#linuxcnc-meet Log v0.1

logged_meetings

Recorded agenda
date_time
id
event
id
user
message
2013-09-28 17:00:04
61
start

archivist
!start 201309
2013-09-28 17:04:51
62
topic
1:
archivist
201309 DISCUSS: 1: formalize the current practice of freezing the agenda on or around the Wednesday
2013-09-28 17:14:08
63
topic
1:
archivist
201309 VOTE: 1: formalize the current practice of freezing the agenda on or around the Wednesday
2013-09-28 17:16:56
64
topic
2:
archivist
201309 DISCUSS: 2: proposal Michael Haberler's 'jog-while-paused'. [mariusl]
2013-09-28 17:49:21
65
topic
MEETING:
archivist
201309 NEXT MEETING: Saturday 2013-10-26 1600UTC
2013-09-28 17:49:56
66
end

archivist
!end 201309


log row count=189
2013-09-28 17:00:25 mariusl joined chan
2013-09-28 17:01:33 cradek one of the items is mine this time. archivist, do you want to run things?
2013-09-28 17:02:21 archivist I cant only you have topic powers
2013-09-28 17:02:37 cradek oh oops
2013-09-28 17:02:45 cradek and I have to go op myself
2013-09-28 17:03:22 cradek I bet you can do it now
2013-09-28 17:05:25 cradek I think it is good to avoid last-minute additions for several reasons: so people have time to thoughtfully consider the agenda items, so they have time to start a discussion, and so they have full knowledge allowing them to decide whether to plan to attend.
2013-09-28 17:05:31 mhaberler fine in principle - I'd suggest a warning mail a day or two in advance
2013-09-28 17:06:14 cradek I like the idea of a reminder on about Monday.
2013-09-28 17:06:23 mhaberler good
2013-09-28 17:06:31 archivist I feel the cut off tends be the point where people get reminded therefore its too late
2013-09-28 17:06:49 cradek will one of you do that job?
2013-09-28 17:06:49 mhaberler thats what I want to avoid
2013-09-28 17:07:06 archivist cut off is only ok is the reminder has been x days before
2013-09-28 17:07:22 archivist is/if
2013-09-28 17:07:31 cradek yeah I'd prefer people put things on the agenda early in the month (so there is time to really figure something out) but that's not really human nature
2013-09-28 17:07:57 mhaberler hopeless ;)
2013-09-28 17:08:00 cradek a good example is today's #2 which mhaberler and I discussed for a few hours but there has been no time to really dig in
2013-09-28 17:08:43 mariusl sorry i am guilty there but it has been raised before
2013-09-28 17:08:50 cradek maybe it should be even earlier? perhaps the reminder at -2 weeks and the freeze at -1 week
2013-09-28 17:09:10 tjtr33 sorry its late. JogWhilePaused has some prev discussion over more than 1 month tho.
2013-09-28 17:09:15 mhaberler no, that's too much of a freeze IMO
2013-09-28 17:09:47 skunkworks joined chan
2013-09-28 17:09:53 archivist late cut-off just means it gets discussed and then it can then be ruminated upon later after a meeting too, the meeting after in devel was a good example last month
2013-09-28 17:10:35 cradek ok should we shoot for reminder on monday (need a volunteer), freeze on wed (I can do it)?
2013-09-28 17:10:46 mhaberler fine
2013-09-28 17:10:52 tjtr33 agreed
2013-09-28 17:10:57 mariusl do we expect to have resolution on every point or can sonme things be discussed later to complete
2013-09-28 17:10:58 cradek the reminder is something a suitably clever person could script
2013-09-28 17:11:02 mhaberler I volunteer
2013-09-28 17:11:22 mhaberler this goes do -dev or -users or both?
2013-09-28 17:11:27 cradek mariusl: lots of things stay undecided
2013-09-28 17:11:29 archivist both
2013-09-28 17:11:31 mhaberler ok
2013-09-28 17:11:44 tjtr33 the discussion was unknown to me, as i monitored the user channel while it went on
2013-09-28 17:11:48 cradek my problem with both is it always starts two separate discussions
2013-09-28 17:12:04 archivist users matter too :)
2013-09-28 17:12:11 mhaberler ha
2013-09-28 17:12:16 cradek yeah, I'd propose only -users
2013-09-28 17:12:17 mhaberler how come
2013-09-28 17:12:21 cradek but I don't have strong feelings
2013-09-28 17:12:29 mhaberler fine, then --users only
2013-09-28 17:12:33 mhaberler .. making repeating calendar entry.. for a year.
2013-09-28 17:12:34 cradek I suspect all people on -dev are also on -users
2013-09-28 17:12:46 mariusl should be
2013-09-28 17:13:06 cradek thanks for taking care of the announcement mhaberler
2013-09-28 17:13:50 cradek I'm ready for a vote on announce-monday freeze-wednesday
2013-09-28 17:13:57 mhaberler so that be on fourth monday every month I guess
2013-09-28 17:13:58 mhaberler fine
2013-09-28 17:14:08 mariusl second that
2013-09-28 17:14:16 cradek mhaberler: no, it's the monday before the 4th saturday
2013-09-28 17:14:35 mhaberler yees, I hope my calendar can pull that off
2013-09-28 17:14:40 cradek heh
2013-09-28 17:14:44 jepler joined chan
2013-09-28 17:14:45 cradek I vote yes
2013-09-28 17:14:51 tjtr33 yes
2013-09-28 17:15:10 mhaberler yes
2013-09-28 17:15:15 cradek jepler: we're voting on #1 which has been changed to announcement on monday before the meeting, then agenda freeze wednesday, mhaberler is going to do the announcement, I'll do the freeze
2013-09-28 17:15:15 mariusl yes
2013-09-28 17:16:02 jepler yes
2013-09-28 17:17:37 archivist mariusl, to you
2013-09-28 17:18:44 mariusl i think it is a requirement that has to be pushed fast. i understand the amount of work but i was hoping to see at least a phased implimentation if possible
2013-09-28 17:19:25 cradek I think mhaberler posted a link to our conversation - did you get a chance to read it?
2013-09-28 17:19:55 mariusl i did an d hence my understanding for the effort involved
2013-09-28 17:20:36 mhaberler well in the form cradek and me agreed it mostly depends on ja3 maturity - I looked into that and it seems its a bit in flux still
2013-09-28 17:20:59 cradek ... well that and you doing a lot of new work
2013-09-28 17:21:05 mhaberler q: should we do a fallback for the case ja3 doesnt happen
2013-09-28 17:21:06 mariusl i suggest that we as some users to get an understanding of what is pressing and what is nice to have maybe that will change the goal
2013-09-28 17:21:43 tjtr33 whats wrong with the code i run now ( from michaels repo ) ?
2013-09-28 17:21:49 mhaberler well as it stands - without wheel and command jog - its fine on master and ubc3
2013-09-28 17:22:11 cradek mhaberler: could you find that link to the archive for tjtr33 please?
2013-09-28 17:22:11 mhaberler it should be integrated with wheel and command jog (from UI) though
2013-09-28 17:22:16 mhaberler sure
2013-09-28 17:22:49 mariusl most only have those to trverse with anyway
2013-09-28 17:22:55 mhaberler http://emc.mah.priv.at/irc/%23linuxcnc-devel/2013-09-26.html#19:13:57
2013-09-28 17:23:14 tjtr33 got, duno if i get thru it inside this meeting
2013-09-28 17:23:23 mhaberler right, a bit longish
2013-09-28 17:24:10 mhaberler I think it mostly is a question if we do a fallback if ja3 doesnt shape up in time
2013-09-28 17:24:16 mhaberler and what the fallback should contain
2013-09-28 17:25:05 mariusl what is minimum functionality
2013-09-28 17:25:16 mhaberler the current one I'd think
2013-09-28 17:25:18 cradek I don't agree with that; I want it based on ja3 and anything else is a waste of effort
2013-09-28 17:25:43 cradek I have worked on stabilizing ja3 and I am willing to do more work of that kind
2013-09-28 17:26:31 mhaberler any gut feeling for 'usefulness for a rebase' timeframe?
2013-09-28 17:26:38 tjtr33 got a synopsis of the agreement , back again
2013-09-28 17:26:42 jepler cradek: so essentially, if the choice was jog-while-paused *OR* ja3, you want ja3
2013-09-28 17:26:52 cradek wish seb was here to answer that
2013-09-28 17:26:54 mariusl i cannot comment on where it must be released as i dont have insight ionto that. as long as we have a way forward
2013-09-28 17:27:13 tjtr33 yes, as long as we have a way forward
2013-09-28 17:27:23 mhaberler I think its rather wheel& command jog integration or not
2013-09-28 17:27:43 skunkworks which ja3 would solve easier?
2013-09-28 17:27:51 cradek jepler: I would hate to see JWP get in the way of merging JA3 but it surely would
2013-09-28 17:28:04 mhaberler rebasing the current code on ja3 will entail some work (I gave it a try) but entirely doable
2013-09-28 17:28:18 mhaberler why do you think so?
2013-09-28 17:28:20 jepler fwiw I have a greater personal interest in ja3 than jwp (a linear delta machine in the basement)
2013-09-28 17:28:52 cradek jepler: I think one would only put JPW on master if he was ignorant of upcoming ja3 - it just makes no sense to me
2013-09-28 17:29:14 tjtr33 jepler i can see that but more than a couple people on email lists have expressed interest in JWP
2013-09-28 17:29:15 archivist I have always rerun from start so no axe to grind here
2013-09-28 17:30:04 cradek frankly I am sure we all would like this functionality
2013-09-28 17:30:06 mariusl we are pushing very hard to get LCNC into industry. some things are very needed like JWP
2013-09-28 17:30:09 mhaberler cradek: what is your feeling for a preview version of ja3
2013-09-28 17:30:15 mhaberler timeframe that is
2013-09-28 17:30:28 cradek but I am mighty uncomfortable voting on "I say someone else should do something complicated"
2013-09-28 17:30:45 cradek mhaberler: what do you mean preview?
2013-09-28 17:30:46 archivist I do see users wanting JWP though
2013-09-28 17:30:49 tjtr33 please, a quick explanation why JA3 and JWP are connected ( not in my synopsis )
2013-09-28 17:31:00 mhaberler meaning: changes to jog modes are stable
2013-09-28 17:31:06 jepler tjtr33: because nontrivial kinematics are more complicated to jog
2013-09-28 17:31:13 tjtr33 k
2013-09-28 17:31:14 jepler there are additional kinds of jogs that do not exist in master branch
2013-09-28 17:31:22 skunkworks But how much of jwp? some are asking for tool length and such - which I don't think is in the current implementation..
2013-09-28 17:31:35 mhaberler no, wont be
2013-09-28 17:31:36 cradek mhaberler: the only upcoming changes I know of (hope for) are world wheel jogs
2013-09-28 17:31:56 mhaberler and that be joint->axis mode?
2013-09-28 17:32:18 mariusl if not now tool length will be the next pressing issue
2013-09-28 17:32:24 archivist industry wants to change an insert so offsets should not need to change
2013-09-28 17:32:24 mhaberler ah, world jog - yes, that I'd like to have before I start
2013-09-28 17:32:59 mhaberler the teleop jog cmds are stable/working?
2013-09-28 17:33:30 jepler fwiw when someone talks about "for industry" it sounds like someone who should hire software developers instead of using this meeting to obligate volunteers to work on a specific part of the software.
2013-09-28 17:33:33 cradek yes I've tested (and fixed) them recently
2013-09-28 17:33:46 cradek I strongly agree with jepler
2013-09-28 17:33:49 mhaberler ah, good, then I will try a rebase of the code
2013-09-28 17:34:10 mhaberler the wheel jogs should be rather similar
2013-09-28 17:34:13 cradek mhaberler: yes continuous and incremental jogs work beautifully in world mode in ja3
2013-09-28 17:34:20 mhaberler very good
2013-09-28 17:34:36 mariusl for industry means all the many guys out there building their own machines to try and make a living - we support them
2013-09-28 17:34:38 cradek I don't know how much work world-wheel is, but I suspect it's not TOO bad
2013-09-28 17:34:45 archivist jepler, I am saying we dont need to worry about offsets for most manufacturing
2013-09-28 17:35:09 skunkworks if micges can find it - I think he already has it done..
2013-09-28 17:35:19 mhaberler no, wheel will be very similar then
2013-09-28 17:35:29 cradek like I said in the linked conversation, over years I have heard that changing offsets is sorely wanted
2013-09-28 17:35:50 cradek for insert changes, tool changes
2013-09-28 17:36:22 cradek also I notice our time is up
2013-09-28 17:36:25 tjtr33 is [making a triv-kins version available ] a bad idea. seems it might be a good testbed
2013-09-28 17:36:27 archivist insert wear has been mentioned but not seen it often
2013-09-28 17:36:54 cradek if we are voting on something I'm not at all sure what it is
2013-09-28 17:37:00 archivist nor me
2013-09-28 17:37:04 skunkworks archivist: maybe part of andys new tool table work?
2013-09-28 17:37:06 mhaberler anyway we need to come to a conclusion here; I would think if it can be made to work on ja3 with world jog then I consider it entirely feasible
2013-09-28 17:37:17 mhaberler thats not going to happen in time
2013-09-28 17:37:25 skunkworks I thing it should be part of ja3 also
2013-09-28 17:37:59 cradek I decline to suggest a time frame for anyone else's work
2013-09-28 17:37:59 skunkworks *ja3 merged to master..
2013-09-28 17:38:22 tjtr33 is the proposal now JWP is part of JA3?
2013-09-28 17:38:35 cradek I will help mhaberler with this as I can
2013-09-28 17:38:38 skunkworks I think a direction is more what people are looking for..
2013-09-28 17:38:46 mhaberler no, it's more about the base I can get at, a ja3 merge into master would help lots
2013-09-28 17:38:53 mhaberler or a rebase
2013-09-28 17:39:07 cradek last I heard, seb has a rebase "mostly" done
2013-09-28 17:39:22 cradek I suspect floods have really paused that work
2013-09-28 17:39:54 skunkworks yes - it has surely effected him personally.
2013-09-28 17:40:54 tjtr33 is there a proposal or a way-ahead now ? im sure time is up
2013-09-28 17:40:54 mhaberler well I am fairly sure given a rebase I should have something within say 2 weeks; leave out 5% probability for a showstopper
2013-09-28 17:41:18 mariusl sounds very good
2013-09-28 17:41:28 mhaberler iow: if ja3 is in 2.6, then I am rather sure it can be made to work in time
2013-09-28 17:41:37 cradek that would be awesome
2013-09-28 17:41:43 tjtr33 yes thx michael, so it sound like a plan
2013-09-28 17:42:31 mhaberler cradek - can you take over the rebase in case Seb
2013-09-28 17:42:35 mhaberler is too busy?
2013-09-28 17:43:09 cradek I will ask him if I can be of any help
2013-09-28 17:43:20 mariusl we will test when you need us
2013-09-28 17:43:23 cradek I absolutely believe you could start working on top of ja3 without waiting for his rebase
2013-09-28 17:43:42 archivist do we need a vote at all as this is all volunteer work
2013-09-28 17:43:44 mhaberler ok, then I'll give that a stab
2013-09-28 17:43:49 mhaberler haha
2013-09-28 17:43:50 cradek after the rebase your changes will still apply just fine
2013-09-28 17:43:54 mhaberler fine
2013-09-28 17:44:44 mariusl awesome stuff mr M
2013-09-28 17:44:50 mhaberler just so I start on the right thing - this is the joints_axes3 branch on g.l.o?
2013-09-28 17:44:55 cradek yes
2013-09-28 17:45:02 mhaberler ok, thanks
2013-09-28 17:45:25 mhaberler we'll I'll do my best
2013-09-28 17:45:43 tjtr33 thanks mhaberler seb cradek et al for a future with JWP
2013-09-28 17:45:57 mhaberler is there any UI which does world mode jog with ja3? Axis?
2013-09-28 17:46:09 mhaberler I saw some patches but didnt try
2013-09-28 17:46:10 cradek yes works in AXIS
2013-09-28 17:46:13 mhaberler ok
2013-09-28 17:46:24 cradek I will help you find a working config etc
2013-09-28 17:46:30 mhaberler great
2013-09-28 17:46:43 cradek I have a rotary-delta config with vismach that works great
2013-09-28 17:47:17 cradek I agree with archivist and feel that a vote is inappropriate/unnecessary
2013-09-28 17:48:22 tjtr33 yes vote not needed, mariusl ? comment
2013-09-28 17:48:30 mhaberler why dont you let me vote on my time, that is utterly undemocratic
2013-09-28 17:48:55 cradek I don't understand
2013-09-28 17:48:59 mhaberler that was a joke
2013-09-28 17:49:02 mariusl thanks a lot guys it sounds great - from the users (industry)
2013-09-28 17:49:05 archivist :)
2013-09-28 17:49:49 cradek archivist: thanks for running it this time!
2013-09-28 17:49:56 archivist !end 201309