#emc-devel | Logs for 2010-05-27

[00:31:40] <dgarr> for consideration to eliminate a confusing message: http://www.panix.com/~dgarrett/stuff/0001-emc.in-eliminate-message-to-DEBUG_FILE-for-nonexiste.patch
[00:32:15] <dgarr> for consideration: http://www.panix.com/~dgarrett/stuff/0001-integ-add-gain-pin.patch
[00:32:29] <dgarr> (analog computer integrators always had a gain pot)
[01:38:06] <KimK> I was going through the examples on the "git" page in the wiki, and under "3. View History", I could not get this one to work: Get a closer look at a particular change by commit: git log -C -p -1 57c609
[01:38:16] <KimK> It says: fatal: ambiguous argument '57c609': unknown revision or path not in the working
[01:38:16] <KimK> Use '--' to separate paths from revisions
[01:39:35] <KimK> Am I doing something wrong? Any advice appreciated.
[01:39:58] <KimK> (I just copied and pasted the example)
[01:40:33] <SWPadnos> maybe there are two changes that could be identified by 57c609
[01:40:41] <SWPadnos> or none
[01:42:20] <KimK> OK, thanks, SWP, I'll work on that awhile and see how far I get.
[01:42:26] <SWPadnos> have fun
[02:24:53] <Dave911> jepler: I'm back if you have any questions about that Modbus patch .. sorry I had to run off.. kid commitment..
[02:39:13] <Dave911> I put those Modbus changes into the latest version of ClassicLadder and they work well there also.. I think I will fix the Modbus config menu the author broke and send him a patch file.. or have Chris do it as I think he has his personal contact info.
[02:49:52] <Dave911> I'd like to incorporate the CL editor, which is in the latest version of CL, into EMC. But I am not sure which makes more sense; adapt the latest CL version to EMC, or modify the current version of CL which is in EMC?
[02:49:53] <Dave911> Does it matter from a license standpoint? Perhaps by running some diffs I can determine the best approach?
[02:55:21] <cradek> Dave911: you should do whatever you think is most likely to be successful - it sounds like you are the best one to decide.
[02:56:46] <Dave911> OK.. I guess there is no license issues either way since the code can be modified at will.. right?
[02:57:16] <cradek> I haven't checked licensing but I heard someone here say they were compatible
[02:57:48] <Dave911> Yep... CL is LGPL version 2.1
[12:09:01] <JT-Dev> what does this error mean when I do a git pull "fatal: Entry 'docs/src/gcode/main.lyx' not uptodate. Cannot merge."
[12:14:37] <jepler> JT-Dev: you have changes in that file that aren't committed (and maybe that the file is also changed in what you just pulled). commit your changes then pull
[12:14:55] <JT-Dev> jepler: thanks
[12:15:12] <jepler> I did comment to gcode/main.lyx recently to restore the NURBS docs
[12:24:36] <JT-Dev> I'm getting all kinds of errors now with git pull... but time to get ready for work now
[13:44:30] <mozmck-6core> Hi JT-Work, did you get the packages to install?
[13:46:33] <JT-Work> no, I had to reload 10.04 this morning it hosed up my video output somehow so all I had was vertical lines
[13:46:56] <JT-Work> rob got his to work but he downloaded all of the debs and I missed one
[13:47:17] <mozmck-6core> Ok, I see.
[13:48:14] <JT-Work> I'll try again this evening if I get time
[15:11:11] <dgarr> for consideration: http://www.panix.com/~dgarrett/stuff/0001-emc.in-eliminate-message-to-DEBUG_FILE-for-nonexiste.patch
[15:11:29] <dgarr> also: http://www.panix.com/~dgarrett/stuff/0001-integ-add-gain-pin.patch
[15:15:46] <jepler> hmph! .. I distinctly remember writing GetFromIniDefault and using it for [EMC]NML_FILE but I see that it's not on linuxcnc.org.
[15:15:49] <jepler> I wonder where I lost it
[15:18:56] <CIA-1> EMC: 03jepler 07v2.4_branch * r8db75d8ab739 10/scripts/emc.in: emc.in: eliminate message for nonexistent [EMC]NML_FILE
[15:19:00] <jepler> anyone else have a thought on putting integrator gain on v2.4_branch? It looks so safe..
[15:19:45] <dgarr> the analog computers i remember always had a gain pot on the operational amplifier integrators:-D
[15:21:28] <jepler> I assume you have a use for it personally as well
[15:21:53] <dgarr> yes
[15:41:50] <CIA-1> EMC: 03jepler 07v2.4_branch * r427d780dd0f3 10/src/hal/components/integ.comp: integ: add gain pin
[15:41:51] <micges> jepler: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?LockedMemory doesn't help on 10.04 rtai
[15:42:40] <mozmck_work1> micges: change hard to -
[15:43:32] <mozmck_work1> try running "ulimit -Sl 20480" in a terminal and you shouldn't have to reboot...
[15:44:27] <micges> mozmck_work1: thanks I'll test it tomorrow
[15:45:06] <micges> mozmck_work1: your packages works fine, no grub issues
[15:45:27] <mozmck_work1> good!
[19:59:51] <alex_joni> mozmck_work1: one thing I've been meaning to talk to you about
[20:00:21] <alex_joni> I think we (as in you) should rename the kernel built to have a way bigger version number in the name
[20:00:43] <alex_joni> instead of make it or something like that
[20:01:11] <alex_joni> I think that will cause grub to always put it first in the list, even if the user upgrade the regular kernels
[20:03:32] <jepler> YUCK
[20:04:16] <Jymmm> jepler: Just say how you REALLY feel =)
[20:05:32] <mozmck_work1> ok, should I or not then :)
[20:06:06] <alex_joni> jepler: I know.. but there were soo many people bitten by that with the 8.04 package
[20:06:13] <Jymmm> mozmck_work1: Unless jepler comes up with a better solution, I'd go with alex_joni's suggestion.
[20:06:26] <alex_joni> you even added an extra emc2 test for the running kernel vs. compiled
[20:06:31] <mozmck_work1> If you install startup-manager you can select the default kernel
[20:06:33] <jepler> yes, when a user encounters this problem just insult him until he leaves and uses different software instead.
[20:06:43] <Jymmm> ROTFLMAO
[20:06:55] <jepler> Jymmm: you *said* to say how I really feel
[20:07:04] <jepler> bbl
[20:07:05] <Jymmm> jepler: Hey, isn't that the SOP for the last 5+ years?
[20:07:15] <mozmck_work1> heh. then there will be less users to have to support right? :)
[20:07:39] <Jymmm> mozmck_work1: Well, how's that for a, errr, um, "solution"?
[20:07:40] <alex_joni> more users who are worth supporting :P
[20:07:57] <mozmck_work1> I think there is an easy way to set the default kernel in some file. I'll check on that.
[20:10:02] <mozmck_work1> Not saying that I won't make the version number higher, just that there may be another option to consider...
[20:10:03] <alex_joni> mozmck_work1: independent on grub1/2/lilo/whatever ?
[20:10:16] <alex_joni> if there's another option to consider I'm all for it
[20:10:24] <mozmck_work1> uh, probably not. so dump my idea :)
[20:10:40] <alex_joni> the version number is really a hack, and I'm all for a better solution
[20:13:24] <cradek> just instruct the user to uninstall the other kernels
[20:13:35] <cradek> people who need both and want to switch between them can handle it
[20:16:16] <aystarik> what if you have a link to latest -rtai kernel and put it permanently into grub.conf at first place?
[20:16:38] <alex_joni> that means messing with grub.conf which is yucky
[20:17:10] <aystarik> memcheck does that as I understand...
[20:21:58] <aystarik> add one more file into /etc/grub.d/ with 09_ prefix :)
[20:22:32] <jepler> if the grub2 "problem" is gone, then I have no problem with doing something that only works for the default bootloader
[20:23:41] <aystarik> jepler, is it solved by removing plymouth altogether?
[20:24:33] <jepler> aystarik: mozmck told me that some update made grub2 start working reliably with rtai kernels on his systems
[20:24:44] <jepler> I haven't kept track of the details
[20:25:02] <aystarik> I've tried server + gnome core, had not seen any issue too
[20:26:50] <mozmck_work1> plymouth is a dependency of mountall, which was updated recently. The computer here which had the worst booting problems seems to have been fixed. I need to test it on some other machines.
[20:27:20] <mozmck_work1> You can't uninstall plymouth without uninstalling just about everything else.
[20:28:39] <jepler> for example, removing it wants to remove telnet (!?)
[20:29:11] <jepler> hostname
[20:31:38] <jepler> so I dunno what plymouth is, but they've sure decided we WILL have it
[20:33:03] <cradek> Description: graphical boot animation and logger - main package
[20:33:19] <mozmck_work1> :) It wants to remove the whole gnome desktop, xorg server, etc etc...
[20:33:20] <cradek> it's the thing that makes it so you can't see the boot messages
[20:33:53] <mozmck_work1> actually, it does more than that apparently. can't remember what but it's part of the fast boot stuff.
[20:35:11] <aystarik> ok, on a server I have 'plymouth-theme-ubuntu-text', which might make everyone happy...
[20:35:45] <jepler> does that make the boot process look like Linux circa 1995?
[20:36:24] <jepler> * jepler doesn't want people to think he's using a mac or something
[20:36:38] <aystarik> no, it is a black screen for about 5 sec, then X starts.
[20:37:22] <aystarik> it boots to full X under 13 sec..
[20:38:49] <aystarik> BIOS screen stays longer :)
[20:38:56] <mozmck_work1> same here
[20:42:24] <cradek> Plymouth isn't really designed to be built from source by end users.
[20:43:19] <cradek> ... Plymouth isn't done yet. It's still under active development and isn't ready for distros to use as-is.
[20:45:50] <jepler> thank you mark shingleroof