#emc-devel | Logs for 2009-02-01

[00:13:38] <cradek> jepler: did that fix the starting-at-zero thing?
[00:45:27] <jepler> cradek: I see it rarely enough that I'm not confident.
[00:45:57] <cradek> I will watch carefully.
[01:25:59] <CIA-2> EMC: 03cmorley 07TRUNK * 10emc2/src/hal/classicladder/module_hal.c: change the allotment of symbols to automatically be large enough for all the variables unless forced smaller on the command line
[02:12:27] <cradek> what timezone is the feature freeze??
[02:22:09] <jepler> cradek: oops, you made the mistake of asking
[02:22:16] <jepler> cradek: what are you going to do if I say "GMT"?
[04:42:18] <cradek> jepler: I was just kidding, anyway
[04:45:24] <seb_kuzminsky> why god why wont they read the frakking manual
[04:45:34] <SWPadnos> heh
[04:45:42] <SWPadnos> I'm typing a response right now
[04:45:42] <seb_kuzminsky> is it so much to ask
[04:45:52] <SWPadnos> "read"
[04:45:55] <SWPadnos> "think"
[04:45:57] <SWPadnos> hmmm
[04:46:00] <SWPadnos> t may be
[04:46:02] <SWPadnos> it
[04:46:09] <seb_kuzminsky> it takes longer to send an email than to just look it up
[04:46:41] <SWPadnos> it takes longer to flail around like Homer Simpson, then write an email, than it does to read about the tools that would help you to debug it :)
[04:46:52] <seb_kuzminsky> i just replied to donnie giving him the section in the integrator's manual and a link to it
[04:47:21] <SWPadnos> well, I'll respond also, we'll see what happens
[04:47:26] <SWPadnos> he did quote me after all
[04:47:55] <seb_kuzminsky> he grepped irc logs, but didnt read the section titled "Homing"
[04:48:25] <SWPadnos> heh
[04:48:35] <SWPadnos> grep != google
[04:48:42] <SWPadnos> (same thing, different users)
[04:49:30] <SWPadnos> note the pin names. that implies an older driver doesn't it?
[04:49:39] <SWPadnos> (with the connector numbers in them)
[04:50:00] <SWPadnos> or is that a current hm2_5i23 (rather than hm2_pci)
[04:50:29] <seb_kuzminsky> it's a bit confusing
[04:50:39] <seb_kuzminsky> hm2_pci is the name of the *driver* for all the pci-based anyio boards
[04:51:06] <seb_kuzminsky> it exports hal objects named hm2_<BoardName>.blah.blah, depending on what board(s) it finds
[04:51:35] <seb_kuzminsky> if the gpio name has the connector name and IO number in it it's 2.2, if it's just IO number it's trunk
[04:52:04] <SWPadnos> ok
[04:52:11] <SWPadnos> so he's using 2.2
[04:52:12] <cradek> heh, for once I gave an answer and seb gave an rtfm
[04:52:24] <seb_kuzminsky> heh
[04:52:27] <SWPadnos> heh
[04:52:55] <cradek> feature freeze!!
[04:52:59] <cradek> I'm excited about 2.3
[04:53:06] <seb_kuzminsky> brrr
[06:03:00] <cradek> two rtfms...
[06:39:10] <seb_kuzminsky> gmane & i only see the one from me so far
[06:39:23] <SWPadnos> I included both the answer and the RTFM
[06:39:47] <seb_kuzminsky> ah
[06:40:01] <seb_kuzminsky> and the rtfSampleConfig, which is pretty good
[06:40:19] <SWPadnos> heh
[06:40:30] <SWPadnos> I don't think it's so good on 2.2 though
[06:40:43] <SWPadnos> but there are others that use index homing (I hope)
[13:04:56] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/ladder/images/stepconf_ladder.png: add image
[14:29:18] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/ladder/images/stepconf_ladder.png: this looks better
[14:30:04] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/ladder/ (classic_ladder.lyx ladder_examples.lyx): add info on stepconf and clean up
[15:25:45] <jepler> jepler has changed the topic to: EMC2 development -- http://linuxcnc.org/ | 2.3 schedule: stabilize; no new features. 2.3~beta on march 1 | channel logged by logger_dev
[15:49:23] <BigJohnT> jepler: I didn't mean to strike a raw nerve yesterday I honestly forgot that I made the commit and missed where cradek had reversed it
[15:50:01] <BigJohnT> the second example I was just following what you had did 2 1/2 years ago to move the X over .5"
[16:02:01] <jepler> BigJohnT: I might make a different decision about that today as well
[16:03:11] <alex_joni> * alex_joni smiles
[16:03:24] <BigJohnT> * BigJohnT grins
[16:27:41] <CIA-2> EMC: 03jepler 07TRUNK * 10emc2/lib/python/rs274/glcanon.py: when the number of 'show's is greater than the number of 'hide's, show the preview
[16:29:23] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/gui/axis.lyx: add info on hide/show must be in pairs
[16:29:32] <alex_joni> there you go
[16:29:39] <BigJohnT> yup
[16:30:15] <alex_joni> BigJohnT: might have earned enough bonus points to be allowed near axis.ngc again
[16:30:22] <alex_joni> * alex_joni is obviously kidding ..
[16:30:34] <BigJohnT> LOL
[16:34:53] <CIA-2> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/gui/axis.lyx: add info on hide/show must be in pairs
[16:36:00] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/gui/axis.lyx: add a couple more words on hide/show cause what I did in 2.2.x looked better
[16:57:15] <BigJohnT> * BigJohnT has to go to town now
[17:13:52] <micges> jepler: I'm trying to discover command path after c.auto() and don't fully undestand s->c->write(run) ? at line 1133
[17:14:43] <micges> command was send where ? task?
[17:14:58] <alex_joni> all commands from axis or emcmodule get sent to task
[17:15:40] <micges> by shared command object in memory ?
[17:16:01] <alex_joni> by NML.. yes
[17:16:25] <alex_joni> depending on how your emc.nml is set up, that can be a shared memory to send the command
[17:16:45] <alex_joni> if AXIS is on a different PC, then the NML command is sent through the network..
[17:17:14] <micges> and task then is a server side
[17:18:01] <micges> ok thanks
[17:18:15] <alex_joni> yes, task needs to be on the same machien as emcsvr
[17:22:37] <CIA-2> EMC: 03jepler 07TRUNK * 10emc2/docs/src/gui/axis_fr.lyx: compensate for moved image
[17:26:07] <jepler> er, no, that's not it
[17:26:16] <alex_joni> oh, so it's failing for you too?
[17:26:36] <alex_joni> make: *** [build-stamp] Error 2
[17:27:45] <jepler> $ make docs
[17:27:46] <jepler> make: Failed to remake makefile `Makefile'.
[17:28:16] <alex_joni> helpful make
[17:29:20] <alex_joni> running make it builds Master_HAL.pdf then bombs out without an error
[17:30:24] <alex_joni> ah, BigJohnT didn't add the image
[17:30:32] <alex_joni> axis-currentandselected.png
[17:32:33] <CIA-2> EMC: 03jepler 07TRUNK * 10emc2/docs/src/gui/axis_fr.lyx: whoops, that wasn't right. revert to rev 1.16
[17:33:58] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/docs/src/gui/images/axis-currentandselected.png: I htink BJT wanted to move it in here
[17:34:13] <alex_joni> like I said.. can't type today :/
[17:35:20] <SWPadnos> that's what you htink
[17:36:23] <alex_joni> sure hting
[17:36:48] <SWPadnos> taht's waht I htought
[17:36:52] <alex_joni> jepler: make clean && make docs seems to work here
[17:37:59] <jepler> good, thanks
[17:38:21] <alex_joni> back to debuild :)
[17:46:34] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/debian/drivers.files.in: add hm2_pci for consistency
[17:47:27] <CIA-2> EMC: 03jepler 07TRUNK * 10emc2/docs/src/Submakefile: try to give a better error when an image file isn't found
[17:48:08] <jepler> alex_joni: I think that instead all the modules in drivers.files.in can be removed
[17:48:18] <jepler> they should be gotten by @MODULE_PATH@/*@MODULE_EXT@ in emc2.files.in
[17:49:05] <alex_joni> jepler: yeah, but drivers.files.in should be only for non-sim
[17:49:16] <alex_joni> otherwise the drivers will end up in the package aswell..
[17:49:42] <jepler> alex_joni: no, it's OK
[17:50:06] <jepler> this @MODULE_PATH@/*@MODULE_EXT@ is looking at files put there by 'make install' so it won't have non-drivers files on sim systems
[17:50:15] <alex_joni> ah, right
[17:50:30] <alex_joni> ok, then I guess drivers.files.in can go away completely
[17:50:38] <jepler> it can't be done for bin, though -- I think because of /usr/bin/comp going in emc2-dev
[17:51:43] <alex_joni> hmm.. there are 3 files which aren't drivers
[17:52:01] <alex_joni> usr/bin/m5i20cfg, usr/bin/pci_write, usr/bin/pci_read
[17:52:25] <jepler> but they're files that don't exist in sim systems
[17:52:48] <alex_joni> since they are userspace.. they might just exist
[17:54:54] <alex_joni> and I don't think they are used anymore
[17:55:18] <SWPadnos> pci_read and pci_write may not be needed for the emc2 package
[17:55:22] <alex_joni> SWPadnos: I notice you are the author on pci_write and pci_read
[17:55:46] <SWPadnos> maybe leave them in the -dev package, since they can be used while writing new drivers for PCI cards
[17:55:49] <SWPadnos> yes
[17:57:10] <alex_joni> usr/bin/m5i20cfg looks like a driverless mode to talk to a 5i20
[17:57:14] <jepler> we can't drop m5i20cfg until we drop the old hal_m5i20 driver
[17:57:22] <jepler> it's used for programming a second card or using a non-default firmware
[17:57:30] <jepler> probably nobody uses it, but oh well
[17:57:32] <alex_joni> jepler: no reason to drop it then
[17:57:39] <SWPadnos> isn't 5i20 supposed to be dropped in 2.3?
[17:57:41] <alex_joni> no harm in having it in the sim package though
[17:58:07] <jepler> SWPadnos: no, I don't think so
[17:58:13] <SWPadnos> ok
[17:58:34] <jepler> the intermediate drivers by seb that were never widely used are removed
[17:59:01] <jepler> but we shouldn't take away the long-useful hal_m5i20 on no notice
[17:59:12] <jepler> (but probably we should add a "deprecated; to be removed in 2.4" message now!)
[17:59:19] <alex_joni> and with little gain..
[17:59:30] <SWPadnos> I agree. I thought notice had been given, but if that was for the early hm2 stuff, I'm fine with leaving 5i20 in there
[17:59:35] <alex_joni> I mean, it's somehow ok to force users to take a different driver if they have any benefits..
[18:00:17] <alex_joni> ok, then I'll move m5i20cfg, pci_read and pci_write to emc2.files.in
[18:00:30] <jepler> umm what?
[18:00:32] <alex_joni> and get rid of drivers.files.in
[18:00:43] <SWPadnos> pci_read and pci_write aren't necessary for 5i20
[18:00:54] <SWPadnos> only m5i20cfg
[18:01:00] <alex_joni> SWPadnos: right, but you said they are usefull sometimes
[18:01:07] <alex_joni> no reason to remove them from the package
[18:01:09] <jepler> m5i20cfg is useless to a sim user, and I don't even think it's built
[18:01:10] <SWPadnos> when developing new drivers
[18:01:35] <alex_joni> * alex_joni checks
[18:04:43] <alex_joni> right, not built
[18:04:57] <alex_joni> ok.. then drivers.files.in it stays, although the name is bogus now
[18:12:00] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/debian/drivers.files.in: no need to specify drivers here, they get picked up from rtlib
[18:24:46] <Roguish> hey, you guys gonna drop the current m5i20 driver? in 2.3? hope not. i don't want to have to rewrite my configs.....
[18:26:28] <alex_joni> Roguish: nope
[18:26:34] <Roguish> thanks.
[18:26:44] <alex_joni> maybe in 2.4
[18:27:30] <Roguish> i just don't see the new driver as 'bullet proof stable' just yet.
[18:57:03] <alex_joni> did you try it out?
[19:01:17] <Roguish> alex_joni: no, haven't had time. been finishing up a big project and starting a new day job.
[19:05:29] <alex_joni> I think it's gotten pretty stable now
[19:05:46] <alex_joni> all errors lately were config/user specific
[19:06:28] <Roguish> ok, maybe have some time over the next few weeks. i am just concerned about all my hal files.
[19:16:42] <jepler> work on a copy of 'em
[20:27:42] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/scripts/emc.in: fix test in case no INTRO_GRAPHIC is defined in the ini
[20:31:22] <alex_joni> humm.. did anybody try to run the splash in lathe mode?
[20:39:58] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/tcl/mini.tcl: hopefully the last remnants
[20:45:55] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/tcl/mini.tcl: it usually helps to be in MDI mode before sending MDI commands
[20:47:49] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/tcl/mini.tcl: disable debug prints
[20:55:53] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/configs/sim/mini.ini: mini sends some MDI commands before homing
[20:56:44] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/configs/sim/servo_sim.ini: reduce debug level
[21:05:26] <CIA-2> EMC: 03alex_joni 07TRUNK * 10emc2/configs/vismach/hbm.var: clear offsets forgotten in the var file
[23:41:54] <jtr_> jtr_ is now known as jtr