Back
[00:03:38] -!-
HDB10 has quit [Read error: Connection reset by peer]
[00:04:03] -!-
theorbtwo has quit [Ping timeout: 246 seconds]
[00:04:07] theorb is now known as
theorbtwo
[00:14:11] -!-
robh__ has quit [Ping timeout: 276 seconds]
[00:17:02] <mhaberler> a fundamental question: why is the g-code interpreter not part of the UI, but a part of the core (task)? it would occur to me that having a 'canon machine' (task) and different languages driving it would be a more natural division
[00:17:15] -!-
rooks has quit [Quit: So long, and thanks for all the fish.]
[00:17:54] <mhaberler> (I'm not announcing a coding project - I just want to find out)
[00:24:00] -!-
adb has quit [Ping timeout: 246 seconds]
[00:24:08] <alex_joni> the interpreter is meant to be interchangeable
[00:25:15] <alex_joni> even if it's linked in to task
[00:25:35] <alex_joni> not hot-swappable, but you could compile different binaries with different interpreters
[00:25:52] -!-
adb [
[email protected]] has joined #emc-devel
[00:25:53] <alex_joni> this also allows multiple GUIs to work (I think)
[00:26:45] <jepler> in fact I have a branch somewhere that dynamically loads the interpreter and has a dummy non-rs274ngc interpreter (canterp). branch plug-interp at git://git.unpythonic.net/emc2-jepler.git
[00:26:57] <mhaberler> the reason why I'm asking this: since interp is in task, one basically has to emulate that in the UI for preview - why not just have one
[00:26:59] <jepler> I forget exactly what shape it was in when I got bored of it
[00:27:31] <jepler> it's not emulated, it's the same interpreter but different code to receive the canonical machine calls
[00:28:01] <mhaberler> but it has local _setup in the UI, riight?
[00:28:08] <alex_joni> right
[00:28:26] <alex_joni> otoh you have lots of UIs in strange languages in which it's not trivial to link the interpreter in
[00:28:39] <alex_joni> like tkemc/mini and halui
[00:28:51] <alex_joni> and there was a java GUI, etc
[00:29:08] <mhaberler> well but that was before the decision was made, I'm just curious
[00:29:43] <alex_joni> I think the early GUIs had all no preview
[00:29:53] <alex_joni> so there was no need for the interp in the UI
[00:30:00] <mhaberler> that is a good point!
[00:30:17] <jepler> emc1 had a standalone program that linked the interpreter and could preview gcode
[00:30:30] <alex_joni> the first preview was in AXIS I think
[00:30:33] <jepler> but none of the UIs had that capability
[00:30:44] <alex_joni> tkemc and mini had a backplot
[00:30:51] <alex_joni> err.. have :D
[00:30:57] <jepler> yes, preview plot and "look sort of like a desktop app" were the two main goals of axis
[00:31:28] <jepler> originally axis had an independent implementation of gcode, but it soon became clear that was a bad idea.
[00:31:44] <alex_joni> mhaberler: so maybe a fresh 2011 design would have it only in the UI
[00:31:48] <jepler> is there a specific problem that's on your mind, that makes you ask about the _setup structure and separate processes?
[00:31:55] <alex_joni> although it leaves the problem of syncing multiple UIs
[00:32:10] <mhaberler> being a control freak, I would have ripped the interp out of task and out in the UI
[00:32:32] <alex_joni> mhaberler: yeah, but still.. what for headless instalations?
[00:32:58] <mhaberler> no specific problem - I just see some issues and try to understand how it came about
[00:33:36] <mhaberler> headless translates into 'background job' for me
[00:34:40] <mhaberler> job drives task, done?
[00:34:52] -!-
pjm has quit [Ping timeout: 250 seconds]
[00:36:52] <mhaberler> real reason: the gladevcp thing reaching deeper into the ui part beyond wiggling and displaying pins - questions on how to access interp state crop up
[00:37:48] <mhaberler> I'm just looking at it from a 'which kind of information do I need where' kind of view, and things fall a bit short on communication mechanisms here
[00:43:58] <mhaberler> I confess it was an esoteric issue . it's 1:40pm and I'll crash
[00:44:13] <mhaberler> thanks for llistening & cu
[00:44:43] -!-
JT-Shop has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.15/20110303024726]]
[00:45:03] -!-
mhaberler has quit [Quit: mhaberler]
[00:46:00] <alex_joni> hmm, he has a point there
[00:46:12] <alex_joni> it really is late
[00:46:24] <alex_joni> night all
[00:53:25] -!-
EDocToor has quit [Read error: Connection reset by peer]
[01:01:30] -!-
servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[01:13:18] -!-
seb_kuzminsky has quit [Ping timeout: 252 seconds]
[01:13:30] -!-
andypugh has quit [Quit: andypugh]
[01:23:49] -!-
EDocToor [
[email protected]] has joined #emc-devel
[01:28:00] cryptopsy is now known as
aborticide
[01:32:28] -!-
adb has quit [Ping timeout: 250 seconds]
[01:39:10] -!-
seb_kuzminsky [
[email protected]] has joined #emc-devel
[01:45:46] -!-
adb [
[email protected]] has joined #emc-devel
[02:08:35] -!-
PCW has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014]]
[02:11:22] -!-
seb_kuzminsky has quit [Ping timeout: 255 seconds]
[02:20:10] -!-
Tom_L has quit []
[02:26:30] -!-
seb_kuzminsky [
[email protected]] has joined #emc-devel
[02:42:30] -!-
seb_kuzminsky has quit [Ping timeout: 276 seconds]
[02:44:20] -!-
adb has quit [Ping timeout: 276 seconds]
[02:45:24] -!-
adb [
[email protected]] has joined #emc-devel
[02:59:06] -!-
tlab has quit [Quit: Leaving]
[03:46:05] -!-
bdale has quit [Ping timeout: 276 seconds]
[03:46:58] -!-
JustinXJS has quit [Quit: Leaving]
[03:57:26] -!-
seb_kuzminsky [
[email protected]] has joined #emc-devel
[04:12:50] -!-
seb_kuzminsky has quit [Ping timeout: 240 seconds]
[04:16:20] -!-
EDocToor has quit [Remote host closed the connection]
[04:29:38] -!-
seb_kuzminsky [
[email protected]] has joined #emc-devel
[04:41:56] -!-
seb_kuzminsky has quit [Ping timeout: 252 seconds]
[04:54:55] -!-
seb_kuzminsky [
[email protected]] has joined #emc-devel
[05:12:08] -!-
psha [
[email protected]] has joined #emc-devel
[06:04:22] -!-
EDocToor [
[email protected]] has joined #emc-devel
[06:12:58] -!-
ve7it has quit [Remote host closed the connection]
[07:25:47] -!-
Connor has quit [Ping timeout: 248 seconds]
[07:29:23] <CIA-5> EMC: 03cmorley 07master * r7867abd8c60e 10/src/emc/usr_intf/stepconf/stepconf.py: stepconf - add checks for realtime kernel before axis tests
[07:29:34] <CIA-5> EMC: 03cmorley 07master * r7400795ab590 10/src/emc/usr_intf/stepconf/stepconf.py: stepconf - switch from md5 to hashlib module
[07:33:32] -!-
kb8wmc has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.14/20110221055140]]
[07:34:16] -!-
seb_kuzminsky has quit [Ping timeout: 252 seconds]
[07:36:18] -!-
mhaberler [
[email protected]] has joined #emc-devel
[07:39:00] -!-
mhaberler has quit [Client Quit]
[07:59:04] -!-
psha has quit [Quit: Lost terminal]
[08:43:14] -!-
adb has quit [Read error: Operation timed out]
[08:49:54] -!-
Dannyboy has quit [Remote host closed the connection]
[09:18:02] -!-
robh__ [
[email protected]] has joined #emc-devel
[09:42:57] -!-
Connor has quit [Ping timeout: 246 seconds]
[10:20:16] -!-
davec has quit [Client Quit]
[10:40:34] -!-
adb [
[email protected]] has joined #emc-devel
[11:07:01] -!-
adb has quit [Ping timeout: 260 seconds]
[11:07:44] -!-
adb [
[email protected]] has joined #emc-devel
[11:20:40] -!-
mk0 has quit [Read error: Connection reset by peer]
[11:31:33] -!-
HDB10 has quit [Quit: Ex-Chat]
[11:59:35] -!-
jthornton has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.14/20110221055140]]
[12:02:08] -!-
jthornton [
[email protected]] has joined #emc-devel
[12:17:04] -!-
cevad has quit [Ping timeout: 264 seconds]
[12:27:31] -!-
tris has quit [Ping timeout: 260 seconds]
[12:40:36] -!-
pingufan has quit [Remote host closed the connection]
[12:47:19] -!-
mikegg has quit [Quit: Visitor from www.linuxcnc.org]
[12:56:41] -!-
skunkworks has quit [Ping timeout: 260 seconds]
[13:05:45] -!-
kvirdi_ has quit [Read error: Connection reset by peer]
[13:08:42] -!-
kvirdi_ has quit [Read error: Connection reset by peer]
[13:19:26] -!-
adb has quit [Ping timeout: 260 seconds]
[13:25:40] -!-
Tom_L has quit []
[13:32:53] -!-
micges_cnc [
[email protected]] has joined #emc-devel
[13:36:32] -!-
x_ [
[email protected]] has joined #emc-devel
[13:37:49] -!-
kvirdi_ has quit [Read error: Connection reset by peer]
[13:39:39] -!-
micges_cnc has quit [Ping timeout: 248 seconds]
[13:40:40] x_ is now known as
micges_cnc
[13:42:37] -!-
cblack001 [
[email protected]] has joined #emc-devel
[13:42:57] -!-
skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.68.115.41.210] has joined #emc-devel
[13:43:44] -!-
Tom_itx has quit [Ping timeout: 246 seconds]
[13:53:11] -!-
mk0 has quit [Quit: Leaving]
[14:05:34] <cblack001> Does anyone know who is working on the SPI for Mesa's cards and where it's development is at. I would love to use their 7i65 card and am willing to help give a hand.
[14:08:56] <cradek> cblack001:
http://git.linuxcnc.org/gitweb?p=emc2.git;a=commit;h=edd9e8d5443a81b1328fbfe9f9f70cef6c0a86af
[14:14:00] <cradek> er, I misread 7i64/65
[14:14:08] <cradek> brb
[14:15:21] <cblack001> I assume sebastian is still the one to talk to....
[14:19:17] -!-
adb [
[email protected]] has joined #emc-devel
[14:21:13] <cradek> I think andy is more interested in the spi than seb
[14:21:33] <cradek> either way, lurk here and you'll find the right person.
[14:22:12] <cradek> did you know there's already support for the 6 axis mux (not spi) card?
[14:22:45] -!-
kvirdi_ has quit [Read error: Connection reset by peer]
[14:23:49] -!-
kvirdi_ has quit [Client Quit]
[14:30:39] <cblack001> I was hoping to use the 7i65 mainly because of the analog inputs.
[14:31:07] <cblack001> Everything I need on one card is nice..
[14:41:58] -!-
skunkworks has quit [Quit: Page closed]
[14:42:57] <cradek> oh cool - I didn't know there was one with analog in.
[14:43:06] <cradek> what is your application where you want analog in?
[15:00:27] <cblack001> I built an prototype 18" drill to attach to a CNC oxy-fuel for a customer, it was all custom programmed and had analog inputs from the servo amps with %torque feedback. I would like to move it over to EMC.
[15:00:35] -!-
mk0 has quit [Quit: Leaving]
[15:01:06] -!-
adb has quit [Read error: Operation timed out]
[15:05:25] <cblack001> If only I knew about EMC when I built the thing the first time, I could have saved a ton of time.
[15:05:37] <cradek> torque feedback - interesting
[15:06:23] <cradek> some are doing feedback from edm (gap voltage I think it is?) and so far their only option is the motenc, which has a bit of adc. but if we had a mesa alternative for adc that would be great.
[15:07:26] <cblack001> It was mostly used so you could tell if a bit was wearing, and also retract out of the hole in a over torque situation.
[15:11:11] <cradek> lots of CNCs have plain old analog meters for stuff like spindle load - they avoid the adc problem/expense that way
[15:12:04] <cradek> it would be nice if we had better - hal handles analog signals fine - we are just at the mercy of the hardware.
[15:16:30] -!-
adb [
[email protected]] has joined #emc-devel
[15:16:39] -!-
pingufan has quit [Remote host closed the connection]
[15:20:13] -!-
x_ [
[email protected]] has joined #emc-devel
[15:20:37] x_ is now known as
micges_laser
[15:22:53] <cblack001> I love mesa products , so I would prefer to stick with them and help out if I can. Motlite looks like a fairly well priced solution as well. I did think about just putting an analog meter but you can't detect over torque and stop... The gantry lifted off the rails when a bit broke without it :)
[15:23:35] <cradek> well, you can if it causes following errors
[15:23:37] -!-
Dallur has quit [Quit: Leaving.]
[15:24:07] <cradek> for that, you could just do analog comparison in hardware and convert to one bit of digital?
[15:24:58] <cradek> but I agree - mesa stuff is good, and the way they work with the free software community is a model to everyone else.
[15:26:51] <cblack001> Yeah, I'm sure I can go another way, with more hardware or just program it into the amp but it would be nice to have..
[15:29:16] -!-
pcw_home has quit [Remote host closed the connection]
[15:38:43] -!-
kb8wmc has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.14/20110221055140]]
[15:42:10] -!-
psha [
[email protected]] has joined #emc-devel
[16:15:42] -!-
nullie has quit [Quit: Ex-Chat]
[16:17:57] -!-
tom3p [
[email protected]] has joined #emc-devel
[16:25:21] -!-
micges_cnc has quit [Quit: Leaving]
[16:39:40] -!-
skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.68.115.41.210] has joined #emc-devel
[16:51:08] -!-
seb_kuzminsky [
[email protected]] has joined #emc-devel
[17:12:15] <CIA-5> EMC: 03seb 07master * r6f3211f6795c 10/VERSION: master is now the 2.6 branch
[17:13:14] <cradek> so many things to remember...
[17:13:39] <seb_kuzminsky> master builds and passes the tests, but the docs dont build in pbuilder (when building debs)
[17:13:57] <seb_kuzminsky> http://buildbot.linuxcnc.org/buildbot/builders/deb-hardy-sim-binary-i386/builds/556/steps/shell_4/logs/stdio
[17:14:58] <cradek> what does that mean? a missing dependency?
[17:15:13] <seb_kuzminsky> cradek: i sort of wish VERSION was autogenerated from git-describe, like the debian/changelog version number
[17:15:30] <seb_kuzminsky> if you dont mind i might do that
[17:16:00] <cradek> you're going to add git as a build dependency then? I'm not sure that's justified
[17:16:12] <cradek> someone can build a tarball without any git programs installed
[17:16:32] <cradek> (although, I sympathize with not wanting to manually update it)
[17:16:47] <seb_kuzminsky> but where did the tarball come from? someone made it from the git repo, so they could make the VERSION file then
[17:17:18] <seb_kuzminsky> but yeah, more build deps = yuck
[17:17:31] <cradek> often people get a tgz from
http://git.linuxcnc.org/gitweb?p=emc2.git;a=snapshot;h=HEAD;sf=tgz
[17:17:46] <cradek> they're not usually made by humans
[17:17:50] <seb_kuzminsky> ah
[17:17:52] <seb_kuzminsky> ew
[17:18:03] <cradek> well, not always -- not usually is a stronger statement than my evidence
[17:18:23] <seb_kuzminsky> hm
[17:20:17] <cradek> I guess you could just fail gracefully
[17:20:43] <cradek> http://git.linuxcnc.org/gitweb?p=emc2.git;a=commitdiff;h=cf3b5270c824c372d8b370fe84391f6b241a00a7
[17:20:50] <cradek> 5 years ago I knew enough about configure to do this
[17:20:57] <cradek> looks like it would fail gracefully if VERSION isn't there
[17:21:38] <cradek> (jeez I've been working on emc for a long time)
[17:30:23] -!-
tom3p [
[email protected]] has parted #emc-devel
[17:31:27] <seb_kuzminsky> bbl
[17:36:48] -!-
mrsun has quit [Remote host closed the connection]
[17:44:08] -!-
JT-Shop [
[email protected]] has joined #emc-devel
[17:45:31] -!-
ve7it [
[email protected]] has joined #emc-devel
[18:02:05] -!-
mrsun has quit [Quit: Leaving]
[18:03:56] -!-
SWPadnos_ [SWPadnos_!~chatzilla@emc/developer/SWPadnos] has joined #emc-devel
[18:04:28] -!-
Dallur has quit [Read error: Connection reset by peer]
[18:13:43] -!-
x_ [
[email protected]] has joined #emc-devel
[18:16:06] -!-
micges_laser has quit [Ping timeout: 240 seconds]
[18:23:11] <psha> seb_kuzminsky: i guess it's due to old asciidoc configs
[18:23:29] <psha> i think new ones have to be placed near docs
[18:23:45] <psha> so they will be not version dependent
[18:24:01] <seb_kuzminsky> right, it fails on hardy and works on lucid
[18:25:40] <psha> there is documented way to create standalone configs, i'll look at it
[18:26:47] <seb_kuzminsky> thanks psha
[18:30:07] <psha> ah, yea, that's "new style tables"
[18:41:12] -!-
JT-Shop has quit [Ping timeout: 250 seconds]
[18:41:17] -!-
cv has quit []
[18:43:10] cv is now known as
snoop
[18:43:44] -!-
jthornton has quit [Ping timeout: 276 seconds]
[18:44:15] -!-
JT-Shop [
[email protected]] has joined #emc-devel
[18:45:04] -!-
jthornton [
[email protected]] has joined #emc-devel
[18:56:38] -!-
andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc-devel
[19:13:00] <andypugh> seb_kuzminsky: I had an email today asking about a SPI Hostmot2 driver. Do you still want to do it, or shall I do it, following the spec on that Wiki page?
[19:14:19] <seb_kuzminsky> i'm realistically not going to have time to do it for a while - too much work of the paying kind, and no personal need atm
[19:14:51] <seb_kuzminsky> i'm not convinced the hm2 spi spec on the wiki is the way to go
[19:14:53] <andypugh> My need is an excuse to be working on "the project" without being in the actual cold garage :-)
[19:14:58] <seb_kuzminsky> heh
[19:15:01] <seb_kuzminsky> i can relate
[19:15:02] <skunkworks> wth - no personal need? bite your tonge!
[19:15:20] <seb_kuzminsky> heh
[19:15:37] <skunkworks> tongue even
[19:15:53] <seb_kuzminsky> i bit my tongs and cracked a tooth, now what?
[19:16:16] <andypugh> I think however you approach SPI, you are going to need task-specific comp modules.
[19:17:01] <seb_kuzminsky> i agree
[19:17:12] <psha> andypugh: re: vcp bits display: you want some automation for vbox-with-many-leds?
[19:17:25] <seb_kuzminsky> the trick is making the interface between the hm2 spi driver and the spi-device drivers clean
[19:17:51] <seb_kuzminsky> iirc, jeff and i talked about this on irc a while back
[19:17:53] <cradek> was your email from cblack001? he's here.
[19:18:26] <andypugh> The guy who emailed me was asking about the 7i65. I can almost see an argument for that being a totally different way of doing it.
[19:18:49] <seb_kuzminsky> we talked about creating a new stream-like interface, possibly in hal or possibly in the kernel, sort of how serial ports and the such-like are handled in linux now
[19:18:54] <andypugh> Dunno, I am hopeless with names ;-)
[19:21:16] <andypugh> cblack001 does sound entirely plausible to be the guy who emailed. Perhaps he wold like to make himself known?
[19:23:01] <andypugh> The 7i65 is always going to use the SPI interface in a very specific way, that rather suggests a specific driver which creates a hm2_5i23.0.7i65.0.current pin, the question is how EMC2 can know that the firmware is 7i65 and not a generic SPI.
[19:29:10] x_ is now known as
micges_cnc
[19:32:17] <seb_kuzminsky> andypugh: i think in general there's no way to know that
[19:32:35] -!-
adb has quit [Read error: Operation timed out]
[19:32:44] <seb_kuzminsky> in general, spi devices dont announce themselves nicely (as opposed to, say the smart-serial devices, which identify themselves clearly)
[19:33:01] <seb_kuzminsky> so that id information needs to come from the user, in some awkward modparam monstrosity
[19:33:25] <andypugh> Unless Pete were to define a new tag for SPI modules that are part of a 7i65 firmware
[19:33:36] <seb_kuzminsky> and since the pins are made by the 7i65 driver, not the 5i23 driver, they should probably be in the 7i65 driver's namespace
[19:34:22] <seb_kuzminsky> in my wildest fever dreams i imagine a new data type for hal pins, representing a stream endpoint
[19:34:51] <seb_kuzminsky> the hm2_5i23.spi.0.endpoint could be netted to the 7i65.0.spi-endpoint, and they could communicate over that
[19:35:04] <seb_kuzminsky> then the 7i65.0 would have all the pins & stuff for the 7i65 board
[19:35:12] <seb_kuzminsky> dont know if that'd actually work
[19:36:29] <seb_kuzminsky> as i understand it, spi modules are too small to have tags
[19:36:37] <seb_kuzminsky> they're itty bitty shift registers
[19:37:31] -!-
motioncontrol has quit [Ping timeout: 248 seconds]
[19:37:32] -!-
adb [
[email protected]] has joined #emc-devel
[19:38:09] <andypugh> I mean the way that they are declared in the firmware. The GTAG
[19:38:42] <andypugh> And yes, there is no reason that there couldn't be a HAL pin that is a pointer to a struct.
[19:39:15] <andypugh> In fact, my assumption as to how the shared memory works, is that "net" just assigns all the pins to the same memory address?
[19:40:24] <andypugh> Not that I have ever bothered checking. It hasn't mattered.
[19:40:42] <seb_kuzminsky> andypugh: i think that's right
[19:40:48] <seb_kuzminsky> about the nets, that is
[19:40:54] <seb_kuzminsky> the gtag won't be able to work that way
[19:40:56] -!-
motioncontrol1 has quit [Ping timeout: 260 seconds]
[19:41:14] <seb_kuzminsky> the gtag identifies the functionality on the fpga, which in this case means "i'm an spi tranciever"
[19:41:29] <seb_kuzminsky> the gtag can't know anything about what device(s) are on the spi cable
[19:42:25] <andypugh> Yes, indeed. However there is no reason that Pete couldn't define two Gtags "I am an SPI transciever" and "I am an SPI transciever in a 7i65-specific-firmware"
[19:42:59] <andypugh> Well, there might actually be reasons. I am making assumptions again.
[19:43:20] <seb_kuzminsky> sure, he could, but i dont think that helps really
[19:43:46] <seb_kuzminsky> the user would have to specify the magic firmware that thinks it has a 7i65 on spi port 0
[19:45:00] <andypugh> Something has occurred to me. We could probably create a hal U32 pin, but malloc it as a huge struct, and write a struct to that address. (There is no type checking about what you are actually putting at a pointer). As long as the other module that uses the same pointer accesses it as the correct struct, it will all work.
[19:46:07] <cradek> that's clever, but only if pointers and ints have the same size
[19:46:27] <andypugh> I don't think that the 7i65 will work with anything other than a firmware written for the job, it needs the muxed encoders and SPI on the right pins, so it pretty specific.
[19:46:44] <cradek> oh wait, I was thinking about it wrong, I think
[19:47:30] <seb_kuzminsky> andypugh: that's all handled by the config modparam, i think it'd be wrong to have a custom firmware and special-case code in the driver for it
[19:49:15] <andypugh> Sorry, when I talk of the other module using the same pointer, I mean using the same pin, really. It is just that AFAIK the pins are just pointers which happen to point to the same memory adresses.
[19:49:46] <cradek> I understand now
[19:50:39] <seb_kuzminsky> i imagine hal_pin_pointer_new(), and everybody better agree what the pointer points to
[19:50:47] <seb_kuzminsky> just like we always do
[19:51:15] <andypugh> seb_kuzminsky: Aye, otherwise it all goes wrong. But we could lie.
[19:51:20] -!-
motioncontrol has quit [Ping timeout: 276 seconds]
[19:51:32] <andypugh> (I think, I have not actually tried this)
[19:52:04] <cradek> maybe something like a counted string: first word is the length
[19:53:07] <andypugh> The problem is preventing those pesky users from netting an SPI stream to ModBus stream.
[19:54:12] <cradek> how do you lock it? you need atomic reads and writes.
[19:58:12] <andypugh> I don't even know what an atomic read is, I am afraid
[19:58:55] <cradek> I write to the buffer one byte at a time - you read one byte at a time
[19:59:15] <cradek> nothing guarantees you get a consistent picture of one of the buffers I intended to write - you might get a combination of two or more.
[19:59:47] <cradek> bbl
[19:59:59] -!-
micges_cnc has quit [Quit: Leaving]
[20:02:26] <CIA-5> EMC: 03seb 07v2.5_branch * re7822915ff7b 10/docs/man/man9/hostmot2.9: hostmot2 manpage: fix an indentation error
[20:08:33] <CIA-5> EMC: 03seb 07v2.4_branch * r9d38e9bac3b0 10/docs/man/man9/hostmot2.9: hostmot2 manpage: add missing pin name info for raw mode & watchdog
[20:12:17] <CIA-5> EMC: 03seb 07v2.5_branch * r9d38e9bac3b0 10/docs/man/man9/hostmot2.9: hostmot2 manpage: add missing pin name info for raw mode & watchdog
[20:12:18] <CIA-5> EMC: 03seb 07v2.5_branch * r7fd555421faa 10/docs/man/man9/hostmot2.9: Merge branch 'v2.4_branch' into v2.5_branch
[20:13:32] <CIA-5> EMC: 03seb 07master * r147e7428fa68 10/docs/html/gcode.html: fix links in the gcode quickref
[20:13:33] <CIA-5> EMC: 03seb 07master * rb91da881830f 10/docs/html/gcode_fr.html: fix links for asciidoc
[20:13:33] <CIA-5> EMC: 03seb 07master * rf9a6a3cf1808 10/docs/ (html/gcode.html src/gcode/main.txt): clean up and update docs for tool touch off g10 l10/l11
[20:13:34] <CIA-5> EMC: 03seb 07master * re7822915ff7b 10/docs/man/man9/hostmot2.9: hostmot2 manpage: fix an indentation error
[20:13:45] <CIA-5> EMC: 03seb 07master * r7fd555421faa 10/docs/man/man9/hostmot2.9: Merge branch 'v2.4_branch' into v2.5_branch
[20:13:45] <CIA-5> EMC: 03seb 07master * r9d38e9bac3b0 10/docs/man/man9/hostmot2.9: hostmot2 manpage: add missing pin name info for raw mode & watchdog
[20:13:46] * seb_kuzminsky gives git a big hug & kiss
[20:13:48] <CIA-5> EMC: 03seb 07master * r54591448bed6 10/docs/ (4 files in 3 dirs): Merge branch 'v2.5_branch'
[20:13:55] <seb_kuzminsky> mmmmwwa!
[20:18:51] -!-
seb_kuzminsky has quit [Quit: Client exiting]
[20:34:39] <cradek> ain't it the truth
[20:45:08] -!-
adb has quit [Ping timeout: 250 seconds]
[20:52:02] -!-
motioncontrol1 has quit [Read error: Operation timed out]
[21:04:04] -!-
micges [
[email protected]] has joined #emc-devel
[21:04:46] -!-
micges has quit [Client Quit]
[21:14:01] -!-
seb_kuzminsky [
[email protected]] has joined #emc-devel
[21:21:36] -!-
x_ [
[email protected]] has joined #emc-devel
[21:25:27] -!-
motioncontrol has quit [Quit: Sto andando via]
[21:25:52] -!-
psha has quit [Quit: leaving]
[21:50:13] -!-
kb8wmc has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101206122310]]
[22:21:40] -!-
Tom_L has quit []
[22:27:50] -!-
adb [
[email protected]] has joined #emc-devel
[22:43:57] -!-
acemi has quit [Quit: WeeChat 0.3.2]
[22:59:46] -!-
Tom_L has quit []
[23:01:47] -!-
uwe_ has quit [Ping timeout: 276 seconds]
[23:03:33] -!-
SWPadnos_ has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.15/20110303024726]]
[23:04:45] -!-
skunkworks has quit [Ping timeout: 245 seconds]
[23:06:56] -!-
tlab has quit [Ping timeout: 255 seconds]
[23:09:47] -!-
micges__ [
[email protected]] has joined #emc-devel
[23:12:42] -!-
micges__ has quit [Client Quit]
[23:13:51] -!-
x_ has quit [Ping timeout: 260 seconds]
[23:17:20] -!-
seb_kuzminsky has quit [Ping timeout: 252 seconds]
[23:21:26] -!-
OoBIGeye has quit [Ping timeout: 260 seconds]
[23:21:46] -!-
robh__ has quit [Ping timeout: 240 seconds]
[23:22:04] -!-
tlab has quit [Max SendQ exceeded]
[23:30:11] -!-
OoBIGeye has quit [Ping timeout: 260 seconds]
[23:40:06] -!-
OoBIGeye has quit [Ping timeout: 260 seconds]
[23:43:42] -!-
tlab has quit [Max SendQ exceeded]
[23:44:57] -!-
robh__ [
[email protected]] has joined #emc-devel
[23:49:56] -!-
adb has quit [Ping timeout: 246 seconds]
[23:50:36] -!-
adb [
[email protected]] has joined #emc-devel
[23:54:41] -!-
OoBIGeye has quit [Ping timeout: 260 seconds]