#linuxcnc-devel | Logs for 2015-10-25

[09:24:12] <skunkworks> new familly member.. http://electronicsam.com/images/house/stellakitty.jpg
[09:46:32] <pcw_home> Almost brought home a cute Calico Lop yesterday from the feed store
[09:46:33] <pcw_home> but it turns out the "no more rabbits" injunction is still in force
[09:54:44] <JT-Shop> I need a goat to mow the weeds
[09:55:16] <Tom_itx> you could have milk too
[09:58:52] <skunkworks> he appeared in front of us on a bike ride..
[09:59:47] <pcw_home> ahh a foundling
[10:00:35] <pcw_home> Our sheep did a good job this year, because of the drought we surely
[10:00:36] <pcw_home> would have got a weed abatement notice by now but the hillside is quite bare
[10:39:10] <jepler> hooray, we're not running a new enough joomla to be affected by the bug that was covered on slashdot today http://it.slashdot.org/story/15/10/25/1244239/joomla-sql-injection-flaw-affects-millions-of-websites
[10:39:18] <jepler> > The flaw first appeared in Joomla 3.2, released in November, 2013
[10:48:42] <seb_kuzminsky> jepler floats like a butterfly
[10:58:08] <archivist> looking at the blog article I would say the problem is more deep seated in joomla
[11:08:50] <pcw_home> https://xkcd.com/327/
[11:14:03] <seb_kuzm1nsky> go home science, you're drunk: https://www.youtube.com/watch?v=u-fbBRAxJNk
[11:14:18] <seb_kuzm1nsky> bbl
[11:15:55] <archivist> or http://www.archivist.info/addedsql.jpg
[11:24:19] <pcw_home> :-)
[11:39:39] <jepler> archivist: nice
[12:15:34] <micges> skunkworks: re mach loop speed; I've read somewhere that 50ms resolution is guaranteed by standard timer on windows xp
[12:58:13] <seb_kuzminsky> ok, the buildbot's back
[12:58:23] <seb_kuzminsky> now: with 90% less dustbunnies
[12:58:42] <andypugh> seb_kuzminsky: Ever read the “User Friendly” webcomic?
[13:03:36] <seb_kuzminsky> linuxcnc-build: force build --branch=master 0000.checkin
[13:03:37] <linuxcnc-build> build #3577 forced
[13:03:37] <linuxcnc-build> I'll give a shout when the build finishes
[13:03:43] <seb_kuzminsky> andypugh: i dont think so, got a link?
[13:03:44] <seb_kuzminsky> bbl
[13:07:02] <andypugh> http://www.userfriendly.org/archives.html (one of the main characters is “The Dust Puppy” found in a server.
[13:09:31] <andypugh> In fact, his first appearance: http://ars.userfriendly.org/cartoons/?id=19971203
[13:32:17] <ssi> andypugh: I used to talk to one of the girls that's depicted in that comic quite a bit
[13:33:47] <andypugh> Small world!
[13:34:08] <ssi> wait I'm thinking of a different comic sry
[13:34:30] <ssi> same era of comic anyway :D
[13:34:30] <andypugh> Interesting to read, it starts off ina world where K56Flex hasn’t quite caught on yet.
[13:34:48] <ssi> UF was a big deal in my early linux days
[13:34:54] <andypugh> That one is still going
[13:37:54] <andypugh> I think LinuxCNC should go back to Ubuntu. “Wily Werewolf” is a cool name.
[13:52:09] <linuxcnc-build> Hey! build 0000.checkin #3577 is complete: Success [3build successful]
[13:52:09] <linuxcnc-build> Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/3577
[14:45:13] <jepler> linuxcnc-build: dance
[14:45:14] <linuxcnc-build> <(^.^<)
[14:45:15] <linuxcnc-build> <(^.^)>
[14:45:16] <linuxcnc-build> (>^.^)>
[14:45:17] <linuxcnc-build> (7^.^)7
[14:45:18] <linuxcnc-build> (>^.^<)
[15:30:00] <andypugh> We seem to have a confused new user. email “cncservo@…” and username “stepmotor”
[15:39:48] <zeeshan|2> :D
[15:40:59] <archivist> or just undecided :)
[21:24:08] <jepler> http://hackaday.com/2015/10/18/fpgas-for-the-raspberry-pi/ -- "hat" with Lattice iCE-HX8K and open source toolchain based on reverse engineering of the FPGA's bitstream
[21:32:51] <jepler> somebody ported xenomai to that 64-bit arm board I have. https://www.96boards.org/forums/topic/announcement-xenomai-on-410c/
[21:36:21] <andypugh> It’s been several days, where is the LinuxCNC port?
[21:37:01] <andypugh> (More seriously, that probably makes the Machinekit guys happy)
[21:37:14] <jepler> possibly. 64 bit machines don't really have much going for them when it comes to cnc
[21:37:38] <jepler> though 64-bit ARMs stand a slightly higher chance of not garbling doubles between threads like the 32-bit ARMs do
[21:38:23] <andypugh> We do use 64-bit types in several important places, but only to make sure we don’t roll over this millenium
[21:38:32] <jepler> sure
[21:38:41] <jepler> you could do that on an 8-bit CPU depending on the performance requirements
[21:38:52] <andypugh> Otherwise, yes, it could all be 16 bit really.
[21:39:23] <andypugh> I lost time chaning from 8bit to 16 bit, and you won :-)
[21:40:34] <andypugh> Is there any fundamental reason that an 8-bit CPU can’t have a 64-bit memory bus?
[21:41:12] <andypugh> I have worked with Z80 and paged memory, but that was a kludge
[21:41:19] <jepler> it's probably just a bad idea, not impossible
[21:42:22] <skunksleep> jepler: did the zenbot move?
[21:42:34] <jepler> most 8-bit CPUs have 16-bit addressing with an 8-bit data bus, right?
[21:43:14] <andypugh> I do recall that on the z80 there was an interupt routine that could be set to call the code at a specific memory location. But on the Sinclair machines that turned out to take the MSB from a register and the LSB from the address bus. So you could choose your 256byte block to be called, but not where in the block
[21:44:05] <andypugh> Now, you might think that was a show-stopper. And when I read the docs, that was my assumption
[21:44:38] <jepler> you just need to waste a whole page with a NOP slide, right?
[21:45:03] <andypugh> But someone with more determination than me noticed that FF was a jump. And FFFFFF was jump to FFFF…
[21:45:32] <andypugh> NOP slide might have worked too, now you mention it.
[21:45:35] <jepler> ah that's better than taking a variable amount of time to reach the real isr
[21:46:43] <andypugh> I can’t recall what you had to put at FFFF, it could clearly only be a single byte
[21:47:24] <andypugh> Maybe it was FEFE? It’s been a few decades
[21:49:28] <andypugh> Anyway, there is nothing to be gained from reminscing about the old days when I could completely understand a computer.
[21:49:45] <andypugh> And a lot more to be gained from some sleeping.
[21:49:56] <andypugh> Night all
[21:51:43] <jepler> night
[21:57:06] <jepler> you geek trapped me but I failed to chase down the details :-/
[21:57:10] <jepler> time to call it a night here too
[21:59:28] <KGB-linuxcnc> 03Dewey Garrett 05master 9511573 06linuxcnc 10scripts/sim_pin sim_pin: make ivalue name consistent * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9511573
[21:59:28] <KGB-linuxcnc> 03Dewey Garrett 05master bf578d3 06linuxcnc 10docs/man/man1/sim_pin.1 10docs/src/hal/tools.txt 10scripts/sim_pin sim_pin: support cmdline mode for bit items * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=bf578d3
[22:32:09] <seb_kuzminsky> i tested out g81 in 2.7 by drilling the 14(!) mounting holes for a 7i43 + 7i54 in the scrounged enclosure for the robot arm, all worked as expected
[22:32:18] <seb_kuzminsky> though the tests are more thorough than i
[22:41:59] <Tom_itx> did it work as advertised?
[22:42:59] <Tom_itx> err no.. it was the G84 i was wondering about a while back
[22:45:29] <Tom_itx> yeah G81 works fine here
[22:51:04] <Tom_itx> as does G83