#linuxcnc-devel | Logs for 2014-11-18

[00:27:26] <KGB-linuxcnc> 03Sebastian Kuzminsky 05master 093b040 06linuxcnc 10src/emc/task/emctaskmain.cc task: reindent for consistency * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=093b040
[02:49:45] <KGB-linuxcnc> 03Norbert Schechner 052.6 bc03663 06linuxcnc 10src/emc/usr_intf/gmoccapy/gmoccapy.glade 10src/emc/usr_intf/gmoccapy/gmoccapy.py 10src/emc/usr_intf/gmoccapy/release_notes.txt gmoccapy_1_3_3 - initialize mouse button mode corrected * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=bc03663
[02:50:27] <KGB-linuxcnc> 03Norbert Schechner 052.7 7cd8b53 06linuxcnc Merge branch '2.6' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7cd8b53
[02:51:13] <KGB-linuxcnc> 03Norbert Schechner 05master 6f96b26 06linuxcnc Merge branch '2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6f96b26
[06:10:45] <KGB-linuxcnc> 03John Thornton 052.6 6d5aa2d 06linuxcnc 10docs/src/gcode/gcode.txt Docs: correct G43 description * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6d5aa2d
[07:46:04] <jepler> seb_kuzminsky: http://emergent.unpythonic.net/files/sandbox/dmesg-in-color.png (with dmesg from 2.25.2-2, probably unrelated to systemd)
[07:46:21] <jepler> also dmesg -H is interesting
[07:46:46] <jepler> [Nov18 07:25] Bluetooth: RFCOMM TTY layer initialized
[07:46:46] <jepler> [ +0.000008] Bluetooth: RFCOMM socket layer initialized
[07:46:46] <jepler> [ +0.000006] Bluetooth: RFCOMM ver 1.11
[08:23:27] <cradek> ha, the real machinist weighs in and says "why not just cut it flat?"
[10:05:00] <archivist> cradek, are you referring to the silly long thread on the bamboo slitter?
[10:05:11] <cradek> heh yeah
[10:05:23] <cradek> it's a fine thread actually
[10:05:44] <cradek> it's just funny how some people want to fix hardware problems in hardware, and others want to fix it in software
[10:05:52] <cradek> you can tell who's a machinist and who's a programmer
[10:05:56] <archivist> is the tooling a slitting saw though , think about it
[10:07:03] <archivist> I do remember seeing some pics of the machine a few years ago
[10:08:28] <cradek> yeah I think it is a saw - not sure how much you can move a saw perpendicular to the cut direction, but I bet he has a good feel for it
[10:09:00] <cradek> sounds like he wants only .003-ish corrections
[10:09:42] <archivist> I still would never move the saw side ways which is what I think he wants
[10:10:21] <archivist> I agree with mechanical solution
[10:10:40] <cradek> yeah I bet everyone thinks that would be best, but not necessarily easiest
[10:11:08] <cradek> get a bunch of appropriate feeler gauge stock and stick it under the whatever parts, done
[10:11:35] <archivist> even adjustable jacks
[10:11:36] <cradek> I used to have a pile of .003 feeler gauges before I fixed my terrible chinese milling vise
[10:11:46] <cradek> (fixed = flycut)
[10:20:06] <seb_kuzminsky> cradek: is that the one that's now my awesome chinese vise?
[10:24:53] <seb_kuzminsky> it bothers me that the pdf docs and html docs have different layout and contents
[10:25:07] <cradek> yeah, iirc, it's quite good now that it's actually flat and square
[10:25:23] <seb_kuzminsky> i bet asciidoc supports multiple levels of include::, so it can be fixed fairly painlessly
[10:25:26] * seb_kuzminsky experiments
[10:25:40] <cradek> it was originally .003 higher on one side, and bumpy
[10:26:26] <cradek> do you call it "awesome" ironically?
[10:26:41] <seb_kuzminsky> it's been great for me, but i dont think i'd notice if it was off by .003, not be a real machinist and all ;-)
[10:26:51] <seb_kuzminsky> no, i love it
[10:27:14] <cradek> a real machinist wouldn't have keyed it to the table, but I sure liked it that way. are the keys still on it?
[10:27:36] <seb_kuzminsky> oh yeah
[10:27:58] <seb_kuzminsky> it's super quick to mount and unmount from the table
[10:29:55] <cradek> the second vise I bought was a kurt, and it probably cost 3x as much, but I didn't have to fix it
[10:30:43] <seb_kuzminsky> you pay with your money or you pay with your time
[10:31:06] <seb_kuzminsky> (or you get lucky like me and buy used from someon who already invested the time!)
[10:33:56] <archivist> I got nice parallel brass jaws on a bench vice because it visited a mill :)
[11:17:43] <archivist> yup plenty of space for feelers to fix that machine, now I have seen the video again
[11:28:25] <cradek> I think he said he's already shimmed it the best he can
[11:28:42] <cradek> I can sure imagine diminishing returns
[11:31:42] <seb_kuzminsky> wow, computers don't behave at all like i remember: http://www.phoronix.com/scan.php?page=news_item&px=MTg0MzU
[11:32:02] <archivist> but it is all on extrusions, they slump, it needs flat references
[11:32:55] <jepler> my OCD would have me constantly trying to make the windows perfectly 90 degree rotated
[11:33:03] <jepler> do. not. want.
[11:34:14] <cradek> the only thing better than a touchscreen keyboard is many touchscreen keyboards
[11:34:55] <jepler> once they abolish typing in computer programming we'll all be better off with just touchscreens
[11:35:25] <jepler> .. say you're trying to fix an off-by-one error in for(i=0; i <= x.length(); i++)
[11:35:34] <jepler> you'd just tap on <= and then use a slider to change it to <
[11:36:04] <jepler> if x is the wrong identifier, you'd just tap on it then tap the correct identifier that pops up next to it
[11:36:28] <archivist> you expect correct guesses!
[11:36:57] <seb_kuzminsky> the building owner at $DAYJOB just sent an email to all tenants asking us not to try to flush little liquor bottled down the toilets because they clog & flood
[11:37:24] <archivist> I am getting fed up of the wildly wrong predictive guessing
[11:37:32] <cradek> seb_kuzminsky: is that a recurring problem?
[11:37:37] <seb_kuzminsky> apparently
[11:37:53] <seb_kuzminsky> my boss just emailed all employees saying he hopes we have enough sense to buy the big bottles
[11:38:16] <cradek> well spoken from a position of privilege
[11:52:52] <skunkworks> it would have been funnier to tell employees to switch to pot..
[11:53:32] <skunkworks> butts don't flood...
[11:54:57] <seb_kuzminsky> we've had cops come and arrest employees from one of the other companies in our building during the work day
[11:55:18] <skunkworks> hmm fun place...
[11:55:32] <seb_kuzminsky> crap, kqt4at5v's computer locks up when he starts the new rtai kernel
[11:56:57] <skunkworks> is the previous hybrid iso out there somewhere still?
[11:57:17] <seb_kuzminsky> lots of folks probably have the old one still
[11:57:42] <jepler> seb_kuzminsky: ugh
[11:57:59] <jepler> skunkworks: it's not explicitly preserved on linuxcnc.org afaik
[11:58:20] <skunkworks> a memcheck still might be in order...
[12:00:23] <seb_kuzminsky> the machine apparently runs fine with 2.6.32-rtai
[12:02:13] <seb_kuzminsky> dgarr had a machine that behaved the same way
[12:03:06] <cradek> "slowly locks up over 1 minute"?
[12:04:44] <seb_kuzminsky> that's odd isnt it
[12:04:53] <pcw_home> that is odd, slowly bogs down?
[12:05:01] <seb_kuzminsky> but latency-test locks up immediately, before showing its window
[12:05:35] <pcw_home> but latency test defaults to 25 KHz base thread which may be way too high
[12:06:00] <pcw_home> 40 KHz I should say
[12:06:23] <seb_kuzminsky> does it?! whoops, that's not very safe
[12:06:32] <jepler> 25000ns
[12:07:12] <jepler> that's been the default for quite a long time
[12:08:42] <pcw_home> I dont think I've had it really lock up even on a a slow machine though since the average latency is going to be much lower
[12:09:41] <seb_kuzminsky> and i guess the latency-test doesn't do hardly anything in that fast thread, just record the latency
[12:10:16] <pcw_home> Yeah so it even mostly works on Preemt-RT
[12:16:02] <skunkworks> dads new acquisition.... DSC02117A.JPG
[12:16:06] <skunkworks> heh
[12:16:18] <skunkworks> http://electronicsam.com/images/KandT/testing/DSC02117A.JPG
[12:16:40] <skunkworks> 84
[12:18:25] <skunkworks> http://electronicsam.com/images/KandT/testing/DSC02122A.JPG
[12:22:34] <seb_kuzminsky> shiny
[12:48:27] <norbert__> Realy lovly small machine ;-), Mine 650 dia x 1500 mm length still waiting for retrofit :-(
[13:20:47] <skunkworks> 12000lbs :)
[13:21:52] <skunkworks> probably replacing the non working Cincinnati..
[13:22:25] <PCW> To me, that is not small...
[13:23:21] <kwallace2> But why would you want a working machine when it's more fun to fix it?
[13:39:16] <skunkworks> I know - right?
[13:40:10] <skunkworks> well - it was working when taken out of service.. so maybe it doesn't....
[13:40:34] <cradek> do you need to make a lot of round things?
[13:41:24] <skunkworks> as of right now - yes..
[13:41:40] <skunkworks> and this has a tail stock also..
[13:42:15] <skunkworks> 20" Max Swing Over Bed, 12" max Cutting Diameter, 40.75" Max Turning Length, 2.5" max Bar Diameter
[13:43:57] <skunkworks> so a nice size :)
[13:44:05] <cradek> that's big bar
[13:45:21] <skunkworks> might save the monarch from being cnc'ed
[15:13:19] <seb_kuzminsky> it's interesting that kq31oijldkqwjd's cpu is also an old AMD, like dgarr's machine that didn't work
[15:15:35] <seb_kuzminsky> dgarr's is: CPU0: AMD Sempron(tm) Processor 2800+ stepping 02
[15:15:49] <seb_kuzminsky> ktqw189's is: CPU0: AMD Athlon(tm) 64 Processor 3700+ stepping 01
[15:16:43] <seb_kuzminsky> both their boot messages complain about the same ACPI error: ACPI Error: A valid RSDP was not found (20120320/tbxfroot-219)
[15:20:21] <cradek> that's unfortunate
[15:33:15] <PCW> the oldelst AMD CPU I have is a Opteron (148) and it works
[15:36:55] <Tom_itx> i'm running a 3200?
[15:37:02] <Tom_itx> i get a few errors with it
[15:37:25] <seb_kuzminsky> i dont know what the amd model numbers mean
[15:37:27] <Tom_itx> but i'm running 2.5
[15:37:30] <seb_kuzminsky> ah
[15:37:44] <Tom_itx> i seldom use that pc but have it in the garage
[15:37:53] <Tom_itx> i generally run my atom 525 on it
[15:37:54] <seb_kuzminsky> i dont know that it's a problem with the amd processor, it could be the bios or some hardware on the motherboard or something else
[15:38:02] <seb_kuzminsky> Tom_itx: that's a sensible thing to do
[15:38:19] <Tom_itx> oh an it's running 8.xx ubuntu
[15:38:32] <Tom_itx> 10.04 wouldn't load on it
[15:39:02] <Tom_itx> i keep it as a backup
[15:39:51] <seb_kuzminsky> haha
[15:48:03] <seb_kuzminsky> if i was an rtai guy i'd totally go find a machine like that and figure out why it doesn't work
[15:48:21] <seb_kuzminsky> but i'm a linuxcnc guy, so i'll just grumpily say "that motherboard's not supported"
[15:48:34] <cradek> I thought jepler found the change that borke it
[15:48:43] <seb_kuzminsky> hmm?
[15:49:37] <seb_kuzminsky> zlog:
[15:51:06] <seb_kuzminsky> sure enough: http://tom-itx.no-ip.biz:81/~tom-itx/irc/logs/%23linuxcnc-devel/2014-06-30.html#07:04:01
[15:51:47] <seb_kuzminsky> someone should build a kernel with jepler's patch
[15:51:50] <seb_kuzminsky> oh wait that's me
[15:52:07] <cradek> and a patch
[15:52:08] <cradek> wow
[15:52:15] <seb_kuzminsky> or brianmorel99....
[15:52:19] <cradek> very competence
[15:52:46] <seb_kuzminsky> heh
[15:55:26] <PCW> I probably shouldn't mention USB doesnt work on my Opteron
[15:56:05] <PCW> of course the MB may be borked since it came out of the junk box
[15:58:31] <seb_kuzminsky> heh
[16:05:56] <PCW> what I do notice is that wheezy runs really well on modern motherboards
[16:07:02] <bjmorel_work> No thanks, i'm working on din rail mounts for my power supply and stepper drivers. Than I get to tear the mill elec. all apart and wire it up properly.
[16:13:33] <seb_kuzminsky> heh cool
[17:05:16] <seb_kuzminsky> not sure what jepler based his patch on (there was a lot of version confusion back then), but the 3.4.55 rtai kernel patch doesnt check for the erratum:
[17:05:19] <seb_kuzminsky> https://github.com/SebKuzminsky/linux-rtai-debian/blob/3.4.55-rtai/patches/features/all/rtai/hal-linux-3.4.55-x86-1.patch#L1443
[17:06:15] <seb_kuzminsky> here's jepler's patch for reference: http://emergent.unpythonic.net/files/sandbox/c1e-nonfatal.patch
[17:06:20] <seb_kuzminsky> so i dont think that's it :-/
[17:12:14] <seb_kuzminsky> the old 3.4.87 patch has the erratum check, but the 3.4.96 one does not
[17:43:53] <Tom_itx> bad thing about my logs is they're not googleable... psha's were
[17:44:54] <Tom_itx> somebody was complaining that what they said in chat could get them in trouble at work and didn't want it public
[17:44:56] <Tom_itx> go figure.
[17:47:18] <jepler> https://freenode.net/channel_guidelines.shtml
[17:47:39] <jepler> >>>
[17:47:39] <jepler> If you're considering publishing channel logs, think it through. The freenode network is an interactive environment. Even on public channels, most users don't weigh their comments with the idea that they'll be enshrined in perpetuity. For that reason, few participants publish logs.
[17:47:41] <Tom_itx> i know what it says
[17:47:44] <jepler> If you're publishing logs on an ongoing basis, your channel topic should reflect that fact. Be sure to provide a way for users to make comments without logging, and get permission from the channel owners before you start. If you're thinking of "anonymizing" your logs (removing information that identifies the specific users), be aware that it's difficult to do it well—replies and general context often provide identifying info
[17:47:50] <jepler> If you just want to publish a single conversation, be careful to get permission from each participant. Provide as much context as you can. Avoid the temptation to publish or distribute logs without permission in order to portray someone in a bad light. The reputation you save will most likely be your own.
[17:47:55] <jepler> <<<
[17:48:28] <jepler> notes that the channels are logged seem to have fallen out of our topics over time
[17:49:28] <andypugh> The log is super-useful in our case for checking what you missed when offline.
[17:49:58] <seb_kuzminsky> i am pro-log as well
[17:50:55] ChanServ changed topic of #linuxcnc-devel to: http://linuxcnc.org | Latest release: 2.6.4 | (this channel is logged)
[17:51:02] <Tom_itx> most users don't weigh their comments with the idea that they'll be enshrined in perpetuity
[17:51:11] <Tom_itx> could be said of facebook, twitter, etc
[17:51:13] <Tom_itx> as well
[17:51:16] <jepler> thank you whoever did that
[17:51:23] ChanServ changed topic of #linuxcnc-devel to: http://linuxcnc.org | Latest release: 2.6.4 | (this channel is logged by the zlog robot)
[17:51:27] * seb_kuzminsky winks
[17:51:41] <Tom_itx> also mah
[17:51:45] <Tom_itx> logger[mah]
[17:51:45] <logger[mah]> Tom_itx: Log stored at http://linuxcnc.mah.priv.at/irc/%23linuxcnc-devel/2014-11-18.html
[17:51:49] <seb_kuzminsky> zlog: now *you're* enshrined in perpetuity
[17:54:54] <Tom_itx> i added a bot.txt because of someone's complaint
[17:55:22] <Tom_itx> 51
[17:55:31] <Tom_itx> 51" of snow in NY
[17:55:45] <Tom_itx> 5/hr
[17:56:43] <seb_kuzminsky> thanks or running zlog, that's super useful
[17:57:54] <Tom_itx> np, it's been there quite a while
[18:00:37] <seb_kuzminsky> i always used psha's logger until it broke just now
[18:00:55] <Tom_itx> looks like his whole site is broke
[18:01:20] <seb_kuzminsky> i hope he's okay
[18:01:26] <seb_kuzminsky> haven't seen him around lately, i think
[18:01:29] <andypugh> Hope he’s OK, I haven’t heard anything for a while
[18:01:43] <seb_kuzminsky> great minds think alike
[18:02:42] <seb_kuzminsky> andypugh: i dig your heat treat furnace
[18:04:14] <andypugh> It works nicely, and through the miracle of eBay was only about £50 all-in
[18:04:25] <seb_kuzminsky> cool
[18:04:42] <seb_kuzminsky> i've been collecting parts for a furnace for a while, but haven't gotten around to putting it together yet...