#emc-devel | Logs for 2010-05-25

[07:18:30] <e-jones> e-jones is now known as e-jones|away
[09:10:12] <e-jones|away> e-jones|away is now known as e-jones
[15:38:20] <cradek> what do you guys think about removing all the stepper sample configs?
[15:41:05] <SWPadnos> heh
[15:41:06] <aystarik> good idea. I've seen one person trying to run it in simulation...
[15:42:23] <jepler> as far as stepper/ goes there's some documentation-ish stuff in standard_pinout.hal, but I think that kind of thing is better covered elsewhere (and it may be all stuff stepconf does anyway)
[15:42:46] <jepler> ditto for stepper-xyza
[15:42:51] <jepler> not so sure about stepper-gantry
[15:43:23] <jepler> Sherline* should probably remain since as far as we know they are applicable to specific machines that are on the market
[15:43:35] <cradek> yeah
[15:43:37] <jepler> nist-lathe? not sure.
[15:43:43] <jepler> demo_step_cl? not sure.
[15:43:47] <cradek> this may not be the time to clean up stuff anyway.
[15:43:53] <cradek> forget I said anything.
[15:44:01] <jepler> when would be the time? I assume you're talking about master anyway..
[15:44:20] <cradek> I meant 2.4.
[15:44:20] <SWPadnos> what's the real benefit?
[15:44:28] <cradek> people go directly to stepconf
[15:44:46] <jepler> oh -- no, I'm not going to do something like that in 2.4!
[15:45:28] <SWPadnos> maybe the thing to do is make a single stepper-based sample config, with very very well documented ini and hal files
[15:45:50] <SWPadnos> including ini options that aren't used, and explanations of exactly why each thing is being done
[15:45:55] <cradek> I don't think those really help anyone
[15:46:16] <jepler> the documentation is the documentation.
[15:46:21] <SWPadnos> sure
[15:46:37] <SWPadnos> most people will use stepconf, so anything else is useless to them
[15:47:14] <SWPadnos> others will want to learn what's going on, and documentation is a great way to do that, but isn't "real world" enough to follow easily
[15:47:45] <cradek> I think yet another kind of documentation isn't going to help that
[15:47:48] <SWPadnos> Jymmm mentioned this idea, it's like the very well commented ini files for apache or similar - you can learn what you need to change by reading the config file
[15:48:05] <SWPadnos> I'm not sure
[15:48:06] <cradek> like the recent "but I want howtos, not documentation" on the list
[15:48:16] <SWPadnos> heh, well that's just different :)
[15:48:55] <SWPadnos> the documentation has all the information (presumably), which has a different use than an example
[15:49:51] <SWPadnos> so I'd say sure, let's remove all but one well-documented stepper config, but I'd prefer to not lose all the case-study based documentation
[16:03:38] <skunkworks_> too many people don't give it a good try... They want to be spoon fed.
[16:06:25] <skunkworks_> btw - I was hooking up pins to the mesa gpio to show dad the buttons that appear in axis... He thought that was cool.. Getting more excited.
[16:09:59] <SWPadnos> yay! my old laptop is actually 8.04, and even has the RTAI kernel on it
[16:21:12] <SWPLinux> so, I get this interesting error when upgrading from emc2.2 to emc2.4 (both emc2 and emc2-dev):
[16:21:30] <SWPLinux> E: /var/cache/apt/archives/emc2-dev_1%3a2.4.0_amd64.deb: trying to overwrite `/usr/share/man/man1/comp.1.gz', which is also in package emc2
[16:22:40] <SWPLinux> ah, there's no 2.4 version of emc2-dev
[16:23:14] <SWPLinux> err, nevermind that last
[16:27:50] <SWPLinux> huh. and on the second try it worked
[16:28:28] <jepler> that's something to do with how the comp manpage got (correctly) moved from the main package to the deb package
[16:28:32] <SWPLinux> so I assume the comp manual was moved from emc2 to emc2-dev, and apt just didn't do it right
[16:28:34] <jepler> did it actually make the install fail?
[16:28:35] <SWPLinux> yep, ok
[16:28:37] <SWPLinux> yes
[16:28:39] <jepler> ugh!
[16:28:50] <SWPLinux> emc2-dev didn't install the first time, but did the second time
[17:52:13] <jepler> I hate the sf website
[18:31:36] <Jymmm> jepler: ?
[18:32:31] <Jymmm> jepler: That's pretty broad, can you be more specific?
[18:33:44] <jepler> to name a few things: it's slow, the organization puts what I want always at least 2 clicks from the project page, and the issue tracker lacks the very obvious "go to report by number, no matter the project or artifact type"
[18:34:43] <Jymmm> jepler: Can you give me an example link?
[18:36:07] <jepler> Jymmm: OK, your task is to find whether tracker item 2917663 is closed; if it is, give me a link to the change that closed it.
[18:36:34] <Jymmm> ok, link to project?
[18:37:47] <jepler> Jymmm: on sf no two projects have the same tracker item number, so you *should* be able to go to the report without knowing the project
[18:37:57] <jepler> but anyway the project is tktoolkit
[18:38:54] <Jymmm> http://sourceforge.net/tracker/index.php?func=detail&aid=2917663&group_id=12997&atid=312997
[18:39:58] <Jymmm> jepler: Is that what you were looking for?
[18:40:46] <jepler> Jymmm: what sequence did you use to get there? I had to use 'support', then in each tracker try entering the number to see if there was a search result.
[18:43:55] <Jymmm> I went to http://tktoolkit.sf.net/ which redirected me to http://tcl.sourceforge.net/ then where it says "Directly jump to a project's bug or patch by number: " (half way down the page) I entered in 2917663 and pressed ENTER.
[18:44:23] <jepler> Jymmm: ah, so you solved the problem by not using the sf site itself, but something the tcl people kindly built to make up for a defiency in it
[18:45:36] <jepler> that's lateral thinking at its best
[18:46:05] <Jymmm> jepler: Well, no not really. You could have easily just done: http://sourceforge.net/support/tracker.php?aid=2917663
[18:47:56] <jepler> Jymmm: yes, that's another thing you can do to go around the sf website
[18:48:39] <Jymmm> jepler: Or just ask in #sourceforge http://sourceforge.net/apps/trac/sourceforge/wiki/Support%20by%20IRC
[18:50:59] <Jymmm> jepler: I'm not necessarily defending them, just saying there are alternatives available
[18:51:04] <jepler> sure
[18:51:44] <jepler> I think it underscores what I was saying, that their website is no good
[18:51:54] <jepler> or at least has some specific rough corners that I always hit
[18:53:20] <jepler> (that link into trac points out that sf doesn't even use its own bug tracker anymore, they love it so much)
[18:53:27] <jepler> http://sourceforge.net/apps/trac/sourceforge/report/1
[18:54:45] <Jymmm> jepler: Well, I know the developers and the infrastructure, it's somewhat narrow minded. They haven't used their tracker in years. It's mostly Trac now.
[18:57:01] <cradek> is this cool or stupid or somewhere in between? http://timeguy.com/cradek-files/emc/newsplash.png
[18:57:50] <Jymmm> cradek: http://www.emc-squared.net/
[18:58:06] <jepler> cradek: does it retain the easy scaling of the current version?
[18:58:56] <cradek> if we like it, I'll make sure it does
[18:58:59] <Jymmm> Is it Axis v2.4 or EMC v2.4?
[18:59:40] <cradek> the version number applies to the combination
[19:00:43] <Jymmm> I'd drop the ^2 and leave everything else then.
[19:01:52] <Jymmm> OR... EMC [v2.4 diagonally] AXIS
[19:02:29] <Jymmm> replacing the fowardslash with the diagonal v2.4
[19:02:52] <Jymmm> eh, just a thought =)
[19:03:01] <cradek> Jymmm: hm, I thought we had an existing logo with ^2, but I'm not finding it on anything
[19:03:32] <Jymmm> cradek: Might be a copyright infringement too
[19:03:59] <cradek> meh
[19:04:19] <jepler> cradek: the emc 2.2 splash did
[19:04:25] <jepler> origin/v2_2_branch:emc2.gif
[19:05:10] <cradek> I was tempted to say EMC/AXIS [rotated]v2.4[/rotated]
[19:05:22] <cradek> I'm never sure how the 2 works
[19:06:11] <Jymmm> What about two lines, with the 2nd being "Enhanced Machine Controller v 2.4"
[19:06:45] <cradek> Jymmm: it turns out a lot of people cut this in stuff, so it has to be reasonably small
[19:07:08] <cradek> well the features should be big so you don't need special tooling, but the overall thing should not be too huge: this means you don't get many letters to work with
[19:07:13] <Jymmm> 2nd line: Version 2.4
[19:07:30] <Jymmm> or 2nd line: http://LinuxCNC.org/
[19:07:49] <Jymmm> self promotion is always a good thing =)
[19:08:12] <Jymmm> bbiab
[19:15:35] <Dave911> Re: removing stepper example configs .. Please don't remove any of them! I have looked at many of them and extracted various bits of info out of them. I even looked at the NIST lathe example recently and found something, but I don't remember what it was. The combination of
[19:15:37] <Dave911> the sample configs and the stepconf program is a good mix. :-)
[19:22:10] <Dave911> Re: Modbus... I sent Chris Morley some revised Classic Ladder Modbus RTU/TCP related source files via the forum for him to test. I tested all of the baud rates and all of the functions and I can't get it to error any longer. I also tested the TCP version with all of the functions and that still works fine. I believe the comm issues are fixed. Modbus RTU Serial is now very quick.
[19:24:56] <jepler> it's better to send as patches prepared with 'git format-patch' and to the mailing list instead of in private
[19:27:34] <Dave911> I put it in the public forum.. anyone can download them.. I was concerned about the format thing when making a patch file... do you use git to create the patch? I can read up on that and send them to this email list .. do you mean the emc-dev email list or the emc email list?
[19:28:20] <SWPadnos> emc-dev list
[19:28:41] <Dave911> ok
[19:28:53] <SWPadnos> the forum exists solely for end-user support. I don't know that any developer reads it with any regularity
[19:31:03] <Dave911> gotcha ... I was going back and forth with Chris about it and he wants to test the routines .. I'll see about making a git patch file and send it to the list.
[19:31:11] <jepler> use 'git commit' (or git gui) to create a commit with accurate authorship information and a description of your change, then 'git format-patch -1' to turn that newest commit into a text patch, then attach the message to an e-mail to the emc-developers list. (if you really trust your mailer you can inline it, but most folks use mailers that will mangle a patch when sent inline)
[19:31:22] <SWPadnos> yeah, that kind of discussion is 100 times better on the mailing list
[19:31:43] <cradek> http://timeguy.com/cradek-files/emc/splash2.png
[19:32:11] <jepler> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Git#Share_your_changes_with_others
[19:32:30] <Dave911> OK... that sounds easy .. thanks .. I'll do that..
[19:32:40] <cradek> with block delete you can toggle the second line on and off
[19:32:59] <cradek> (wonder if that's a good thing?)
[19:33:35] <Dave911> I like the splash :-)
[19:34:01] <SWPadnos> re: blovk delete, I'd say it's not a good thing - it's one more variable
[19:34:06] <SWPadnos> err, block
[19:34:21] <Jymmm> cradek: Maybe drop the http:// and jsut have "LinuxCNC.org" (case preserved)
[19:35:10] <cradek> 14:07 < Jymmm> or 2nd line: http://LinuxCNC.org/
[19:35:24] <cradek> when you and you decide what you want, let me know :-)
[19:35:39] <Jymmm> cradek: Yeah I know. But seeing it, the http:// looks cluttered imo
[19:35:52] <cradek> it might be a bad font - it all looks cluttered
[19:35:58] <Jymmm> lol
[19:52:25] <skunkworks_> Can the axis preview have clickable url's? ;)
[19:57:05] <cradek> much cleaner looking: http://timeguy.com/cradek-files/emc/splash3.png
[19:59:55] <cradek> hm, the gcode is 10x as big as before...
[20:00:24] <alex_joni> I like lowercase "emc2.4"
[20:00:57] <alex_joni> cradek: maybe "www.linuxcnc.org" instead of the http://.. ?
[20:01:11] <cradek> hm, that's probably a good idea
[20:05:50] <cradek> http://timeguy.com/cradek-files/emc/splash4.png
[20:08:28] <alex_joni> better
[20:13:28] <cradek> fbofw...
[20:13:28] <CIA-1> EMC: 03cradek 07v2.4_branch * re66c1f9661d2 10/configs/sim/ (axis_9axis.ini lathe.ini): These need separate var files.
[20:13:39] <CIA-1> EMC: 03cradek 07v2.4_branch * r0d01a2294fea 10/share/axis/images/axis.ngc: new splash screen for 2.4
[20:24:26] <ries_> ries_ is now known as ries
[23:19:05] <ries> cradek: if somebody would try to make this file, would it turn the spindle on? the current version of 2.3 doesn't do that I believe...