Back
[09:48:42] <memleak> seb_kuzminsky, what RTAI are you using?
[09:49:15] <seb_kuzminsky> memleak: the 3.4.55 one we worked on way back when
[09:50:03] <memleak> ok
[09:52:13] <memleak> shabbyx has been refactoring all of my RTAI work and cleaning it all up for me as i havent had the time and energy to do it myself, in turn i havent tested any of the new stuff on the current master branch on the RTAI github repo
[09:53:06] <memleak> the master that you know of is now old-master and maybe some time I'll get around to testing and working on the new master branch
[09:53:07] <seb_kuzminsky> ok, thanks for that information
[09:53:49] <seb_kuzminsky> i want to get the old-master version out for wider testing, then i'll talk with you and shabby about newer versions
[09:54:14] <seb_kuzminsky> it sounds like paolo is making progress on his branch too, do you know anything about that?
[09:54:24] <memleak> i have no idea what is going on with the RTAI cvs tree but I'm not going to be touching the new master branch until RTAI 4.0 is stable and tarred up
[09:54:53] <memleak> (when its not a test release anymore)
[09:55:10] <seb_kuzminsky> yeah, i hear you
[09:55:28] <seb_kuzminsky> i too am pretty excited about the idea of a real rtai release
[09:56:08] <memleak> once 4.0 is really out I'll take the new master branch on github probably and make our own tarball stable release on there too
[09:56:54] <seb_kuzminsky> is there a lot in your branch that's not in the testing prerelease of 4.0?
[09:58:04] <memleak> it's probably eventually all going to be an RTAI patch set sort of like crosstool-ng or something that is only based off stable releases with maybe some CVS fixes merged in
[09:58:53] <memleak> instead of constantly running cvs update and merging the entire tree in every few days, just work off a stable release instead and only pull in what has been tested and works.
[09:59:18] <seb_kuzminsky> ok
[09:59:27] <memleak> having two completely different trees of basically the same thing when it could just be a patch set is a lot more work
[09:59:43] <memleak> because then shabby and I need to follow literally every single change in the CVS repo
[10:00:20] <seb_kuzminsky> so you'll have a 'shabby-memleak' branch based on the most recent official rtai release
[10:00:28] <memleak> correct.
[10:00:30] <seb_kuzminsky> and when a new release comes out you'd rebase your work on top of that
[10:00:35] <memleak> yes.
[10:00:54] <seb_kuzminsky> and any fixes that you think belong upstream you'll mail to paolo for inclusion in the mainline?
[10:01:09] <seb_kuzminsky> oop, my bus stop is coming up, bbl
[10:01:11] <memleak> with intermediate changes along the way such as critical fixes and such that Paolo mocked up in CVS that have not yet been tarred
[10:01:18] <seb_kuzminsky> thanks for your work on rtai, it helps linuxcnc a lot!
[10:01:51] <memleak> Paolo takes my stuff for whenever he wants it or needs it, not by when I ask him to, heh
[10:02:05] <memleak> he completely ignores me pretty much..
[10:02:17] <memleak> just goes after my code (which i don't really mind)
[14:31:03] <skunkworks> is there an up/down counter in hal?
[14:36:47] <seb_kuzminsky> skunkworks: updown
[14:37:26] <Tom_itx> how fitting
[14:38:24] <skunkworks> seriously? I swear I looked - I fail at the alphabet.
[14:38:33] <Tom_itx> how's that different from a single channel encoder?
[14:39:35] <skunkworks> single channel encoder doesn't have a way to count down
[14:39:40] <Tom_itx> looks like it's got a few more options
[14:39:49] <Tom_itx> quad then
[14:40:56] <Tom_itx> are you counting widgets?
[14:41:47] <skunkworks> I was just thinking of a way to feed step/dir into linuxcnc and have it calculate velocity and acelleration
[14:42:32] <seb_kuzminsky> might make sense to add step/dir as an input type to our encoder component. maybe up/down too
[14:42:43] <skunkworks> sure
[14:42:50] <seb_kuzminsky> our stepgen module has these different output modes... it's be symmetric somehow
[14:43:05] <skunkworks> ocd?
[18:18:46] <alex_joni> jepler: around?
[18:19:03] <alex_joni> you talked about some bbb changes to make AXIS work faster over ssh
[18:21:59] <alex_joni> I applied those, but I can't remember what they were