Ping timeout: 245 seconds]
[11:03:12] -!- GoSebGo [GoSebGo!~Seb@] has joined #linuxcnc-devel
[11:34:09] -!- WalterN has quit [Read error: Connection reset by peer]
[11:34:09] -!- H264 has quit [Read error: Connection reset by peer]
[11:52:56] -!- cncbasher [cncbasher!~quassel@cpc15-hart9-2-0-cust101.11-3.cable.virginmedia.com] has parted #linuxcnc-devel
[12:42:44] <jthornton> mhaberler: thanks
[12:42:59] <mhaberler> does it work for you?
[12:43:19] <jthornton> I have not been out to the shop yet... still waking up
[12:43:27] <mhaberler> ah, me too;)
[12:43:36] <jthornton> I assume I need a RIP of master to run it in now?
[12:44:29] <mhaberler> yes, to set the active colors you need the recent gladevcp patch in master (note 'man gladevcp', option -R , needs to be in ini file)
[12:45:11] <psha[work]> mhaberler: .rc?
[12:45:14] <jthornton> ok, it will be this evening before I can work on it as work gets in the way today
[12:45:16] <mhaberler> I found no side effects and wanted to get you going, but I think I'll have cmorley and psha have a second look, so I didnt make it permanent
[12:46:17] <mhaberler> yes, these properties (active etc) dont need python code, it can be done with gtk styles which you put into .gtk-2.0 file
[12:46:54] <mhaberler> see gtkrc file here: http://live.gnome.org/GnomeArt/Tutorials/GtkThemes
[12:47:41] <psha[work]> i'll check patch to be sure but generaly this is good feature
[13:11:18] -!- psha[work] has quit [Quit: Lost terminal]
[13:31:14] <jthornton> hmmm, in
[13:31:49] <jthornton> Axis 2.5 "R" will not run a program only "r" and the quick reference shows R as does the popup
[13:43:48] <skunkworks> same here
[14:11:04] -!- mk0 has quit [Quit: поздравляю всех с кредитом. халява продолжается.]
[14:14:37] <CIA-125> 03mhaberler 07v2.5_branch * r7d9d151935d1 10/ (5 files in 2 dirs): gladevcp: add custom label example
[14:15:29] <mhaberler> on push: remote: git_buildbot: ERROR: Could not connect to buildbot.linuxcnc.org:5133: User timeout caused connection failure.
[14:17:53] <JT-Shop> mhaberler: did you just add my fix to 2.5?
[14:18:17] <CIA-125> 03mhaberler 07master * rdc909f9217d5 10/ (5 files in 2 dirs): Merge branch 'v2.5_branch'
[14:18:26] <mhaberler> was 'your fix' the '-R' thing?
[14:19:01] <JT-Shop> dunno lol
[14:19:12] <JT-Shop> the need for me to use master for your label example
[14:19:19] <mhaberler> I see
[14:19:46] <mhaberler> I want to have it reviewed by cmorley and psha ; it does have f….p potentiaö
[14:20:16] <JT-Shop> ok, I just passed through the shop and saw the commit
[14:40:02] -!- hm2-buildmaster [hm2-buildmaster!~hm2-build@97-122-119-19.hlrn.qwest.net] has joined #linuxcnc-devel
[14:40:04] -!- linuxcnc-build [linuxcnc-build!~linuxcnc-@97-122-119-19.hlrn.qwest.net] has joined #linuxcnc-devel
[14:43:21] <GoSebGo> My dodgy network blinked again... I'll try to bring the buildbot up before leaving for work
[14:50:09] -!- linuxcnc-build has quit [Ping timeout: 245 seconds]
[14:50:29] -!- seb_kuzminsky has quit [Ping timeout: 256 seconds]
[14:50:31] -!- hm2-buildmaster has quit [Ping timeout: 252 seconds]
[14:52:51] -!- factor has quit [Remote host closed the connection]
[14:56:48] <GoSebGo> Should work now
[14:57:20] -!- seb_kuzminsky [seb_kuzminsky!~seb@97-122-107-227.hlrn.qwest.net] has joined #linuxcnc-devel
[15:03:33] -!- linuxcnc-build [linuxcnc-build!~linuxcnc-@97-122-107-227.hlrn.qwest.net] has joined #linuxcnc-devel
[15:03:35] -!- hm2-buildmaster [hm2-buildmaster!~hm2-build@97-122-107-227.hlrn.qwest.net] has joined #linuxcnc-devel
[15:16:25] -!- Poincare has quit [Read error: Operation timed out]
[15:55:57] <jepler> cradek was talking to me in meat-space about interpreters. I haven't followed the thread, but I am reminded of some work that is living in a branch in my personal repository: http://pastebin.com/raw.php?i=y9bdcGyv it makes the interpreter pluggable at runtime, and turns canterp into a second, pluggable interpreter. It's pretty old (started in 2010ish!).
[15:56:15] <jepler> hm, and somehow I added a .rej file. how embarassing!
[16:01:08] <cradek> that looks really useful
[16:02:50] <jepler> I wish I remembered if I thought it was stable or not
[16:07:02] -!- x_cnc_x has quit [Client Quit]
[16:08:35] -!- e-ndy has quit [Quit: Ex-Chat]
[16:39:44] -!- psha [psha!~psha@] has joined #linuxcnc-devel
[16:49:19] -!- e-ndy [e-ndy!~jkastner@static-84-242-102-36.net.upcbroadband.cz] has joined #linuxcnc-devel
[17:01:38] -!- WalterN has quit [Ping timeout: 240 seconds]
[17:06:18] -!- e-ndy has quit [Remote host closed the connection]
[17:06:48] -!- X704 has quit [Ping timeout: 245 seconds]
[17:10:51] -!- micges [micges!~x@apn-77-114-244-144.dynamic.gprs.plus.pl] has joined #linuxcnc-devel
[17:16:56] <psha> mhaberler: ack on gtkrc
[17:17:06] <mhaberler> thanks!
[17:17:13] -!- cncbasher [cncbasher!~quassel@cpc15-hart9-2-0-cust101.11-3.cable.virginmedia.com] has joined #linuxcnc-devel
[17:18:02] <psha> changing wheel in ~-20C and wind is not nice thing )
[17:20:06] <micges> it can be little cold ;)
[17:21:13] <micges> here is -15C
[17:28:17] -!- sliptonic-shop has quit [Ping timeout: 252 seconds]
[17:29:13] <mhaberler> jepler: what was the purpose of a 'pluggable interpreter'?
[17:30:20] <cradek> so you can pick which one you want at runtime, like kinematics
[17:30:49] <cradek> would open up the ability (for instance, at minimum) of lathes having different gcode dialect
[17:30:55] <mhaberler> but you remain with a single instance ?
[17:31:21] <mhaberler> oh, I see. An abstract base class and loading a something from a shlib
[17:31:49] <cradek> eh I don't know how it's implemented, but I know what he meant by pluggable
[17:32:38] <mhaberler> I read 'pluggable' and wished I heard 'multiple independent instances';)
[17:32:58] <cradek> yeah that feels like a totally different goal
[17:33:13] <cradek> but maybe you can base that on his work? I dunno.
[17:33:40] <mhaberler> I wish I had seen that earlier; some of the refactoring I already did looks like duplicated work
[17:33:50] <cradek> yuck
[17:34:06] <mhaberler> any other private treasures out there ;-?
[17:34:34] <mhaberler> it will be good read anyway
[17:34:51] <cradek> you never know what's lurking in git repositories worldwide...
[17:35:28] <mhaberler> throwing in the pluggable idea makes sense and with bit of c&p its not so much extra work
[17:36:13] <cradek> seems like once a user can choose at runtime, you might be a bit more free to innovate and try things
[17:36:35] <skunkworks> sounds cool
[17:37:34] -!- ve7it [ve7it!~LawrenceG@S0106001c10b7770f.pk.shawcable.net] has joined #linuxcnc-devel
[17:37:44] <mhaberler> btw: were you aware that canon already had partially evolved from C into a C++ class? I'm unsure wether its anywhere else, it seems code from peter vavaroutsos: http://fennetic.net/irc/emc3/src/emc/canonical/
[17:37:52] <mhaberler> I'm redoing that, too :-/
[17:39:00] <cradek> unfortunately I don't understand why that's interesting or important
[17:39:36] <mhaberler> did you rewrite it into C? just curious - not blaming
[17:40:49] <cradek> to the best of my knowledge, those files are not an ancestor of anything in the project
[17:40:56] <mhaberler> I see
[17:41:16] <mhaberler> thanks - I'm trying to understand the history so I dont go full circle needlessly
[17:43:04] Guest3403 is now known as fragalot
[17:43:10] -!- fragalot has quit [Changing host]
[17:43:57] <cradek> it looks to me like they are the result of a "what if" type discussion (because of the path irc/emc3)
[17:44:54] <cradek> and that canon interface has things like cyclehold, optionalstop, etc., that aren't part of the current canon interface
[17:46:08] <mhaberler> I talked to Fenn and it is a dead branch - he said Peter did most of it
[17:52:13] <cradek> mhaberler: petev c++ized some other parts of the system, and those changes were completed and incorporated - I think ini-reading was a part he redid
[17:53:00] <mhaberler> is he still around/active?
[17:53:11] <cradek> no
[17:53:16] <mhaberler> aja, ini sticks out
[17:53:57] <cradek> I don't remember if there were others
[17:54:56] -!- JT-BP [JT-BP!~chatzilla@216-41-156-59.semo.net] has joined #linuxcnc-devel
[17:56:10] <mhaberler> btw: this is IMO a showstopper bug due to language ambiguity:
[17:56:37] <JT-BP> 2.6 RIP make gives me an error that I don't understand "Makefile:55: Makefile.inc: No such file or directory"
[17:56:47] <JT-BP> Makefile:66: *** Makefile.inc must specify RTPREFIX and other variables. Stop.
[17:56:56] <mhaberler> O100 if [#<crash_machine>] LT 10
[17:57:05] <mhaberler> accidental code execution possible
[17:57:41] <mhaberler> it is executed as 'if #<crash_machine> > 0'
[18:01:26] <mhaberler> re canon c++ifying: its part of factoring out machine state into separate classes
[18:01:34] -!- theorbtwo has quit [Ping timeout: 272 seconds]
[18:01:54] theorb is now known as theorbtwo
[18:02:52] <awallin> for cutting-simulation, theres work by mpictor on canon-stuff. see http://code.google.com/p/cam-occ/source/browse/#git%2Ftrunk%2Fsrc%2Fg2m
[18:03:56] <awallin> mpictor's code used the output of the "rs274" binary that ships with LinuxCNC, parsed that text output, and created internal c++ classes of the machine state and motions in order to drive a cutting simulation
[18:04:03] <cradek> mhaberler: is the fix to make unexpected junk after the closing ] an error? I'm not sure I see the right fix
[18:04:43] -!- mozmck has quit [Ping timeout: 244 seconds]
[18:05:12] <cradek> I have to go look at the docs to "remember" whether the condition wants [] around it
[18:05:34] <cradek> yes it does
[18:05:51] <mhaberler> in the minimum, yes
[18:05:51] <cradek> so you can make non-comment junk after the ] the error
[18:05:58] <mhaberler> yes
[18:06:26] <cradek> I'll happily review a patch for that for 2.5
[18:06:47] <cradek> you said you found that in one of our example files?
[18:07:02] <mhaberler> yes, my very own blunder
[18:07:13] -!- mozmck [mozmck!~moses@client-] has joined #linuxcnc-devel
[18:07:52] -!- WalterN has quit [Ping timeout: 248 seconds]
[18:08:05] <mhaberler> http://git.linuxcnc.org/gitweb?p=emc2.git;a=blob;f=nc_files/m6demo.ngc;h=85a3c0f7f031a7735e1660403f22beee90302b4e;hb=dc909f9217d5f55425df64c57948e0527cb2ed1c#l30
[18:08:30] <mhaberler> actually the bison parser spotted it because it insisted on a bracketed expression without extra fluff
[18:09:09] <mhaberler> incidentally, in that case the result is the same which is why I didnt notice
[18:09:51] <mhaberler> ok, I'll see wether I can patch that
[18:09:55] <cradek> if (settings->test_value != 0.0) // true - execute forward
[18:10:12] <cradek> minor correction: it's maybe not >0 ?
[18:11:01] <cradek> fix goes near interp_read.cc:1697 I think?
[18:11:34] <cradek> perhaps also check repeat, elseif, while
[18:12:28] <mhaberler> I think its a parsing issue, not execution
[18:13:01] <mhaberler> parser sees a bracketed expression, calls it a day for that line and doesnt look further (I assume)
[18:13:33] <cradek> yes I think that's where I'm pointing
[18:13:52] <cradek> but nevermind, you're capable of fixing it without my pointers
[18:22:09] <mhaberler> yep, thats the issue - rs274 eats 'o100 if [10] !"§$"$%!' just fine
[18:32:16] -!- crib has quit [Ping timeout: 276 seconds]
[18:33:33] -!- GoSebGo has quit [Quit: Bye]
[18:33:48] -!- GoSebGo [GoSebGo!~Seb@] has joined #linuxcnc-devel
[18:38:04] -!- GoSebGo has quit [Ping timeout: 245 seconds]
[18:45:45] -!- e-ndy [e-ndy!~jkastner@static-84-242-102-36.net.upcbroadband.cz] has joined #linuxcnc-devel
[18:49:10] -!- A2Sheds has quit [Ping timeout: 276 seconds]
[19:04:10] -!- JT-Work [JT-Work!~chatzilla@216-41-154-150.semo.net] has joined #linuxcnc-devel
[19:14:54] -!- rob_h [rob_h!~rob_h@5ace704a.bb.sky.com] has joined #linuxcnc-devel
[19:24:38] -!- micges has quit [Ping timeout: 240 seconds]
[19:34:49] -!- e-ndy has quit [Quit: Ex-Chat]
[19:54:57] -!- GoSebGo [GoSebGo!~Seb@] has joined #linuxcnc-devel
[20:06:28] -!- L84Supper has quit [Quit: puff of smoke]
[20:06:59] -!- JT-Work has quit [Quit: ChatZilla 0.9.88 [Firefox 5.0/20110615151330]]
[20:21:27] -!- acemi has quit [Quit: WeeChat 0.3.2]
[20:24:12] -!- psha has quit [Quit: Lost terminal]
[20:29:25] <CIA-125> 03jthornton 07v2.5_branch * r2fd74b9df10d 10/docs/src/gui/ (ngcgui.txt ngcgui_de.txt ngcgui_es.txt ngcgui_pl.txt): Docs: add info on demo configs
[20:41:58] -!- syyl has quit [Quit: Leaving]
[21:16:25] -!- isssy has quit [Quit: Bye Bye]
[21:20:00] <cradek> mhaberler: please review and/or test http://timeguy.com/cradek-files/emc/0001-Error-on-unexpected-junk-after-if-.-etc.patch
[21:21:31] <cradek> (applies to v2.5_branch)
[21:27:54] <mhaberler> there's not a whole point in two folks doing the same patch.. I'll look.
[21:28:00] <mhaberler> I'm almost done too
[21:28:06] <cradek> ah crap
[21:28:10] <cradek> wonder if you found a better way
[21:28:50] <mhaberler> try your patch with runtests. I would guess it fails
[21:28:54] <cradek> I thought it was a one-liner but I got mired in read_o not properly consuming its arguments
[21:29:02] <mhaberler> positively not.
[21:29:04] <cradek> uh I did that
[21:29:11] <mhaberler> it IS ok?
[21:29:25] <cradek> no I just sent it to you to waste your time
[21:29:33] <mhaberler> sorry
[21:29:39] <mhaberler> I'll give it a try-
[21:29:39] <cradek> me too
[21:29:50] <cradek> it does pass the runtests and the manual testing I did
[21:30:04] <cradek> what looks wrong to you?
[21:31:25] <mhaberler> it's ok - sorry; I came up with pretty much the same thing
[21:31:44] <cradek> ok: o100 if [1 eq 0] (a) (b) ;c (d)
[21:31:54] <cradek> bad: o100 if [1 eq 0] (a) (b) c ;c (d)
[21:32:55] <cradek> other than wanting to allow comments there, is there any twist you can think of?
[21:33:38] <cradek> I don't like that I had to skip (don't notice) the error on sub definition, only later when it runs
[21:34:10] <cradek> if you don't have that problem in your patch, please finish
[21:37:18] <mhaberler> http://git.mah.priv.at/gitweb/emc2-dev.git/shortlog/refs/heads/oword-execution-fix
[21:38:24] <mhaberler> no. same thing here; but I guess Ken wasnt sure wether expressions should be executed in subs during parsing
[21:38:46] <mhaberler> I guess it would be safe to do because I cant think of side effects of that
[21:39:01] <mhaberler> or so he says, ha. Boils down to a try, I guess
[21:39:46] <mhaberler> and if Ken sells me one more time on the superiority of hand-written parsers I get a hysteric attack
[21:40:41] <mhaberler> I'll see what happens to runtests if I enable expression evaluation during sub parsing
[21:41:35] <cradek> I don't trust the runtests to be anything but a sanity check here
[21:41:40] <mhaberler> the whole thing would change the moment we let oword functions return values directly within expressions
[21:42:01] <mhaberler> because of possible side effects
[21:42:11] <mhaberler> currently it is ok
[21:42:57] <mhaberler> did you spot any difference between your and my patch ;-? I first thought I accidentially am looking at my repo
[21:44:29] <cradek> the difference between _setup.skipping and _setup.defining_sub
[21:44:52] <cradek> err skipping_o
[21:45:02] <cradek> I don't pretend to know exactly how those work
[21:47:33] <mhaberler> probably skipping_o is the better condition
[21:47:42] <cradek> style issue: I prefer strlen(literal) to sizeof(literal)-1
[21:48:08] <mhaberler> I take compile time over runtime when there's no ambiguity
[21:48:16] <cradek> I think the patches are more or less equivalent
[21:48:31] <mhaberler> but whatever the condition, it is unclear to me why we cant eval expressions in either case
[21:48:52] <cradek> we could enable that in master
[21:49:04] <mhaberler> I think that would be a better idea
[21:49:05] <cradek> we should not unnecessarily change 2.5
[21:49:17] <cradek> (I'm nervous about doing this much)
[21:50:13] <mhaberler> I guess I'll send Ken a mail and ask him why that cant be done, he surely had a reason and I have no intent to rediscover it
[21:50:39] <cradek> maybe there was just no need to do it?
[21:51:57] <mhaberler> I had looked at line 1583 here: http://git.mah.priv.at/gitweb/emc2-dev.git/blob/b53a1669c5db046b2139952a033ee51152c08927:/src/emc/rs274ngc/interp_read.cc#l1584
[21:52:14] <mhaberler> maybe one condition is a superset of the other
[21:53:01] <mhaberler> whatever 'just skip on out' means semantically..
[21:53:09] <cradek> else if(block->o_type == O_else) block->o_type = O_else;
[21:53:17] <cradek> I have to admit to being puzzled by all these ^^
[21:53:55] <mhaberler> try removing it - was there already.
[21:54:09] <JT-BP> I downloaded a fresh master from git and did autogen.sh and configure and I get this error from configure "configure: error: boost::python is required to build LinuxCNC"
[21:54:16] <mhaberler> it breaks something, maybe just syntax, cant remember
[21:54:19] <JT-BP> I assume I'm missing something
[21:54:40] <mhaberler> is that on a fresh machine?
[21:54:51] <JT-BP> but can't find anything about that on the wiki page for installing LinuxCNC
[21:54:54] <JT-BP> yes
[21:55:16] <mhaberler> try sudo apt-get install libboost-python-dev
[21:56:00] <JT-BP> thanks I'll add that to the wiki
[21:59:30] -!- A3Sheds has quit [Client Quit]
[21:59:51] -!- mhaberler has quit [Ping timeout: 248 seconds]
[22:00:56] -!- mhaberler [mhaberler!~mhaberler@] has joined #linuxcnc-devel
[22:02:03] <mhaberler> replugged into net
[22:03:19] <mhaberler> jt-Shop: it IS in the wiki..
[22:03:24] <mhaberler> hold on
[22:05:33] <mhaberler> probably needs to be moved, at least the package deps: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?GladeVcp
[22:06:06] <mhaberler> cradek: I suggest you push your patch to master, and I'll ask Ken
[22:06:32] <JT-Shop> I need to add that to the install page
[22:07:03] <Ekkeri> is there any news about pre3?
[22:07:21] <mhaberler> sorry, wrong page. those are the gladevcp deps
[22:07:31] <mhaberler> the boost stuff is only noted here: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?RemappingStatus
[22:08:14] <mhaberler> the libboost-python-dev dep would be good to add to the install section, yes
[22:09:41] <JT-BP> does that only apply to master?
[22:09:48] <mhaberler> yes
[22:10:07] <mhaberler> required by the remapping stuff
[22:11:53] <mhaberler> btw psha ok'd my gladevcp gtkrc patch, so I think it is ok to make it default, and turn off by flag; I think that should go into 2.5: http://git.linuxcnc.org/gitweb?p=emc2.git;a=commit;h=f62708319375ce2f26dccd686692c270b495f4d3
[22:12:08] <mhaberler> its only a fix, and all optional
[22:18:00] <JT-BP> cool
[22:24:04] -!- FinboySlick has quit [Quit: Leaving.]
[22:25:03] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[22:55:27] -!- joe9 has quit [Quit: leaving]
[23:11:41] -!- GoSebGo has quit [Quit: Bye]
[23:12:54] -!- GoSebGo [GoSebGo!~Seb@] has joined #linuxcnc-devel
[23:14:35] -!- syyl_ has quit [Quit: Leaving]
[23:15:50] <mhaberler> cradek: based on Ken's answer I think its ok to evaluate expressions in if/elsif/while/repeat during load
[23:16:11] <mhaberler> either you do it, or I, but not both;)
[23:26:09] -!- Lasivian has quit [Ping timeout: 240 seconds]
[23:29:27] <mhaberler> ok, the issue is: params in evaluated expressions are undefined and fail
[23:38:49] -!- linuxcnc-build has quit [Quit: buildmaster reconfigured: bot disconnecting]
[23:38:57] -!- linuxcnc-build [linuxcnc-build!~linuxcnc-@97-122-107-227.hlrn.qwest.net] has joined #linuxcnc-devel
[23:49:29] -!- jr_ has quit [Quit: Ex-Chat]