Back
[00:01:52] -!-
robh__ has quit [Ping timeout: 246 seconds]
[00:04:13] -!-
theorbtwo has quit [Ping timeout: 255 seconds]
[00:04:16] theorb is now known as
theorbtwo
[00:04:51] -!-
Dannyboy has quit [Remote host closed the connection]
[00:23:10] joel is now known as
Guest6248
[00:23:38] -!-
skunkworks has quit [Ping timeout: 240 seconds]
[00:31:03] davec is now known as
Guest63696
[00:31:52] -!-
mhaberler has quit [Ping timeout: 272 seconds]
[00:34:32] -!-
servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[00:34:46] -!-
dgarr [
[email protected]] has joined #emc-devel
[00:38:09] -!-
skunkworks [skunkworks!~chatzilla@str-bb-cable-south2-static-6-412.dsl.airstreamcomm.net] has joined #emc-devel
[01:10:48] -!-
tlab has quit [Quit: Leaving]
[01:23:28] -!-
Guest6248 has quit [Quit: Visitor from www.linuxcnc.org]
[01:38:33] -!-
frank222 has quit [Quit: Visitor from www.linuxcnc.org]
[02:11:59] -!-
fatpandas has quit [Quit: leaving]
[02:32:46] -!-
ries has quit [Quit: ries]
[02:34:15] -!-
krushia has quit [Ping timeout: 240 seconds]
[02:54:41] -!-
pcw_home has quit [Ping timeout: 255 seconds]
[03:01:53] -!-
rooks has quit [Ping timeout: 264 seconds]
[03:51:09] -!-
rooks has quit [Quit: So long, and thanks for all the fish.]
[04:27:01] -!-
skunkworks has quit [Ping timeout: 255 seconds]
[04:39:48] -!-
Valen has quit [Quit: Leaving.]
[05:30:55] -!-
odiug has quit [Ping timeout: 255 seconds]
[05:37:33] -!-
dgarr has quit [Ping timeout: 276 seconds]
[05:39:40] -!-
Dallur has quit [Quit: Leaving.]
[05:41:33] -!-
jj_ has quit [Quit: Ex-Chat]
[06:30:49] -!-
Emcrules_Laptop has quit [Ping timeout: 240 seconds]
[06:42:09] -!-
micges [
[email protected]] has joined #emc-devel
[06:47:48] -!-
mhaberler [
[email protected]] has joined #emc-devel
[06:53:46] -!-
JuniperJaxx has quit [Ping timeout: 272 seconds]
[07:00:22] -!-
LawrenceG has quit [Remote host closed the connection]
[07:02:03] -!-
adb has quit [Ping timeout: 276 seconds]
[07:26:44] -!-
drill has quit [Ping timeout: 265 seconds]
[07:35:58] -!-
micges has quit [Quit: Ex-Chat]
[07:41:59] -!-
psha [
[email protected]] has joined #emc-devel
[07:51:52] -!-
Dannyboy has quit [Remote host closed the connection]
[08:42:17] -!-
mhaberler has quit [Ping timeout: 276 seconds]
[08:53:03] -!-
Paragon39 has quit [Ping timeout: 240 seconds]
[09:13:31] -!-
mhaberler [
[email protected]] has joined #emc-devel
[09:23:49] -!-
mhaberler has quit [Ping timeout: 240 seconds]
[09:26:00] -!-
awallin [
[email protected]] has joined #emc-devel
[09:29:31] -!-
mhaberler [
[email protected]] has joined #emc-devel
[09:34:43] -!-
robh__ [
[email protected]] has joined #emc-devel
[09:57:30] -!-
maximilian_h [
[email protected]] has joined #emc-devel
[09:58:56] -!-
mhaberler has quit [Ping timeout: 250 seconds]
[10:07:41] -!-
mozmck has quit [Ping timeout: 265 seconds]
[11:00:14] -!-
mozmck [mozmck!~moses@client-173.225.233.241.dfwtx.partnershipbroadband.com] has joined #emc-devel
[11:14:06] -!-
maximilian_h has quit [Quit: Leaving.]
[11:20:34] -!-
micges_work [
[email protected]] has joined #emc-devel
[11:25:14] -!-
ries [
[email protected]] has joined #emc-devel
[11:39:30] -!-
mhaberler [
[email protected]] has joined #emc-devel
[12:41:25] -!-
mozmck has quit [Ping timeout: 240 seconds]
[12:41:29] -!-
rooks has quit [Ping timeout: 255 seconds]
[12:45:12] -!-
mozmck [mozmck!~moses@client-173.225.233.241.dfwtx.partnershipbroadband.com] has joined #emc-devel
[12:49:46] -!-
Guest958 has quit [Client Quit]
[12:56:01] -!-
psha has quit [Quit: leaving]
[13:14:00] -!-
SteveStallings [
[email protected]] has joined #emc-devel
[13:16:32] -!-
steves_logging has quit [Ping timeout: 250 seconds]
[13:19:27] -!-
skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.68.115.41.210] has joined #emc-devel
[13:21:36] SteveStallings is now known as
steves_logging
[13:33:12] -!-
micges_work [
[email protected]] has parted #emc-devel
[13:45:37] -!-
psha [
[email protected]] has joined #emc-devel
[14:05:02] -!-
adb [
[email protected]] has joined #emc-devel
[14:12:39] <mhaberler> psha: presente?
[14:24:19] <psha> yea
[14:28:08] <mhaberler> I think I nailed the real problem with the two o-word calls per block. It's much simpler than that.
[14:28:10] <mhaberler> since o-word calls are handled outside 'block' execution, calling a sub from within execute_block() changes the _setup state... and overwrites m_modes if the subroutine calls m commands for example
[14:28:48] <mhaberler> it's an even stronger argument to excise the global _setup and make it an instance variable
[14:29:28] <mhaberler> I naively thought calling execute() with the o-sub call will not trample on _setup. well it does, big time
[14:30:04] <mhaberler> if one had a way to stack _setup, this would be a non-issue
[14:30:17] -!-
nullie has quit [Quit: Ex-Chat]
[14:31:43] <mhaberler> the assumption seems to be sort of 'atomic execution' of a block wrt _setup, and that of course is violated if you replace testing a t or m flag and call a sub instead, which modifies _setup
[14:33:09] <mhaberler> example: assume I have t2m6 to execute
[14:33:10] <mhaberler> the T sub has an M66 statement in it
[14:33:12] <mhaberler> by the time I get around in the 't2m6' execute_block m6 has been overwritten by m66 :-/
[14:33:30] <mhaberler> which of course fails to wexecute the m6 after the t2..
[14:34:18] <mhaberler> anyway, gotta run, but at least I know where the 'last remaining issue is' (famous last words ;-)
[14:35:21] <mhaberler> naive me: I see a proper class definition in c++ for Interp, I think - great, OO design, local state and all, and bang - there you run into a FORTRAN COMMON block
[14:44:46] -!-
dgarr [
[email protected]] has joined #emc-devel
[14:46:04] -!-
nullie has quit [Quit: Ex-Chat]
[14:51:00] <psha> mhaberler: btw, drop a notice about 'realtime start' to the man on emc-user list
[15:01:33] -!-
kent_cnc has quit [Ping timeout: 265 seconds]
[15:39:51] -!-
isssy has quit [Quit: Visitor from www.linuxcnc.org]
[15:48:44] -!-
adb has quit [Ping timeout: 240 seconds]
[15:49:33] -!-
EDocToor has quit [Read error: Connection reset by peer]
[15:51:16] -!-
wobblybootie has quit [Ping timeout: 265 seconds]
[15:53:15] -!-
KimK [
[email protected]] has joined #emc-devel
[16:01:40] -!-
awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc-devel
[16:02:11] -!-
micges [
[email protected]] has joined #emc-devel
[16:02:20] -!-
EDocToor [
[email protected]] has joined #emc-devel
[16:03:08] -!-
EDocToor has quit [Client Quit]
[16:21:36] -!-
vezza has quit [Quit: Sto andando via]
[16:31:03] -!-
drill has quit [Ping timeout: 240 seconds]
[16:34:19] -!-
tom3p [
[email protected]] has joined #emc-devel
[16:42:48] -!-
KimK [
[email protected]] has parted #emc-devel
[16:47:01] -!-
adb [
[email protected]] has joined #emc-devel
[17:03:14] -!-
Paragon39 has quit [Quit: ChatZilla 0.9.86 [Firefox 3.5.16/20101130074636]]
[17:06:53] -!-
fatpandas has quit [Quit: leaving]
[17:11:00] -!-
nullie has quit [Quit: Ex-Chat]
[17:16:05] -!-
LawrenceG [
[email protected]] has joined #emc-devel
[17:53:22] -!-
Dallur has quit [Quit: Leaving.]
[18:26:05] -!-
vezza has quit [Remote host closed the connection]
[18:41:52] -!-
andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc-devel
[18:44:05] <andypugh> cradek: So, to recreate the sim problem with s64: make clean; ./autogen.sh; ./configure --enable-simulator; make. Should show it?
[18:45:50] <andypugh> Ah, yes, it seems to.
[18:47:37] -!-
tom3p [
[email protected]] has parted #emc-devel
[18:53:38] -!-
Guest852 has quit [Ping timeout: 255 seconds]
[18:56:37] <cradek> can we fix that with some typedefs somewhere or some other trivially easy solution?
[18:56:43] <cradek> and by we, I mean jepler
[18:57:15] <andypugh> changing the s64 to __s64 seems to do the trick with the patch I sent you
[18:57:36] <andypugh> Should I send a whole new patch file, or do you want to just try the change?
[18:57:45] <cradek> if you send again, would you fix your git email please? It says simply "Andy <...>" and it should be your full name
[18:58:04] <cradek> I can do either, since I patched up your git emails in the ones I applied already
[18:58:09] <andypugh> Yeah, I was wondering about that.
[18:58:13] <cradek> but it'll have to wait until late tonight-ish
[18:58:46] <cradek> whichever you want me to do is fine with me
[18:59:33] <andypugh> Well, I just moved the variable definitions below the description text, so that it is more obvious that they exist, so I will try to make a new patch
[18:59:54] <cradek> ok, cool
[19:01:22] <jepler> cradek: we went down this road once before (commit c24f046a) and ultimately elected to prefer the __s32 types instead (commit 8b77b38a).
[19:01:36] <jepler> Unfortunately I don't recall the surrounding discussion to recall why we made that decision at the time
[19:01:49] <cradek> oh ok.
[19:03:45] <andypugh> Is there a list of preferred nomenclatures? I assume we don't use __int and __char?
[19:04:05] <jepler> andypugh: not that I know of
[19:04:43] <jepler> maybe it belongs in intro.3rtapi or something
[19:04:47] <andypugh> Is this only relevant to comp? Or should Icheck everything?
[19:05:06] <jepler> it's relevant to any code that is built with --enable-simulator
[19:05:15] <jepler> so hardware drivers don't count
[19:06:20] <andypugh> OK. Not that bldc.comp is much use without hardware drivers to talk to.
[19:07:05] <jepler> reasons to build it for the simulator include being able to use gdb
[19:10:57] <andypugh> I sometimes wish I knew how that worked
[19:12:40] <jepler> looks like the discussion of __s32 vs s32 was here:
http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2007-04-15.txt down at 17h
[19:13:30] -!-
dgarr has quit [Ping timeout: 240 seconds]
[19:14:02] <jepler> walkthrough of debugging realtime components with --enable-simulator and gdb:
http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?DebuggingRtapi
[19:32:28] <seb_kuzminsky> andypugh: the interface to new gearchanger in master is way better than the one in 2.4, thank you
[19:33:00] -!-
mhaberler_ [
[email protected]] has joined #emc-devel
[19:33:38] <andypugh> Can you push that patch which inverts the scale?
[19:33:51] <seb_kuzminsky> the gear_number pin doesn't work as advertised
[19:33:58] <seb_kuzminsky> it can't select gear 0 i think
[19:34:46] <andypugh> I will have a look later. ( I am not currently in a branch which contains it)
[19:34:47] -!-
mhaberler has quit [Ping timeout: 240 seconds]
[19:34:54] mhaberler_ is now known as
mhaberler
[19:37:40] <seb_kuzminsky> andypugh: i'm looking at the gearchange patch you mailed to the list on jan 23
[19:38:05] <seb_kuzminsky> the scale change looks good
[19:38:15] <seb_kuzminsky> (but i haven't tested it yet)
[19:38:35] <seb_kuzminsky> there are some whitespace changes that make it a little more confusing than it could be
[19:44:25] <seb_kuzminsky> all the arithmetic with embedded conditionals is a bit awkward for me to parse
[19:46:26] -!-
CIA-4 has quit [Ping timeout: 272 seconds]
[19:47:41] -!-
acemi has quit [Quit: WeeChat 0.3.2]
[19:53:14] <andypugh> I can't remember why I did it that way. The whole thing started as a forum post that I wrote without access to a compiler or linux.
[19:55:15] -!-
CIA-2 [
[email protected]] has joined #emc-devel
[20:01:41] <seb_kuzminsky> heh
[20:06:57] -!-
awallin_ has quit [Remote host closed the connection]
[20:18:36] -!-
micges has quit [Ping timeout: 250 seconds]
[20:19:53] <seb_kuzminsky> grrr
[20:20:53] <seb_kuzminsky> if you're running a kvm virtual machine, and the host computer runs out of disk, the guest's disk silently becomes corrupted :-(
[20:21:27] <andypugh> That sounds suboptimal.
[20:22:48] <psha> seb_kuzminsky: is that bad?
[20:23:17] <psha> i mean why not to run build slave with COW disk?
[20:23:36] <seb_kuzminsky> i was able to recover my vm by booting off an ubuntu install cd and fscking the root partition
[20:23:40] <seb_kuzminsky> it's not the buildslaves
[20:24:21] <psha> ah, then it's bad ;)
[20:26:29] <andypugh> So many of my acquaintances use "fscking" as a stand-in for something much ruder on online forums that I tend to always read it the rude way. Which puts a different slant on that statement.
[20:26:41] <seb_kuzminsky> heh
[20:28:02] <seb_kuzminsky> debsums and "git --fsck" both pass, so i guess there's no subtle damage lurking...
[20:29:00] <psha> i've recently upgraded server and got one broken memory bank
[20:29:21] <psha> i was running git --fsck for two days trying to figure out _where_ it's broken
[20:29:37] <psha> until figured that fs _was_ ok (before checking ;) :)
[20:30:51] -!-
andypugh_ [andypugh_!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc-devel
[20:30:51] -!-
andypugh has quit [Read error: Connection reset by peer]
[20:30:51] andypugh_ is now known as
andypugh
[20:31:41] <andypugh> If I default gear_number to -1, that should fix it?
[20:38:30] <seb_kuzminsky> i think what you're saying is this:
[20:38:40] <seb_kuzminsky> by default, gear_number is -1 and all the gear_N_sel bits are 0, and gear 0 is selected
[20:39:23] <seb_kuzminsky> if any gear_N_sel bits are 1, then select the highest-numbered gear that has its gear_N_sel bit set
[20:39:48] <seb_kuzminsky> if gear_number is greater than -1 (and less than personality), then select the indicated gear
[20:40:01] -!-
WalterN has quit [Read error: Connection reset by peer]
[20:40:06] <andypugh> Indeed
[20:40:15] <seb_kuzminsky> that sounds ok
[20:40:36] <seb_kuzminsky> i think i'd prefer to get rid of the pin named "sel"
[20:40:43] <andypugh> I am also currently doing away with "personality"
[20:40:56] <seb_kuzminsky> how are you choosing the number of gears?
[20:41:09] <andypugh> num_gears as a modparam
[20:41:17] <seb_kuzminsky> nice
[20:41:47] <seb_kuzminsky> i'd expect num_gears=4 to give me gears 0-3
[20:42:02] <seb_kuzminsky> rather than the personality+1 behavior it currently has
[20:42:02] <andypugh> Yes. I was wondering what to do about that
[20:42:36] <andypugh> My logic was that 5 gears would have 5 select bits, and 0 indicates neutral.
[20:43:04] <seb_kuzminsky> interesting
[20:43:21] <andypugh> Habit. Neutral is gear 0 at work :-)
[20:43:27] <seb_kuzminsky> ah :-)
[20:44:06] <andypugh> As for the "sel" pin, it is stupid, but needs to be there to not break old configs.
[20:44:46] <seb_kuzminsky> i dont mind breaking old configs when we have a new release branch, if it makes the new situation cleaner
[20:45:26] <seb_kuzminsky> backwards compatibility is nice, in moderation ;-)
[20:46:40] <andypugh> ditto the "reverse" bit which is also a special case
[20:49:37] <andypugh> hmm, it seems that modparam int num_gears will only pass one integer. So gearchange can be a singleton, or I need to get a bit cleverer.
[20:51:39] <andypugh> (jepler did hint that the reason that the modparam statement in comp is undocumented is that it doesn't work too well)
[21:10:51] -!-
CraigW76 has quit [Remote host closed the connection]
[21:15:39] -!-
CraigW76 has quit [Remote host closed the connection]
[21:22:19] <andypugh> seb_kuzminsky: I have spotted a good reason to have the inverted ratio that I had before. A gear with an undefined or zero ratio leads to zero output if you mutliply, but does daft stuff otherwise..
[21:28:20] <seb_kuzminsky> you want to use a ratio of 0 for your neutral?
[21:34:08] -!-
micges [
[email protected]] has joined #emc-devel
[21:35:46] -!-
fatpandas has quit [Quit: leaving]
[21:41:55] -!-
CraigW76 has quit [Remote host closed the connection]
[21:44:24] -!-
e3m has quit [Ping timeout: 250 seconds]
[21:46:46] e3m_ is now known as
e3m
[22:02:08] -!-
psha has quit [Quit: leaving]
[22:04:28] <andypugh> Seems sensible
[22:05:10] <andypugh> My thinking is that you can have zero output if not in gear.
[22:05:20] <andypugh> Though that might get a bit confusing.
[22:06:16] <andypugh> Anyway, I have now massively modified the comp.
[22:07:34] <andypugh> It takes an optional num_gears modparam. If you choose to use that then the "sel" and "reverse" pins disappear. If you don't use that then it defaults to looking just like the old version.
[22:11:25] -!-
rooks has quit [Ping timeout: 240 seconds]
[22:16:00] -!-
skunkworks has quit [Ping timeout: 265 seconds]
[22:25:30] -!-
odiug has quit [Ping timeout: 240 seconds]
[22:26:30] -!-
micges has quit [Ping timeout: 240 seconds]
[22:29:35] -!-
micges [
[email protected]] has joined #emc-devel
[22:38:42] -!-
micges has quit [Quit: Ex-Chat]
[22:38:48] -!-
e3m has quit [Quit: Life without danger is a waste of oxygen]
[22:54:15] -!-
Fox_Muldr has quit [Ping timeout: 240 seconds]
[22:59:25] -!-
toastyde1th has quit [Ping timeout: 255 seconds]
[23:08:43] -!-
CraigW76 has quit [Remote host closed the connection]
[23:14:46] <seb_kuzminsky> andypugh: i think that's the wrong direction of development. I'd rather see the new, better behavior replace the old, more confusing behavior
[23:15:49] <andypugh> It does, unless you use it the old, confusing way.
[23:16:02] <seb_kuzminsky> the legibility of the code is important to me, and backwards compatibility often makes thing more convoluted and hard to follow
[23:18:29] <andypugh> let me send you the latest file:
[23:20:13] <andypugh> I only started fiddling with gearhange again because you complained that my updates had broken your old config with poor backwards-compatibility <sulk>
[23:20:33] <seb_kuzminsky> i'm sorry to be such a complainer
[23:21:00] <seb_kuzminsky> i saw the commit message and thought "great, i wont have to change my config"
[23:21:39] <seb_kuzminsky> but then when i realized i actually did have to change the config, i thought "i wish the new config i just wrote was cleaner"
[23:22:21] <andypugh> I am trying for the best of both worlds now.
[23:22:36] <seb_kuzminsky> thanks :-)
[23:23:48] <andypugh> Old configs just work, and have all the pins they used to have. If you specify a number of gears, though, (even 2) you get the new, cleaner, more sensible pinout, (sel and reverse just disappear)
[23:24:55] <seb_kuzminsky> I can see how that's convenient for users, who don't have to update their old 2.4 configs when they upgrade to 2.5, that's a good thing
[23:25:43] <andypugh> I have a file transfer window open, if you want to acknowledge it
[23:25:51] -!-
tlab has quit [Quit: Leaving]
[23:26:08] <seb_kuzminsky> i'm behind a nat, so i can't receive dcc :-(
[23:26:28] -!-
dgarr [
[email protected]] has joined #emc-devel
[23:26:52] <andypugh> email?
[23:27:07] <andypugh> highlab?
[23:27:16] <seb_kuzminsky> the problem i have with doing gearchange like that (and maybe it's just me being cranky and particular), is that we now have two gearchange components in 2.5, they share a name and a manpage and a comp, but they behave differently from each other
[23:27:42] <seb_kuzminsky> it's confusing when you read the code, and it's confusing when you read the manpage, and it's confusing when you read a hal configuration
[23:27:46] <andypugh> Yeah. Very linuxy I think :-)
[23:27:55] <seb_kuzminsky> heh
[23:28:02] <seb_kuzminsky> my email is
[email protected]
[23:28:15] <seb_kuzminsky> or you can pastebin it and copy the link to this irc channel
[23:29:47] <andypugh> forgot about that.
[23:29:49] <seb_kuzminsky> got it
[23:30:14] <andypugh> it segfaults in num_gears mode, at the moment :-)
[23:30:40] <andypugh> Consider it light reading, not sofware
[23:34:07] <andypugh> Maybe the solution is to document that the old version is deprecated, and call the new one "gearchange_multi" or something?
[23:34:36] <andypugh> though that gives all the pins silly names, so I don't like it.
[23:46:47] -!-
CraigW76 has quit [Read error: Connection reset by peer]
[23:51:51] -!-
skunkworks [skunkworks!~chatzilla@str-bb-cable-south2-static-6-412.dsl.airstreamcomm.net] has joined #emc-devel
[23:52:50] -!-
owhite has quit [Ping timeout: 240 seconds]
[23:52:56] <seb_kuzminsky> andypugh: i vote for replacing the old broken gearchange with the nice new gearchange, and forget about backwards compatibitlity
[23:53:15] -!-
Guest820 has quit [Quit: Visitor from www.linuxcnc.org]
[23:53:17] <seb_kuzminsky> there's only one sample config that uses gearchange (tormach), i don't know how widely used it is in people
[23:53:22] <seb_kuzminsky> 's private custom configs
[23:53:38] <seb_kuzminsky> just my opinion
[23:53:41] <andypugh> The old gearchange isn't actually broken, but it is a bit silly.
[23:54:00] <seb_kuzminsky> btw, i totally dig that the new one lets you set the scale for all gears, instead of just the second gear
[23:54:11] <seb_kuzminsky> bbl
[23:54:17] <seb_kuzminsky> thanks for working on it, it's much better now