#emc-devel | Logs for 2006-08-13

[00:10:01] <alex_joni> I noticed an error during booting (this is a livecd)
[00:10:13] <alex_joni> and now I installed the deb with the non-free modules
[00:10:22] <alex_joni> was wondering if I can make it work without rebooting
[17:18:17] <alex_joni> jepler: help
[17:20:32] <alex_joni> what does this mean?
[17:20:34] <alex_joni> make: Failed to remake makefile `Makefile'.
[18:35:12] <jepler> alex_joni: I got it earlier when I didn't do an update -d at the top level
[18:35:37] <jepler> alex_joni: the Makefile includes other files, some of which are generated. If Make can't generate all its include files, it prints that message
[18:35:46] <jepler> alex_joni: it's not very helpful as a message
[18:35:55] <jepler> and it should probably be treated as an error, but I don't see how to do that.
[18:36:04] <alex_joni> I still get it
[18:36:12] <alex_joni> and I did 'cvs up -dP'
[18:36:15] <alex_joni> from emc2/
[18:36:56] <jmkasunich> for what its worth, I just did "cvs up -dP ; cd src ; ./cponfigure ; make" and it was clean
[18:37:16] <jmkasunich> well, it would be clean if I did ./configure, not ./cponfigure
[18:38:05] <alex_joni> jepler: any way to figure out what the files are it depends on?
[18:38:11] <alex_joni> I might have one dated oddly ?
[18:39:01] <jmkasunich> hmm, seems like lyx leaves droppings - in docs/src/hal:
[18:39:01] <jmkasunich> halshow.aux
[18:39:02] <jmkasunich> halshow.lyx
[18:39:02] <jmkasunich> halshow.lyx.aux
[18:39:02] <jmkasunich> halshow.tex
[18:39:15] <jmkasunich> halshow.lyx is the source, the rest are droppings
[18:39:15] <alex_joni> these are from various exports
[18:39:32] <alex_joni> I had tex all over the place, after trying an latex export
[18:39:35] <jmkasunich> wonder if the make should clean them up
[18:39:42] <alex_joni> the aux I've never seen
[18:39:48] <alex_joni> might be your text editor?
[18:39:56] <jmkasunich> I have the same three for ecery lyx file
[18:40:03] <jmkasunich> nope, not from an editor
[18:40:23] <alex_joni> hmm.. what exports did you run?
[18:40:27] <alex_joni> any from the GUI?
[18:40:35] <jmkasunich> hard to say, lots of stuff lately
[18:40:38] <alex_joni> :)
[18:40:43] <jmkasunich> they are dated 8/5
[18:40:50] <jmkasunich> I'm gonna clean them, and see if they come back
[18:42:29] <alex_joni> jepler: can you look at the axis install diff if it's ok?
[18:44:26] <alex_joni> jmkasunich: seen http://www.motionsystemdesign.com/full_story.php?WID=15301?
[18:44:33] <alex_joni> was posted on the users list
[18:46:02] <jmkasunich> that url fails
[18:46:11] <jmkasunich> lemme look on the list, maybe it got truncated
[18:46:23] <alex_joni> jmkasunich: seen http://www.motionsystemdesign.com/full_story.php?WID=15301 ?
[18:46:36] <alex_joni> wrongly put question mark :)
[18:49:12] <jmkasunich> interesting
[18:49:22] <jmkasunich> more than my brain can process at the moment
[18:49:27] <alex_joni> same here
[18:49:38] <alex_joni> that's why I thought I'd let it bounce off :)
[18:52:45] <Lerneaen_Hydra> sounds interesting
[18:52:58] <jmkasunich> someday I'd like to work on an autotuner
[18:53:05] <jmkasunich> tell it "tune pid loop 0"
[18:53:25] <Lerneaen_Hydra> it doesn't look all too complex
[18:53:29] <jmkasunich> and it would use halcmd to find pid.0.command, see what signal is connected to it, disconnect that and connect its own signal
[18:53:54] <jmkasunich> and start issuing "setp pid.0.pgain <value>" etc
[18:55:35] <jmkasunich> btw, that aux and tex stuff didn't come back after doing make docs, must be from something else
[18:57:03] <alex_joni> jmkasunich: good
[20:36:34] <alex_joni> jmkasunich: still around?
[20:47:55] <alex_joni> jmkasunich: was looking for the canonical description, found it in docs/src/hal/general_ref.lyx
[20:56:20] <jepler> alex_joni: I looked and did not spot any problems
[20:56:51] <alex_joni> jepler: good... I wasn't able to build a deb, because of the cannot rebuild Makefile problem
[20:57:00] <alex_joni> dpkg-buildpackage just stops on that
[20:57:21] <jepler> alex_joni: I get that message too, I'll see if I can figure out what it's about
[20:57:30] <alex_joni> yay
[20:57:47] <alex_joni> cd src && /usr/bin/make depclean -s
[20:57:47] <alex_joni> make[1]: Failed to remake makefile `Makefile'.
[20:57:47] <alex_joni> make: *** [clean] Error 2
[20:58:04] <alex_joni> this is the error from dpkg-buildpackage
[21:00:58] <jepler> Must remake target `../docs/src/Master_developer.d'.
[21:00:58] <jepler> Failed to remake target file `../docs/src/Master_developer.d'.
[21:02:16] <jepler> DOC_SRCS should list Master_Developer.lyx (upperacse 'D' of developer)
[21:02:47] <jepler> fix checked in
[21:06:17] <alex_joni> jepler: oh.. thanks
[21:42:00] <alex_joni> the STG1 is a mess
[21:42:15] <alex_joni> at least the relevant parts about index pulse
[21:43:08] <jmkasunich> yeah, ppmc is the same way - the hardware docs don't really tell you how it handles index
[21:44:19] <alex_joni> here it says.. but it's really a MESS
[21:44:32] <alex_joni> 4 or 5 registers need to get updated
[21:45:07] <jmkasunich> damn
[21:45:11] <alex_joni> then to reset the latched value you need to load some 2-bit value into an register (to select channel 0 or 1,2 or 3,4 or 5,6 or 7)
[21:45:34] <alex_joni> and to really reset it you read from another register (to reset the odd ones), or yet another to reset the even ones
[21:45:37] <jmkasunich> the canononical interface was designed to be as simple as possible, hoping that just about any hardware could do it
[21:45:37] <alex_joni> * alex_joni pukes
[21:46:23] <jmkasunich> don't get any on yourself
[21:46:45] <alex_joni> "The latched-high-when-active output of the index pulse pair selected by IXS1 and IXS0 will appear here. The active level is set by IXLVL signal in register INT.C (see section 5.1.8 for more information). To reset the even bit, read BRDTST; to reset the odd bit, read ODDRST. The bits are level triggered, and can not be reset if they are active."
[21:47:06] <jmkasunich> eww eww eww
[21:47:08] <alex_joni> one example of info I need to read quite a few times before I start to understand what they meant :/
[21:47:38] <alex_joni> and of course the STGv2 is completely different :(
[21:49:16] <jmkasunich> hard to believe it was once the best motion control card
[21:50:12] <alex_joni> yeah, and from the fact that no-one ever complained about a missing index pulse.. I'm tempted to say no-one uses it
[21:50:25] <alex_joni> I mean homing with index pulse
[21:51:35] <alex_joni> * alex_joni loses interest for tonight :/
[21:53:39] <alex_joni> I see the emc1 driver used to poll this