Back
[01:07:28] <adam3999_> hey guys, i posted a note on the forum with my parport issue hoping to reach a wider audience. any help would be appreciated greatly:
http://linuxcnc.org/index.php/english/forum/18-computer/28929-no-input-signals-w-asrock-q1900m-onboard-parport
[09:57:29] <skunkworks> pcw_home,
http://linuxcnc.org/index.php/english/forum/18-computer/28929-no-input-signals-w-asrock-q1900m-onboard-parport#56467
[09:57:33] <skunkworks> ^ a thought...
[09:59:00] <pcw_home> yeah posting the hal/ini files might help also
[10:00:17] <skunkworks> eh - probably not.. The itx version has a physical printer port...
[10:00:27] <skunkworks> (which seems to have the same problem)
[10:01:38] <pcw_home> I have a Asrock miniitx MB and its PP works
[10:02:08] <pcw_home> I dont have the 1900M pro though
[10:02:33] <mozmck> Is there a config option to make axis not prompt every time you exit?
[10:02:53] <pcw_home> Isn't that a new feature
[10:03:11] <mozmck> I think so. It's in 2.7
[10:07:13] <cradek> pcw_home: that thread is getting worse. I sure wish people were better at knowing when they don't know things.
[10:08:04] <pcw_home> Yeah I probably should have resisted posting
[10:08:56] <cradek> the gpl2 is about one page. I can't imagine ranting about it in a public forum without even reading it first.
[10:10:03] <cradek> I suppose people are used to inscrutable license agreements like EULAs you click through while shrugging
[10:10:24] <cradek> I suppose they think they know what's in those, based only on their feelings, too
[10:10:51] <pcw_home> Yeah (Ive done that as well)
[10:14:51] <pcw_home> If I understand it correctly, Tormach is obliged to provide source to their users but not
[10:14:52] <pcw_home> to anyone else, so if the source is on their DVD (is most of it python so its source anyway?)
[10:14:54] <pcw_home> they have fulfilled their obligations
[10:15:42] <cradek> to be more specific: anyone they distribute binaries to
[10:16:36] <pcw_home> Yes, but they are not obliged to distribute binaries to the public
[10:16:45] <cradek> no, certainly not
[10:17:01] <cradek> they can distribute to whoever they want
[10:17:42] <cradek> as soon as they do that, they have to give the source to that person too (or just offer it, there are several ways)
[10:18:23] <cradek> but I bet you know this because you've read the gpl2 and the faq :-)
[10:18:30] <pcw_home> Nor can they restrict re-distribution (AFAIK)
[10:19:02] <cradek> right, all their users have exactly the same rights tormach has, that's the whole point
[10:19:14] <pcw_home> the FAQ is really good for non-lawyers
[10:22:31] <cradek> there should be a little quiz about the faqs on our web forum. it would require you to get the questions right before you could post in a thread containing "GPL" in the title.
[10:22:55] <pcw_home> :-)
[11:31:39] <CaptHindsight> is it possible to hard code accel and speed limits to an axis after the customer discovers how to modify them?
[11:32:35] <seb_kuzminsky> CaptHindsight: linuxcnc gets vel & accel from the ini file, or choose which ini file linuxcnc reads, controls the vel & accel
[11:32:43] <seb_kuzminsky> also: your question worries me
[11:32:47] <CaptHindsight> also, rather than not crashing the machine, the customer now wants extra actuators and parts so that they can replace them :)
[11:33:20] <CaptHindsight> seb_kuzminsky: yes, I'll set limits in the ini, they make changes, break something and then play dumb
[11:33:34] <cradek> this is not a software problem
[11:34:02] <archivist> punitive invoice medicine
[11:34:07] <seb_kuzminsky> heh
[11:34:12] <CaptHindsight> in the works
[11:35:18] <archivist> CaptHindsight, a checksum on the ini files for your protection
[11:36:26] <CaptHindsight> they changed the accel and max speed and stripped the lead screw this time
[11:36:58] <CaptHindsight> last time they snapped the coupling between the stepper and screw shaft
[11:37:10] <seb_kuzminsky> if i was selling a preconfigured machine, i would put the machine config in a debian package, maybe next to our sample configs
[11:37:32] <seb_kuzminsky> then you can use debsums to verify that the files are unchanged
[11:40:28] <archivist> we had a name for that "customer participation"
[11:43:03] <CaptHindsight> about 8-10 months ago I asked about being able to mask out ranges of motion for when one axis should not be moved when other are in certain positions in expectation of their other problem
[11:43:32] <CaptHindsight> trying to prevent them from being able to crash the machine
[11:43:59] <skunkworks> cradek, rob figured out the tapping issue. says says he mistakenly made a function return an int instead of a double.. (that actuall makes sense to me with the way it failed...)
[11:44:10] <CaptHindsight> these guys all have Ph.D's, they are going to crash them no matter what :)
[11:44:10] <skunkworks> at 1
[11:51:15] <mozmck> CaptHindsight: PhD - Post hole Diggers
[11:52:08] <ssi> CaptHindsight: creative use of file permissions and lack of giving them the root password perhaps? :D
[12:09:07] <seb_kuzminsky> skunkworks: that's great :-)
[12:12:58] <CaptHindsight> archivist has the best solution
[12:23:13] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 f7f7f82 06linuxcnc Merge remote-tracking branch 'rob-ellenberg/hotfix/threading-regression-2.7' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=f7f7f82
[12:57:44] <linuxcnc-build> build #3040 of 1202.rip-lucid-amd64 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1202.rip-lucid-amd64/builds/3040 blamelist: Sebastian Kuzminsky <seb@highlab.com>, Robert W. Ellenberg <rwe24g@gmail.com>
[12:58:37] <skunkworks> uh oh
[13:01:16] <linuxcnc-build> build #3050 of 0000.checkin is complete: Failure [4failed] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/3050 blamelist: Sebastian Kuzminsky <seb@highlab.com>, Robert W. Ellenberg <rwe24g@gmail.com>
[13:11:34] <skunkworks> why only lucid
[13:14:24] <cradek> looks like that's another weird linuxcncrsh thing
[13:24:39] <skunkworks> cradek, your test gcode works as expected
[13:25:32] <cradek> yay!
[13:26:12] <skunkworks> magic
[13:33:24] <seb_kuzminsky> the t0 tests fail intermittently like that. i've got a long list of examples from the buildbot, they all looked like dropped mdi commands
[13:33:32] <seb_kuzminsky> not sure if it's linuxcncrsh or task dropping them
[13:33:47] <seb_kuzminsky> haven't made time to look into it deeply
[13:34:13] <seb_kuzminsky> linuxcnc-build: force build --branch=2.7 0000.checkin
[13:34:14] <linuxcnc-build> build forced [ETA 53m45s]
[13:34:14] <linuxcnc-build> I'll give a shout when the build finishes
[14:23:21] <linuxcnc-build> Hey! build 0000.checkin #3051 is complete: Success [3build successful]
[14:23:22] <linuxcnc-build> Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/3051
[17:23:03] <KGB-linuxcnc> 03Dewey Garrett 052.7 ea3c55f 06linuxcnc 10lib/hallib/xhc-hb04.tcl xhc-hb04.tcl: improve assign of coords to switch * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ea3c55f