#linuxcnc-devel | Logs for 2016-03-20

[00:51:15] <trentster> Hi all, wondering if someone can help me with the command to reset machine absolute home position for a single axis after "home all" has been previously run?
[00:52:39] <trentster> e.g. I have proximity sensors on X and Y but not on Z, when I click home all, Z gets set to "0" at the current position it is. Later when I run an auto tool height probe I get an error that it will exceed the limits. I need to reset Z so that the probe succeeds.
[00:53:24] <trentster> I am asking in here as I have asked this question in the regular channel but don't ever get a reply ;-)
[01:04:13] <trentster> https://monosnap.com/file/i8hxOG5A2SNL9VokEpTLr3Q2jWPw1D
[01:04:24] <trentster> https://gist.github.com/anonymous/e36edb3ed1fd348fb369
[05:06:47] <emcPT> Yestarday I try to make my first commit.
[05:06:57] <emcPT> And I am missing the sign off
[05:07:41] <emcPT> that is mandatory. Today I created a gpg key and I commited using: git commit -S (that should make a signature)
[05:08:22] <emcPT> however I do not see any difference in the commit output, although when I issue git log --show-signature -1, I can see
[05:08:35] <emcPT> gpg: Signature made Sun 20 Mar 2016 09:36:59 AM WET using RSA key ID 1A2601AA
[05:08:36] <emcPT> gpg: Good signature from "Filipe Tomaz (emcPT) <eusurplus@gmail.com>"
[05:08:36] <emcPT> Author: Filipe Tomaz <eusurplus@gmail.com>
[05:08:36] <emcPT> Date: Sun Mar 20 09:36:28 2016 +0000
[05:08:56] <emcPT> So I did it ok, or I am missing something?
[05:36:06] <ikcalB> seb_kuzminsky: tnx for the build w/o the "unexpected realtime error". running fine since 3+ days. the machine is not heavily used, but so far no regression, and ofc, no unexepected delay warning. going to report back in a week or so
[08:39:57] <jepler> emcPT: http://linuxcnc.org/docs/2.7/html/code/contributing-to-linuxcnc.html#_signed_off_by_policy -- it is not about public key signatures, but about saying that you contribute your changes under the terms of our license.
[09:28:24] <emcPT> jepler: Do you know where the tooltable is save? on src/emc/iotask/ioControl.cc and src/emc/iotask/ioControl_v2.cc there is saveToolTable() but it seams that the code does not pass though it when the user press "saveFile" under the ToolEdit menu
[09:42:07] <emcPT> it is probably arround line 756 in tcl/tooledit.tcl
[10:19:18] <KGB-linuxcnc> 03Dewey Garrett 05joints_axes12 f419731 06linuxcnc 10src/emc/motion/command.c 10src/emc/motion/control.c 10src/emc/motion/motion.c motion: initialize all axis->locking_indexer items * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=f419731
[10:19:18] <KGB-linuxcnc> 03Dewey Garrett 05joints_axes12 e1a6774 06linuxcnc 10docs/src/getting-started/updating-linuxcnc.txt 10src/emc/usr_intf/axis/scripts/axis.py axis.py: bind [,] keys using trajcoordinates (JA) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e1a6774
[10:28:15] <emcPT> second attempt patch sent. I hope I did it correctly now.
[12:36:16] <KGB-linuxcnc> 03Jeff Epler 05jepler/ja/axis-jogging-config 9fc3c58 06linuxcnc 10configs/sim/axis/axis_foam.ini configs: foam: this has no locking indexer * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9fc3c58
[12:36:16] <KGB-linuxcnc> 03Jeff Epler 05jepler/ja/axis-jogging-config 65b58eb 06linuxcnc 10docs/src/config/ini-config.txt 10docs/src/gui/axis.txt 03docs/src/gui/images/axis-foam.png docs: document AXIS's foam mode * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=65b58eb
[12:36:16] <KGB-linuxcnc> 03Jeff Epler 05jepler/ja/axis-jogging-config b2e468d 06linuxcnc 10docs/src/config/ini-config.txt 10docs/src/getting-started/updating-linuxcnc.txt 10src/emc/usr_intf/axis/scripts/axis.py axis: keyboard jogging improvements * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b2e468d
[12:37:21] <jepler> dgarr: if you're lurking, please let me know if you think that branch is on the right track. if so I can merge to ja12.
[12:37:29] <jepler> (or rebase)
[14:40:35] <rob_h> emcPT, are you saying the info on wiki i put is wrong? as i took this from a fanuc 31 turning manual and comapred it to other mauals also for front & rear post, alot of websites i found had it down wrong to any fanuc manual
[14:45:56] <emcPT> Yes, I am afraid is wrong. At least for the rear post is wrong, but I also believe that the tooltip is equal if you are working with a rear or a front toolpost.
[14:46:05] <emcPT> I have fanuc manuals with me.
[14:46:42] <rob_h> what manual you have?
[14:46:53] <rob_h> might have a 16 or 21 here
[14:47:09] <emcPT> Fanuc, haas, siemens, ... all use the same
[14:47:35] <rob_h> i know we have all them controls apart from a hass never ever lol
[14:48:02] <emcPT> For example: https://diy.haascnc.com/imaginary-tool-tip-and-direction
[14:48:24] <emcPT> Take in consideration that X++ is vertical up and Z++ is horizontal right
[14:53:05] <emcPT> I have 0i, 0i-TD, 21, 18 manual. All are in line with the same tooltip.
[14:57:45] <rob_h> bin along time since i looked at them
[14:59:15] <rob_h> something like this in manual . https://ibin.co/2avqyDjrldfr.jpg
[14:59:54] <rob_h> as you say wiki image is abit upside down
[15:13:15] <emcPT> Correct the link is from a Fanuc manual and it is the same image that I sent to linuxcnc forum.
[15:13:31] <emcPT> And it does not comply with the image on the wiki
[15:13:46] <rob_h> agree
[15:15:05] <emcPT> The path that I sent hopefully solves this
[15:15:10] <emcPT> patch
[15:15:19] <rob_h> did you have a slant lathe too?
[15:15:29] <rob_h> i looked at your forum post quick
[15:15:38] <emcPT> Currently 4
[15:16:38] <emcPT> The one in the post was the 2nd that I retrofitted with linuxcnc, back in 2013/2014
[15:17:42] <rob_h> as the rotate screen trick always hides the tool shapes in axis
[15:18:51] <emcPT> I also coded a method to overcome that
[15:18:59] <emcPT> that I am able to share
[15:19:14] <rob_h> nice
[15:19:32] <rob_h> i mean it is there just flat in the Y plane
[15:19:51] <emcPT> it allows the user to define shapes and to "draw" tools
[15:20:25] <rob_h> so when i make that fancy tool in the cam i can put it into linuxcnc?
[15:20:41] <emcPT> but it is only useful in lathes, and I do not know what would be the behaviour in mills
[15:21:10] <emcPT> I will try to take a snapshoot to show
[15:21:22] <rob_h> nice
[15:21:31] <rob_h> lathe does need abit of work ncie to see someone working on it
[15:23:28] <rob_h> my problem is i have a 4 axis lathe to convert next with power tooling
[15:34:22] <emcPT> check https://ibin.co/2aw1OCRf6Wm6.png
[15:38:46] <emcPT> ibin is not working. Try postimg.org/image/gvmkkgn7v
[15:46:54] <rob_h> nice work
[15:48:54] <emcPT> I will try to clean up my code and also submit a patch, but first I would like to know if the patch that I sent works, so that I do not make errors twice.
[15:50:06] <jepler> The Q- number in G10 L10 should probably obey whichever system the inifile specifies. with the patch I looked at, you have to enter the linuxcnc-style number even if the inifile says to use fanuc numbers.
[15:57:22] <emcPT> jepler: correct. The patch does not address a G10 that changes the orientation of a tool.
[16:04:12] <jepler> I would rather see more complete documentation than what that second patch has, but I understand it is difficult to write documentation.
[16:05:37] <emcPT> it wa s
[16:05:53] <emcPT> it was in line with the documentation of the random toolchange
[16:06:02] <emcPT> that was present on the line before
[16:07:58] <jepler> we should probably not needlessly use the name "fanuc" in the configuration item's name if it's actually a standard of multiple industrial controls.
[16:08:47] <jepler> TOOLTIP_ORIENTATION_STANDARD = 1
[16:09:01] <emcPT> For what I know, linuxcnc is the only different.
[16:09:21] <rob_h> y not just tooltip_orientation= ? why any extra
[16:09:44] <emcPT> Why not use the standard always?
[16:10:01] <rob_h> becasue you brake alot of peoples setups
[16:10:33] <emcPT> I think people would like to use the standard.
[16:10:54] <jepler> I doubt we intended to do the opposite of the standard the first time around, but we did
[16:11:05] <jepler> if a release goes out that makes everybody's tool tables change meaning, that's REAL BAD
[16:11:16] <jepler> just like if we made a change that made all the tools 25mm wrong in length
[16:12:01] <emcPT> In fact, I am not sure if there are so many users of linuxcnc for lathes that use G41 G42. The tooltip orientation only changes the behaviour in G41 G42
[16:12:47] <jepler> I know the original author of the tool orientation code in linuxcnc uses G41/G42 when he uses his lathe
[16:12:54] <emcPT> And the machining result would be at most 2*tool_radius difference
[16:13:04] <emcPT> so no big damage
[16:13:25] <emcPT> (unless a large tool is used :-)
[16:17:14] <emcPT> But still it is strange to have a TOOLTIP_ORIENTATION_STANDARD = 1
[16:17:41] <emcPT> like the other option was to have TOOLTIP_ORIENTATION_NONSTANDARD
[16:19:29] <jepler> I just want to not use a trademarked name needlessly
[16:22:11] <jepler> (yes there are places the software and documentation already say 'fanuc' and yes on reflection I'd rather get rid of each one of them, even where it is used in hal pins and would therefore break user configurations in the wild)
[16:22:44] <jepler> (because getting a trademark cease and desist letter from a law firm is not a fun event, and that has happened in the past to this project)
[16:23:25] <emcPT> sure. I will change that and make a patch again. There is no need to make more things wrong just because there are already done before.
[16:23:46] <jepler> thank you for understanding my point of view.
[16:29:35] <emcPT> Regarding the use of the G10 changing the orientation, it is not trivial to change the code. I must pass the information about the current state of the fanuc_tooltip_orientation to where the G10 is processed, and I do not know how to do that - at least for now.
[16:35:20] <jepler> G42.1 L- is another case which would need its supporting c++ code changed
[16:38:09] <emcPT> Damm I am starting to feel really small
[16:41:00] <jepler> It's not my intent to make you feel small. I am really glad you have offered your change, and I know your goal is to improve linuxcnc!
[16:42:55] <jepler> I am a lousy machinist but I have been working on linuxcnc for a long time which sometimes lets me understand more quickly how a change will affect the software
[16:46:39] <emcPT> I know that (not that you are a lousy machinist)! Your contribution is widely known.
[16:47:33] <emcPT> Ok. Name was changed using find . -type f -print0 | xargs -0 sed -i 's/FANUC_TOOLTIP_ORIENTATION/STANDARD_TOOLTIP_ORIENTATION/g'
[16:48:00] <emcPT> so it will stick to STANDARD_TOOLTIP_ORIENTATION
[16:50:02] <emcPT> I will try to look about the G10 and G41/G42.1, but probably it is over my knowledge.
[16:54:21] <emcPT> Does the G42.1 also writes the tool table? No?
[16:54:24] <jepler> no
[16:58:19] <jepler> afk, thank you for taking the time to talk about your patch with me.
[17:10:04] <emcPT> I think I manage to work with the G10. In Interp::convert_setup_tool(block_pointer block, setup_pointer settings)
[17:10:30] <emcPT> I will pass the current state of STANDARD_TOOLTIP_ORIENTATION as part of the settings
[17:10:36] <emcPT> and act accordinally
[17:18:43] <emcPT> yes. it works
[19:40:19] <KGB-linuxcnc> 03Chris Morley 052.7 cf1fd1e 06linuxcnc 10(5 files) stepconf: dynamically show how step scale is calculated * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=cf1fd1e
[19:40:20] <KGB-linuxcnc> 03Chris Morley 052.7 b9d47a8 06linuxcnc 10src/emc/usr_intf/stepconf/stepconf.py stepconf: fix default pitch for a axis * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b9d47a8