#linuxcnc-devel | Logs for 2012-01-22

[00:02:33] -!- A2Sheds has quit [Ping timeout: 252 seconds]
[00:03:35] -!- theorbtwo has quit [Ping timeout: 255 seconds]
[00:03:49] theorb is now known as theorbtwo
[00:04:41] <seb_kuzminsky> thanks john
[00:04:42] <seb_kuzminsky> bbl
[00:06:38] <jepler> was there anybody arguing against naming the main script just 'cnc'? It's not brand-y enough, maybe?
[00:06:53] * seb_kuzminsky likes brandy
[00:07:15] <jepler> sorry, I don't actually have any brandy
[00:07:47] <jepler> I'm presently debating between a homebrew blonde ale or a g&t, though, so if you can help me decide about that I'd be in your debt.
[00:07:51] <andypugh> Ah, jepler, do you have a moment to talk about something else? EmcPT on the other forum has noticed that Axis seems to pay more attention to T than H
[00:08:06] <JT-Shop> I like cnc, nice and short
[00:08:06] <andypugh> jepler: Which gin?
[00:08:40] <seb_kuzminsky> g&t definately
[00:09:07] <JT-Shop> no, Remy Martin... nice and smooth
[00:09:09] <jepler> andypugh: the bottle that was 2 brands away from being the cheapest the day I bought it
[00:09:23] <andypugh> Blonde homebrew then.
[00:09:32] <JT-Shop> that is for cleaning toilets I think
[00:09:49] <andypugh> Have you tried Hendrik's?
[00:09:58] <JT-Shop> for sure homebrew over that
[00:10:02] <JT-Shop> me? no
[00:11:15] <jepler> turns out it's tanqueray. no, I don't think I've had Hendrick's.
[00:11:22] <JT-Shop> I don't drink gin only a glass of congnac or really good whiskey from time to time on a cold night just before bed
[00:12:08] <andypugh> Anyway, As far as I can see, the Pocket number in the tooltable is not used at all unless you have RANDOM_TOLCHANGER. So, typically the T word sets the carousel slot, and the H word sets the offset. So, the offsets in the bottom status bar, and the tool shape preview should come from the tool table, based on H. T doesn't have any link at all to the tool table.
[00:12:25] <JT-Shop> wow Remy Martin Loius XIII is $1700
[00:13:25] <jepler> andypugh: not sure what pocket numbers have to do with anything. both G43 H and M6 T are ways to refer to toollll numbers.
[00:13:28] <JT-Shop> I've seen people drink tanqueray and not die...
[00:13:49] <jepler> toolllll numbers?
[00:13:59] <andypugh> EmcPT sells a machine for making seals and O-rings to order, but has some 4-tip tools. So, he will use 1 T word with 4 different H words, but the bottom status bar always shows the data from the T-line of the stautus bar.
[00:14:03] <jepler> ok, here's the deal: before I buy another bottle of gin, I'll consult both with you guys and with my accountant
[00:15:06] <jepler> (oh, I see the $1700 bottle was not gin .. phew)
[00:15:53] <JT-Shop> not that I've ever seen much less touched a bottle of Louis XIII
[00:16:26] <JT-Shop> andypugh: is he saying that because the Axis display is not updating or did he check the parameter to see that they didn't change?
[00:16:51] <andypugh> jepler: I am not sure that, technically, T is a tool number. There is a shortcut that an unqualified G43 will choose an H to match the current T, but I don't know if there is any time that a bare "T" will cause any access to the tooltable (Except in the case of a random toolchanger, where it is used to find where the tool has been put)
[00:16:59] <JT-Shop> I do like the Remy Martin 1738 for $40 and a bottle of that lasts about 6 months
[00:17:47] <andypugh> The Axis preview updates the position if you G43 Hx, but not the tool shape picture (on a lathe). I am not sure if it changes the tool diameter in Mill.
[00:17:58] <jepler> so .. the default tool table with axis has tool 1 as .511 long and tool 2 as .1 long
[00:18:26] <jepler> m6 t1 g43 h2 says to load tool 1 but use the length of tool 2 for TLO
[00:18:37] <jepler> right?
[00:19:04] <andypugh> Yes, and the status bar at the bottom will show Tool1 Zo .511
[00:19:17] <jepler> so the request is that it should show "tool 2" now
[00:19:24] <andypugh> Yes.
[00:19:28] <andypugh> No
[00:19:49] <jepler> now program m6 t1 g43 h2 g41 d3
[00:20:05] <andypugh> It should show "Tool 2" but the offset should now be shown as ".1"
[00:20:11] <jepler> so that cutter comp is using the .201 diameter from tool 3
[00:20:29] <jepler> should it show "tool 1, offset 0.1, diameter .201"?
[00:20:35] <andypugh> I think so.
[00:20:49] <jepler> where none of the 3 come from the same row in the tool table.
[00:21:02] <jepler> I'm going to need a more expensive gin.
[00:21:07] <andypugh> It should be showing the current tool holder, and what offsets are being used.
[00:21:15] -!- jts has quit [Remote host closed the connection]
[00:21:38] <andypugh> jepler: That was true even before I brought this up.
[00:21:50] <jepler> oh snap
[00:22:22] <andypugh> Good gin is very good. Bad gin is an industrial solvent.
[00:23:13] <andypugh> I am trying to parse that "oh snap" :-)
[00:23:24] <jepler> it's american slang. it's what you say when someone (me) gets told
[00:23:31] <andypugh> Gin-related, disagreement or agreement?
[00:23:48] <JT-Shop> oh snap has me wondering too what that is
[00:24:06] <jepler> I think it's mostly used by people who are a good 20 years younger than me
[00:24:40] <jepler> basically, I acknowledge your excellent put-down
[00:24:46] <JT-Shop> ah, ok that explains it then
[00:25:01] <jepler> http://www.flickr.com/photos/vjornaxx/294168009/
[00:25:06] <JT-Shop> my friends are usually my age or older if they are still alive
[00:26:19] <JT-Shop> kinda like when Alton Brown says Oh Bother?
[00:26:46] <jepler> dunno. is that something I'd need a TV to know about?
[00:27:03] -!- clytle374 has quit [Read error: Connection reset by peer]
[00:27:09] <JT-Shop> that or internet, but a tv would be faster
[00:27:23] <jepler> andypugh: it looks like the H and D numbers are not in the stat buffer
[00:27:32] <jepler> the Z offset 0.511 vs 0.1 is
[00:27:34] <andypugh> I am guessing that you have conceded on the Gin, not on what Axis should display?
[00:27:49] -!- jts [jts!~jts@cpe-67-49-64-146.socal.res.rr.com] has joined #linuxcnc-devel
[00:27:59] <JT-Shop> http://altonbrown.com/
[00:28:09] <jepler> I wish cradek would chime in. he has done 2 tools on one post in his lathe with TC
[00:28:56] <jepler> there are 8 positions for the turret. I think he has it set up so that T1 and T9 both move the turret to the same position, but have different characteristics
[00:29:30] <jepler> in ladder this can be as simple as ignoring some high bits of the tool number going into the toolchange logic
[00:29:42] <andypugh> It's easy enough to do in practice. You accept that the toolchanger only ever sees the T word and then apply the offsets with H. It does all work fine, but the status bar not updating and the tool shape preview staying with the T makes it unclear that it works.
[00:29:43] <jepler> is "x & 7" one of the things you can write in the expression language?
[00:29:48] -!- clytle374 [clytle374!~clytle374@] has joined #linuxcnc-devel
[00:30:16] <jepler> it looks like a whole path interp -> task -> [stat buffer] -> axis would need to change to really show what is desired
[00:30:27] <JT-Shop> andypugh: are the parameters changing?
[00:31:04] <jepler> particularly if you want tool touch off to be able to use the G43 D number
[00:31:13] <andypugh> Yes, the tool offsets all change correctly. I am pretty sure that it works, it is just that Axis doesn't show it working.
[00:32:48] <jepler> (if axis *does* use the d number then I should go study that code and figure out what it does)
[00:35:04] <andypugh> This is the (Rather cool) machine that brought this up, machining O-rings and seals on-demand. 12 position toolpost but many more tool tips. http://www.youtube.com/watch?v=ddno7YzyS6o
[00:36:29] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust492.basl.cable.virginmedia.com] has parted #linuxcnc-devel
[00:36:59] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust492.basl.cable.virginmedia.com] has joined #linuxcnc-devel
[00:38:12] -!- Gast109 has quit [Quit: Bye Bye]
[00:39:02] <jepler> andypugh: did you understand what I was saying about using distinct T numbers for multiple tools at one position? I agree it's not a perfect solution
[00:39:13] <jepler> right now I don't think that just by changing axis it's possible to implement the desired behavior
[00:39:25] -!- clytle374 has quit [Quit: Leaving]
[00:40:54] -!- testtest [testtest!~jepler@emc/developer/pdpc.professional.jepler] has joined #linuxcnc-devel
[00:41:08] -!- testtest has quit [Client Quit]
[00:42:18] <andypugh> It seems that EMC2 works perfectly well now, as long as you accept that the T-word is a carousel position, and the H-word in G43 is the tool number from the tool table. There is a default that the H word is assumed to be the same as the T, but that can be seen as a convenience. Unless I have the wrong end of the stick, the T-word is only ever seen by the toolchanger code, and it is the H word that references tool n
[00:42:18] <andypugh> in the tool table.
[00:42:54] <jepler> I'm pretty sure axis touch off is going to use the T number too
[00:43:03] <jepler> that could lead to some unfortunate consequences
[00:43:12] <andypugh> Ah.
[00:43:19] <jepler> tool touch off, that is
[00:43:42] <jepler> workpiece touch off should be using the active offsets
[00:44:13] <andypugh> Is there a G-code route to do whatever it is that Axis tool touch-off does? And would that use T or H?
[00:45:03] <jepler> axis touch off always works by issuing gcode in mdi
[00:45:19] <andypugh> (I think G10 L20 uses the current offsets, so, again, that would be H not T)
[00:46:16] <jepler> G10 L10 / G10 L11 are using s.tool_in_spindle for P
[00:46:26] <jepler> so that'll be the T number
[00:46:36] -!- tlab has quit [Client Quit]
[00:47:37] <andypugh> That seems wrong, as that isn't the offset that is currently in use..
[00:47:46] -!- Gast109 has quit [Quit: Bye Bye]
[00:48:07] <andypugh> No, sorry, different codes
[00:49:00] <jepler> if you issue M6 T1 G43 H2, then tool touch off, axis will issue G10 L1x P1 ...
[00:49:09] <jepler> and modify the tool table for tool 1
[00:49:16] <mhaberler> there's ongoing confusion of tool vs pocket throughout the code down to nml fields - IMO the 'pocket' concept has no place in the code since this is a changer implementation detail; I plan to clean this up eventually
[00:50:00] <andypugh> Aye, "Pocket" in the toolchanger has no meaning, and is not used, in all but a very small minority of machines.
[00:50:23] <andypugh> Sorry, I mean "pocket" in the tool_table_
[00:50:31] <jepler> so I keep coming back to: this isn't a 30-second "whoops, one simple thing is overlooked" thing, and re-iterating the "use distinct tool numbers and changer logic that maps them onto the tool slots"
[00:50:59] <cradek> andypugh: if you'd do M6 T1, G43.1 Z1 X2 Y3, which tool table entry would you have tool touch off change?
[00:51:01] <jepler> I would be pleased to see the additional information available to axis so that it can do something smarter
[00:51:47] <jepler> cradek: hi. did I get the explanation of how you do multiple tools on one turret position approximately right?
[00:53:15] <cradek> I haven't spotted your explanation - which channel?
[00:53:54] <andypugh> cradek: Does G43.1 access the tooltable at all?
[00:54:06] <jepler> cradek: here
[00:54:07] <cradek> this comes down to: the stat buffer has the loaded tool number and the numeric/axis values of the tool offset, and nothing else
[00:54:11] <cradek> andypugh: no
[00:54:24] <andypugh> Ah, you mean G43.1, then a Touch-off?
[00:54:30] <jepler> <jepler> there are 8 positions for the turret. I think he has it set up so that T1 and T9 both move the turret to the same position, but have different characteristics
[00:54:33] <jepler> <jepler> in ladder this can be as simple as ignoring some high bits of the tool number going into the toolchange logic
[00:55:13] <cradek> yes I use modulo tool number
[00:55:25] -!- Nick001-Shop has quit [Remote host closed the connection]
[00:55:30] <cradek> it's pretty easy with a turret
[00:56:05] <cradek> andypugh: yes that's what I mean. the tool offsets aren't necessarily from the tool table at all.
[00:56:51] <andypugh> There are a number of ways to do this. One way (but it is a big change) would be to send the pocket number to the tool-request pins, and allow multiple tools in a pocket.
[00:58:06] <cradek> is his complaint that orientation doesn't get changed by G43H2 where 2 != loaded T number?
[00:58:06] <andypugh> The current situation is that you can use T to send a pocket number to the changer, and H to load offsets from the tool table, but the touch-off behaviour then becomes ill-defined.
[00:58:25] <cradek> or is this a complaint about touch off
[00:58:31] <cradek> or is it two problems
[00:58:38] <andypugh> Yes, because he depends on tool radius compensation.
[00:58:41] <cradek> I thought it was the former
[00:59:04] <cradek> G43H has nothing to do with radius compensation
[00:59:17] <andypugh> Eh?
[00:59:44] <cradek> G43 loads axis offsets
[00:59:52] <andypugh> But not tool radius?
[00:59:54] <cradek> if you want radius compensation != tool number, you use D
[01:00:05] <cradek> no it's totally separate
[01:00:05] <andypugh> Ah, yes, OK
[01:00:27] <cradek> so if his complaint is about orientation or radius comp, he's just mistaken
[01:01:24] <andypugh> Does the D-word load the orientation from the tool table?
[01:01:30] <cradek> if you want tool touch off to follow something other than tool number, you need to define what to do when there's no offset loaded, or when the offset didn't come from the tool table at all, and if you do that I think you've made it muddier, not clearer
[01:01:35] <andypugh> Or just the radius?
[01:01:47] <cradek> it should certainly do orientation and radius
[01:02:23] <cradek> (did you see me hedge there?)
[01:03:58] -!- rob_h has quit [Ping timeout: 260 seconds]
[01:04:22] <cradek> yes the code says it does do it right
[01:04:34] <cradek> CHP((find_tool_pocket(settings, tool, &index)));
[01:04:42] <cradek> ...
[01:04:43] <cradek> radius = USER_TO_PROGRAM_LEN(settings->tool_table[index].diameter) / 2.0;
[01:04:46] <cradek> orientation = settings->tool_table[index].orientation;
[01:06:03] <andypugh> i am starting to think that sending the Pocket number to the toolchanger pins is perhaps better.
[01:06:25] <cradek> what does that have to do with this?
[01:07:46] <cradek> no matter how the tool changer works, you've got the same problems
[01:08:49] <cradek> you've got H and D, and if you don't specify them they default to T's value
[01:08:55] <andypugh> Well, if the T number is what determines where the offsets get stored when you touch-off, then that seems like a cleaner starting point.
[01:08:57] <cradek> that seems sane to me
[01:09:08] <cradek> ok you're talking about touch off now
[01:09:15] <cradek> I agree it'd be nice if that could be smarter ... somehow
[01:09:46] <andypugh> I am talking about the whole relationship between toolchanger position and offsets.
[01:09:48] <cradek> but you can have T, H, and D in effect from three different tool table entries
[01:10:45] <andypugh> There is an assumption that any one toolchanger position can only have one tool controlled point. It's not a strong assumption, but it is there.
[01:11:11] <cradek> I disagree. the whole reason why you can specify H,D != T is because that is not true.
[01:11:34] <andypugh> No, that fact is what makes it a weak assumption ;-)
[01:11:56] -!- clytle374 [clytle374!~clytle374@75-105-212-57.cust.wildblue.net] has joined #linuxcnc-devel
[01:13:30] <andypugh> I am mainly thinking about multiple tool tips in a single carousel position here. (and that is largely a lathe issue, but could apply to generic robots with interchangeable effectors too).
[01:14:05] <cradek> I feel like modulo T number (and using unspecified H,D) is a great way to handle that
[01:14:37] <cradek> it feels a tad hackish, but for a fixed size carousel it's pretty foolproof
[01:14:47] -!- tlab has quit [Quit: Leaving]
[01:14:52] <jepler> without arguing that linuxcnc is perfect as-is (you know how I love to do that), I also think that it's worth exploring "modulo T" as a way to solve this user's difficulty without changing code..
[01:15:14] <cradek> yeah I'd never argue that either, but it's a fine solution
[01:15:54] <andypugh> My assumption until last night (and EmcPT's too) was that the toolchanger was sent the pocket number, and you could put multiple tools in one pocket. Unfortunately the tool table insists that pocket numbers can't be repeated. Also, the toolchanger pin sends tool, not pocket.
[01:16:19] <cradek> to be precise, it sends both
[01:16:29] <cradek> you can use whichever you want in hal
[01:16:45] <andypugh> As the mapping (with RANDOM_TC) is always tool to pocket, I am not sure why it doesn't allow multiple tools to share a pocket
[01:16:46] <cradek> it is correct that you can't have more than one tool in a pocket.
[01:17:33] <andypugh> AFAIK from reading, the pocket-selected pin is only created with RANDOM_TOOLCHANGER?
[01:17:52] <cradek> oh is it? my mistake then.
[01:21:05] <andypugh> Is there any reason not to always make the pocket pin available? Also, why disallow multiple tools in one pocket?
[01:22:36] <cradek> unfortunately to both of those my answer is I don't know
[01:25:32] <andypugh> We could make the changes, and see what breaks. As far as I can see it involves removing two checks.
[01:25:43] <andypugh> But no new code.
[01:26:15] <cradek> I support that strategy!
[01:26:32] <cradek> but only if by we, you mean you :-)
[01:28:13] <andypugh> I am game for a laugh, and the guy is already using 2.6 anyway. I guess we can always revert the patch.
[01:28:50] <cradek> whoah, test before you decide whether to push
[01:28:50] <cradek> there are sims for both random and nonrandom
[01:29:05] <cradek> I think modulo T is his cleanest answer today
[01:29:28] <andypugh> Well, as the guy is running 2.6 anyway, and I think is building from Git, then he ought to be able to apply a patch
[01:30:00] <cradek> maybe we're better to hold out for mhaberler to redo the whole mess sanely
[01:30:08] <cradek> hope springs eternal
[01:30:26] <mhaberler> I am actually working on it
[01:30:30] <cradek> :-)
[01:32:01] <mhaberler> I reordered tasks as follows: 1. make interp and task instantiable and spin out state into POD classes (world model, interp state, config, modal state)
[01:33:27] <mhaberler> 2. add in redis (done at call level) to provide distributed late binding of variables (to break the emcstatus limitations); that worked in the old code but makes no sense to do before 1) is complete
[01:33:41] <mhaberler> 3. try to reach the status ante again;)
[01:35:05] <mhaberler> 1) twinkers but needs serious testing in an automated fashion
[01:37:10] <mhaberler> unfortunately refactoring is terminally boring work
[01:38:22] <cradek> yeah I bet. yikes.
[01:38:53] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[01:41:41] -!- sumpfralle has quit [Ping timeout: 248 seconds]
[01:42:39] <mhaberler> the idea with task instances is to enable switching between different task/interp instance stacks ontop of motion (read as: MDI while paused - with limitations)
[01:44:01] <mhaberler> ah, and canon. that is 'classified' as well, partially.
[01:48:41] <mhaberler> btw: this means - no more static variables and free functions in the interpreter please
[01:49:13] <andypugh> I must be getting close to finishing a project, I am looking for new things to start. I seem to have a phobia about finishing things.
[01:50:09] <mhaberler> "there's a lot to do - let's call in sick!"
[01:53:00] <andypugh> it's more that I got my encoder boards today, so need to assemble the mill and the Sekrit Project, and see if it works, so suddenly looking at toolchangers for gang tooling seems important (I have no toolchanger, and no gang tooling)
[01:59:13] <jepler> wait a minute -- andypugh is going to have a mill?
[01:59:39] -!- EmcPT [EmcPT!~EmcPT@bl12-43-32.dsl.telepac.pt] has joined #linuxcnc-devel
[02:01:17] <andypugh> I have had a mill for ages.
[02:01:46] <jepler> you just don't use the mesa stuff on it?
[02:01:55] <andypugh> Yes, it has a 7i43
[02:01:55] <jepler> or for it
[02:02:01] <jepler> huh, shows what I know
[02:02:41] <andypugh> The new mill will have a 5i23, 7i64, 3x8i20 and a 7i39. I bought one of those :-)
[02:05:00] <andypugh> I find myself looking for projects for all these Mesa cards I seem to have acquired. I think I have 13 of them. Possibly more.
[02:05:53] <jepler> har
[02:07:11] -!- clytle374 has quit [Quit: Leaving]
[02:07:37] -!- EmcPT has quit [Remote host closed the connection]
[02:36:57] -!- andypugh has quit [Quit: andypugh]
[02:39:26] -!- clytle374 [clytle374!~clytle374@] has joined #linuxcnc-devel
[02:47:31] -!- cstop has quit [Quit: Leaving]
[03:15:14] <jepler> seb_kuzminsky: does buildbot use a signing key for its package repository?
[03:16:10] <jepler> looks like the answer is "no"..
[03:17:32] <jepler> I'm playing with a revamped installer shell script; it aims to cover all supported distributions with a single script automatically, and to change between buildbot vs released package and between sim and rt with trivial edits..
[03:17:36] <jepler> http://emergent.unpythonic.net/files/sandbox/linuxcnc-install.run
[03:17:56] <jepler> I'd love to make it accept an apt signing key for buildbot if there was one
[03:24:57] -!- GoSebGo has quit [Quit: Bye]
[03:25:13] -!- GoSebGo [GoSebGo!~Seb@] has joined #linuxcnc-devel
[03:27:57] -!- GoSebGo has quit [Client Quit]
[03:29:02] -!- tlab has quit [Quit: Leaving]
[03:30:00] <CIA-99> 03jepler 07v2.5_branch * r326234ba23fe 10/scripts/rtapi.conf.in: include requried kernel version in rtapi.conf
[03:37:55] -!- mhaberler has quit [Quit: mhaberler]
[03:43:39] <jepler> 'night all
[04:56:21] elmo401 is now known as elmo40
[05:00:30] -!- elmo40 has quit [Ping timeout: 248 seconds]
[05:42:05] -!- clytle374 has quit [Quit: Leaving]
[06:11:19] -!- n2diy has quit [Quit: Ex-Chat]
[06:15:03] -!- psha [psha!~psha@] has joined #linuxcnc-devel
[06:15:12] -!- garage_seb [garage_seb!~seb@174-16-121-79.hlrn.qwest.net] has joined #linuxcnc-devel
[06:30:52] -!- cmorley has quit [Ping timeout: 244 seconds]
[06:47:09] -!- cmorley [cmorley!~chris@S010688ae1d61f51a.no.shawcable.net] has joined #linuxcnc-devel
[06:49:58] -!- ve7it has quit [Remote host closed the connection]
[06:57:47] <garage_seb> making chips with linuxcnc 2.5.0~pre2-732 :-)
[07:09:06] -!- cmorley has quit [Ping timeout: 244 seconds]
[07:23:26] -!- cmorley [cmorley!~chris@] has joined #linuxcnc-devel
[07:35:46] -!- mhaberler [mhaberler!~mhaberler@] has joined #linuxcnc-devel
[07:41:00] -!- rob_h [rob_h!~rob_h@5ace704a.bb.sky.com] has joined #linuxcnc-devel
[08:19:03] -!- psha has quit [Quit: Lost terminal]
[08:19:52] -!- garage_seb has quit [Quit: Leaving]
[08:25:36] -!- cmorley has quit [Ping timeout: 240 seconds]
[08:30:16] -!- cmorley [cmorley!~chris@] has joined #linuxcnc-devel
[08:32:32] -!- LawrenceSeattle has quit [Ping timeout: 255 seconds]
[08:32:32] LawrenceSeattle_ is now known as LawrenceSeattle
[08:47:49] -!- vladimirek [vladimirek!~vladimire@] has joined #linuxcnc-devel
[09:30:15] -!- mhaberler_ [mhaberler_!~mhaberler@] has joined #linuxcnc-devel
[09:30:15] -!- mhaberler has quit [Read error: Connection reset by peer]
[09:30:16] mhaberler_ is now known as mhaberler
[09:40:32] -!- factor has quit [Read error: Connection reset by peer]
[10:07:59] -!- mhaberler has quit [Quit: mhaberler]
[10:09:53] -!- LawrenceSeattle has quit [Remote host closed the connection]
[10:12:32] -!- LawrenceSeattle has quit [Client Quit]
[10:22:47] -!- vladimirek has quit [Remote host closed the connection]
[10:22:50] -!- LawrenceSeattle has quit [Quit: LawrenceSeattle]
[10:40:05] -!- e-ndy [e-ndy!~jkastner@static-84-242-102-36.net.upcbroadband.cz] has joined #linuxcnc-devel
[10:41:41] -!- sumpfralle has quit [Ping timeout: 255 seconds]
[10:46:06] -!- skunkworks has quit [Ping timeout: 244 seconds]
[10:47:23] -!- skunkworks [skunkworks!~chatzilla@str-bb-cable-south2-static-6-425.dsl.airstreamcomm.net] has joined #linuxcnc-devel
[11:15:01] -!- EmcPT has quit [Ping timeout: 248 seconds]
[11:23:29] -!- EmcPT has quit [Quit: Konversation terminated!]
[11:24:30] -!- EmcPT has quit [Remote host closed the connection]
[11:25:53] -!- tissf [tissf!5a13dfd5@gateway/web/freenode/ip.] has joined #linuxcnc-devel
[11:26:19] -!- tissf [tissf!5a13dfd5@gateway/web/freenode/ip.] has parted #linuxcnc-devel
[11:26:48] -!- tissf_ [tissf_!5a13dfd5@gateway/web/freenode/ip.] has joined #linuxcnc-devel
[11:37:35] -!- bootneck1ad has quit [Ping timeout: 252 seconds]
[11:55:09] -!- Loetmichel_ has quit [Ping timeout: 245 seconds]
[12:26:20] -!- syyl has quit [Ping timeout: 244 seconds]
[13:38:05] -!- syyl_ has quit [Ping timeout: 255 seconds]
[13:42:17] -!- yoyo has quit [Ping timeout: 244 seconds]
[13:45:40] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust492.basl.cable.virginmedia.com] has joined #linuxcnc-devel
[14:00:22] -!- yoyo has quit [Client Quit]
[14:01:38] -!- awallin has quit [Read error: Operation timed out]
[14:02:36] -!- n2diy has quit [Quit: Ex-Chat]
[14:07:30] -!- n2diy has quit [Client Quit]
[14:13:18] <andypugh> cradek? I am not sure if you are there (surely it's early there?) but it seems we were both wrong about tool-prep-pocket (or both right). It does exist without RANDOM_TC but shows the tool number, not the pocket number from the tool table.
[14:13:31] <andypugh> You can hand-edit the tool table to put more than one tool in one pocket, but then RANDOM_TC can only find one of the tools (I guess is searches the pockets for tools, not tools for pockets)
[14:25:55] -!- awallin [awallin!awallin@lakka.kapsi.fi] has joined #linuxcnc-devel
[14:31:01] -!- isssy has quit [Quit: Bye Bye]
[14:51:42] -!- mhaberler [mhaberler!~mhaberler@] has joined #linuxcnc-devel
[14:53:34] -!- EmcPT has quit [Read error: Operation timed out]
[14:53:44] -!- the_wench has quit [Ping timeout: 240 seconds]
[14:54:34] <CIA-99> 03jthornton 07v2.5_branch * r7237a7d79130 10/docs/src/gcode/ (gcode.txt gcode_de.txt gcode_es.txt gcode_pl.txt): Docs: fix links
[14:55:07] -!- archivist has quit [Ping timeout: 240 seconds]
[14:59:24] -!- Mjolinor has quit [Ping timeout: 240 seconds]
[15:03:45] <andypugh> I have figured out how the tool table works. And it seems that I can't make the changes I wantd.
[15:05:46] <andypugh> Effectively EMC2 has a 56 element array of tool data which is its internal representation of the tool file. The file is parsed such that the "Pocket" number in the file determines which array element is loaded with the data. That means that offsets are associated with pockets, not with tools, inside the code. (This is not necessarily the case in the tool file)
[15:09:32] -!- the_wench [the_wench!~the_wench@host81-149-189-98.in-addr.btopenworld.com] has joined #linuxcnc-devel
[15:09:32] -!- archivist [archivist!~archivist@host81-149-189-98.in-addr.btopenworld.com] has joined #linuxcnc-devel
[15:13:58] <andypugh> I think that a static array of all tool data with array elements corresponding to both tool number and pocket number would make more sense. At the moment RANDOM_TC moves data around the array to keep track of tools during toolchanges. (Or so I think, I haven't found that code yet)
[15:15:28] -!- EmcPT has quit [Ping timeout: 276 seconds]
[15:17:44] -!- Mjolinor has quit [Ping timeout: 240 seconds]
[15:22:53] -!- psha [psha!~psha@] has joined #linuxcnc-devel
[15:46:27] <andypugh> cradek: I came back to your idea of modulo-toolnumbers and letting the toolchanger handle it. Anything else is a heap of work.
[15:51:53] -!- elmo40 has quit [Ping timeout: 252 seconds]
[15:56:17] <mhaberler> very neat widgets - unfortunately in Ada:-/: http://www.dmitry-kazakov.de/ada/aicwl.htm cradek might like the clock widgets
[16:06:54] <CIA-99> 03jthornton 07v2.5_branch * r18002a663c90 10/docs/src/gcode/ (8 files): Docs: fix links
[16:08:12] -!- EmcPT has quit [Remote host closed the connection]
[16:10:12] -!- mhaberler has quit [Quit: mhaberler]
[16:12:44] <CIA-99> 03jthornton 07v2.5_branch * rdaaae954de88 10/docs/src/gcode/ (8 files): Docs: fix links
[16:17:30] -!- micges [micges!~x@apn-77-115-110-99.dynamic.gprs.plus.pl] has joined #linuxcnc-devel
[16:23:45] -!- factor has quit [Ping timeout: 252 seconds]
[16:26:03] -!- GoSebGo [GoSebGo!~Seb@] has joined #linuxcnc-devel
[16:30:12] <jepler> there have to be nearly 0 people who think this is actually a good user interface element. http://www.dmitry-kazakov.de/ada/aicwl.htm#Gtk.Gauge.Rectangular_70s
[16:32:07] <andypugh> It's, err, stylish.
[16:32:37] <JT-Shop> I think I saw one like that on a Mach screen
[16:33:04] -!- mhaberler [mhaberler!~mhaberler@] has joined #linuxcnc-devel
[16:51:28] -!- syyl_ws has quit [Quit: Verlassend]
[16:57:19] -!- clytle374 [clytle374!~clytle374@75-105-212-57.cust.wildblue.net] has joined #linuxcnc-devel
[17:00:24] <CIA-99> 03jthornton 07v2.5_branch * rf14fa1448892 10/docs/src/ (16 files in 3 dirs): Docs: fix links
[17:01:32] <jthornton> I have 5 more emails from Kent to work through so if anyone wants to tackle the config readme files re-branding it would make my mouse finger happy
[17:01:46] -!- jts has quit [Remote host closed the connection]
[17:09:35] -!- sumpfralle has quit [Ping timeout: 255 seconds]
[17:12:50] -!- jprentice has quit [Quit: Ex-Chat]
[17:15:56] <tissf_> I'll do it John
[17:17:10] <JT-Shop> thanks Francis
[17:17:15] -!- jts [jts!~jts@cpe-67-49-64-146.socal.res.rr.com] has joined #linuxcnc-devel
[17:19:41] -!- elmo40 has quit [Ping timeout: 252 seconds]
[17:20:34] -!- jts has quit [Client Quit]
[17:27:22] -!- e-ndy has quit [Quit: Ex-Chat]
[17:32:21] <tissf_> What is the consensus for the new name of xemc?
[17:32:39] -!- ve7it [ve7it!~LawrenceG@S0106001c10b7770f.pk.shawcable.net] has joined #linuxcnc-devel
[17:33:50] <JT-Shop> wasn't there something about that on the dev mailing list?
[17:36:20] <GoSebGo> The current name is xlinuxcnc, there's been some talk about xcnc but no decision yet
[17:37:10] <psha> we really need 'emc' python module
[17:37:36] <psha> that prints warning about obsolete name and imports linuxcnc
[17:37:37] <psha> btw it was very dumb idea to rename _everything_...
[17:38:06] <tissf_> thanks GoSebGo
[17:42:42] <micges> psha: seems like agreement with EMC corporation included removing all instances of 'emc' from source code
[17:43:27] <psha> heh, announce on mailing list gives other impression
[17:44:09] <psha> that question was about global naming
[17:45:05] <psha> maybe they want to remove every occurance of emc from git history?
[17:46:39] <psha> for example i've installation with some out-of tree GUI
[17:46:40] <micges> who knows, request for change emc name was enough stupid
[17:46:44] <psha> it depends on 'import emc'
[17:46:45] -!- GoSebGo has quit [Quit: Bye]
[17:46:54] <psha> and now it'll break on package upgade
[17:47:10] -!- GoSebGo [GoSebGo!~Seb@] has joined #linuxcnc-devel
[17:48:18] <micges> that's reason that my installation can't upgrade emc packages
[17:48:28] <micges> *installations
[17:50:16] <jepler> there is no written agreement about exactly what we "must" change. a Python module called 'emc' that prints a deprecation warning but otherwise works would probably be fine.
[17:51:01] <jepler> there are going to be lots of uses of the substrings emc and EMC (for instance, we're not going to break all inifiles that specify [EMCIO])
[17:54:27] <psha> emc is used nearly everywhere in glade files
[17:54:31] <psha> and so on
[17:54:45] <psha> replacing every occurance is nearly impossible
[17:55:22] <seb_kuzminsky> micges: if you "apt-get install linuxcnc", it should replace the old emc2 package correctly
[17:55:39] <seb_kuzminsky> bbl
[17:57:53] <jepler> $ python -c 'import emc; print emc.stat'
[17:57:53] <jepler> -c:1: DeprecationWarning: the emc module is deprecated; use the linuxcnc module instead
[17:57:56] <jepler> <type 'linuxcnc.stat'>
[17:58:17] <jepler> psha: would this make you feel better?
[17:59:50] <jepler> note that as far as specifying [DISPLAY]DISPLAY goes, there is a special case in the linuxcnc script where xemc and tkemc are rewritten to xlinuxcnc and tklinuxcnc.
[18:01:18] <JT-Shop> any more thoughts on just having xcnc and tkcnc?
[18:02:36] <psha> jepler: sure
[18:02:46] <psha> at least it will make transition a bit painless
[18:03:16] <psha> also what was descision about lcnc?
[18:03:25] <psha> at least for me it sounds nice
[18:03:38] <jepler> lcnc is a non-starter. it is used by some other company for a cnc control
[18:03:38] <psha> tklinuxcnc and xlinuxcnc are toooooo looooong
[18:03:44] <psha> ah
[18:03:45] <andypugh> But belongs to someone else. As does lxcnc
[18:03:46] <psha> got it
[18:04:05] <andypugh> How about we switch ti cyrillic?
[18:04:25] <psha> heh лцнц is not used i guess :)
[18:04:34] <JT-Shop> :)
[18:04:35] <psha> or even lснс
[18:05:14] <psha> btw o/о and c/с tricks are very popular in government contracts :)
[18:09:45] <jepler> you must have me confused with someone who enjoyed this trademark-induced nightmare enough that this time he's deliberately trying to pick a fight
[18:13:11] <CIA-99> 03tissf 07v2.5_branch * rcea1e1cd3d16 10/ (20 files in 19 dirs): French docs - rename in README signaled by Kent
[18:17:09] <JT-Shop> Thanks Francis
[18:17:57] -!- e-ndy [e-ndy!~jkastner@static-84-242-102-36.net.upcbroadband.cz] has joined #linuxcnc-devel
[18:18:31] <tissf_> oups! bad commit message
[18:19:17] <tissf_> can't --amend after push?
[18:19:40] <JT-Shop> I know what you meant :)
[18:20:50] <JT-Shop> bbl
[18:21:03] <tissf_> ok
[18:21:06] -!- GoSebGo has quit [Ping timeout: 260 seconds]
[18:23:34] <psha> tissf_: if nobody pulled tree - you may
[18:23:54] <tissf_> how please?
[18:25:16] <tissf_> git commit --amend cea1e1cd3 don't work
[18:27:12] <tissf_> git message: did not match any file know to git
[18:32:33] <tissf_> it's like that!
[18:37:26] <micges> tissf_: did you fixed it?
[18:37:50] <tissf_> no it's like that :)
[18:42:46] <tissf_> micges: you know git command for to correct a commit message after pushing?
[18:43:13] <cradek> I recommend just leave it
[18:43:47] <cradek> it can cause big problems for others if you push a non-fastforward
[18:44:16] <cradek> it is ok in emergencies if you warn everyone like we have done for incorrect merges
[18:44:40] <tissf_> ok cradek I leave it
[18:44:42] -!- n2diy has quit [Quit: Ex-Chat]
[18:45:18] <cradek> I recommend before every push always: git push --dry-run, then git log
[18:50:02] <tissf_> ok
[18:55:36] <CIA-99> 03jepler 07v2.5_branch * r56368c9de829 10/scripts/emc: commandline compatibility script
[18:55:37] <CIA-99> 03jepler 07v2.5_branch * rb602d2cbf857 10/src/Makefile: install commandline compatibility script
[18:55:38] <CIA-99> 03jepler 07v2.5_branch * r11cc72ddb360 10/lib/python/emc.py: python compatibility module
[18:59:01] -!- EmcPT has quit [Remote host closed the connection]
[19:00:03] -!- awallin has quit [Remote host closed the connection]
[19:03:11] <CIA-99> 03jepler 07master * rf074bdb96cb2 10/ (145 files in 53 dirs): Merge remote branch 'origin/v2.5_branch'
[19:03:13] <CIA-99> 03jepler 07master * r642ca70e98ad 10/lib/python/emc.py: Revert "python compatibility module"
[19:03:13] <CIA-99> 03jepler 07master * r48340ad73650 10/src/Makefile: Revert "install commandline compatibility script"
[19:03:15] <CIA-99> 03jepler 07master * rb6228b22acf6 10/scripts/emc: Revert "commandline compatibility script"
[19:04:21] <jepler> (so the compatibility names will be in 2.5 but not post-2.5)
[19:07:52] -!- awallin [awallin!awallin@lakka.kapsi.fi] has joined #linuxcnc-devel
[19:08:27] -!- IchGucksLive has quit [Quit: ChatZilla 0.9.87 [Firefox 9.0.1/20120117091749]]
[19:32:34] syyl__ is now known as syyl
[19:35:04] -!- Mjolinor has quit [Ping timeout: 240 seconds]
[19:43:08] <CIA-99> 03micges 07v2.5_branch * rf9c92c04fa9a 10/src/po/pl.po: Update Polish translation, also few rebrandings
[19:48:47] -!- micges has quit [Quit: Ex-Chat]
[19:49:48] -!- psha has quit [Quit: Lost terminal]
[19:54:05] -!- vladimirek [vladimirek!~vladimire@bband-dyn226.178-40-36.t-com.sk] has joined #linuxcnc-devel
[19:55:25] -!- isssy has quit [Quit: Bye Bye]
[20:07:02] <CIA-99> 03jthornton 07v2.5_branch * rd42a1d1e169b 10/docs/src/code/NML_Messages.txt: Docs: remove double back tics
[20:07:02] <CIA-99> 03jthornton 07v2.5_branch * rc7c33032a146 10/docs/src/config/ (4 files): Docs: fix link
[20:07:03] <CIA-99> 03jthornton 07v2.5_branch * r6731b66ebe21 10/docs/src/ (index.tmpl index_de.tmpl index_es.tmpl index_pl.tmpl): Docs: fix title
[20:25:24] -!- e-ndy has quit [Quit: Ex-Chat]
[20:35:33] -!- pfred1 has quit [Quit: switching machines]
[20:42:24] -!- syyl has quit [Quit: Leaving]
[20:57:17] -!- sumpfralle has quit [Ping timeout: 252 seconds]
[21:21:24] -!- Mjolinor has quit [Ping timeout: 240 seconds]
[21:22:38] -!- pfred1 has quit [Quit: Lost terminal]
[21:51:44] -!- vladimirek has quit [Ping timeout: 240 seconds]
[21:53:25] -!- sumpfralle has quit [Read error: Operation timed out]
[22:12:10] -!- factor has quit [Read error: Connection reset by peer]
[22:56:38] -!- mhaberler has quit [Quit: mhaberler]
[23:05:48] -!- acemi has quit [Quit: WeeChat 0.3.2]
[23:19:29] <CIA-99> 03jepler 07v2.5_branch * rdcbaa105aaa5 10/src/emc/usr_intf/axis/extensions/emcmodule.cc: rebrand linuxcnc python module
[23:28:30] -!- EmcPT [EmcPT!~EmcPT@bl13-194-203.dsl.telepac.pt] has joined #linuxcnc-devel
[23:29:12] -!- ries has quit [Read error: Connection reset by peer]
[23:34:08] -!- jonnyATroot has quit [Ping timeout: 240 seconds]