cradek: I wonder if motion should do something defensive to turn certain buggy index behavior into an abort
for instance, if in the cycle after an index reset the fabs(spindle-position) > 1 rev, abort
or whether failure to follow the thread could be turned into a following error
I agree that might be nice (either one)
(perhaps you've entertained these ideas before; if so, just ignore me)
I wish I knew the least-bad thing to do when something's going wrong during synced motion
on the surface, "the same as ferror" makes intuitive sense to me
with g76 you could theoretically know where a safe X is
I'm a bit afraid to touch it at this stage
man am I shot. I replaced all the fuel lines on the bus today.
now it will be roadworthy after a tire repair
wow, that butterfly is quite a nice close-up shot
they're cool critters
BigJohnT_ is now known as BigJohnT
seems I broke building sim packages and the buildbot didn't call me on it
EMC: 03jepler 07TRUNK * 10emc2/src/Makefile: fix building of sim packages: don't install firmwares, and arrange for the emc2-sim-dev.files file list to exist
EMC: 03jepler 07TRUNK * 10emc2/debian/configure: fix building of sim packages: don't install firmwares, and arrange for the emc2-sim-dev.files file list to exist
jepler: any plans for today?
alex_joni: trying to figure out if the deb build process is right yet
also I have been setting up a vmware 32-bit hardy in case I want to do the whole build process myself
yes, it'd be nice to get beta2 out today
what can I do?
umm, make a change at the last minute without considering the consequences?
too bad none of us has a machine that stepconf is appropriate for .. "test stepconf" would be nice
stupid vmware. it frequenly removes all the modifiers from xmodmap when changing focus to/from vmware (I think that's it, anyway; it never seems to happen when I'm not running the vmware viewer, but it's happened 3 times today
yeah stepconf is very important, and pretty complicated, and none of us use it at all
it's pretty hard to test pointy-clicky stuff isn't it? how do you simulate points and clicks?
when I "test" stepconf, I make a small range of configurations (maybe just one) and look at it to see whether it looks plausible
the user interaction stuff is mostly OK, it's the generated hal files that I worry about
(and that *could* be tested, in terms of input / expected output / actual output -- there's a mode in stepconf that just regenerates the config files ... but I haven't taken advantage of that)
EMC: 03jepler 07TRUNK * 10emc2/VERSION: bump version for release. note changes.
EMC: 03jepler 07TRUNK * 10emc2/debian/changelog: bump version for release. note changes.
huh -- what is wrong with this command? cvs -d:ext:firstname.lastname@example.org/cvs co -p -rRELEASE_2_3_0_BETA2 emc2/VERSION
it works if I change BETA2 to BETA1
I remember having this problem back when I tagged BETA1, but then it magically resolved itself
[19:01:00] <jepler> http://pastebin.ca/1368277
it seems to work doing a full checkout, but I haven't reied it with -p or just getting VERSION
oh, and I'm doing it as me, not anon
anon makes no difference
most commands worked
but that specific "co -p -r" fails
it's working now
(for me, using the anon line as you have it)
so it is
wfm now too
wow, there's more stuff than I expected in the new changelog
good job, everyone
EMC: 03jepler 07TRUNK * 10emc2/VERSION: TRUNK is open for new features again
EMC: 03jepler 07TRUNK * 10emc2/debian/changelog: TRUNK is open for new features again
jepler has changed the topic to: EMC2 development -- http://linuxcnc.org/
| 2.3 schedule: beta2 on package server soon, development open on TRUNK | channel logged by logger_dev
EMC: 03jepler 07v2_3_branch * 10emc2/VERSION: bump version number after release
EMC: 03jepler 07v2_3_branch * 10emc2/debian/changelog: bump version number after release
I'm sure you've got a thing or two to put on TRUNK by now
it's interesting that in the graphical view of revisions of VERSION, each branch's topmost point is about on the same level as the others'
in the 2.0, 2.1 and 2.2 branches, that is
jepler: I'm checking 2.3.0-beta2 out now
my upload will probably be a bit faster
(although the download is probably way slower ;)
jepler: 17 revisions on 2.1 and 2.2
a couple more on 2.0 because we did the branching earlier
alex_joni: I've got hardy 32- and 64-bit built
no need for you to build anything
but.. but I want to :P
did you build all variants? (sim too?)
did you manage to do anything with how stepconf looks?
I tried to, but I failed
you mean pickconfig?
I don't know what you mean, then
since it grew in vertical size
the logo on the left looks pretty bad
with the brown extension top left
I am surprised I hadn't noticed that myself
maybe you just clicked away at it :)
put it on a bug list somewhere
all I managed to do with it is add an icon to the window (top right)
which caused the title to be a bit bigger, which made it slightly nicer
the other idea is to resize the picture
there's no guarantee any size will be taller than the tallest content
jepler: maybe you can find a way to resize it then, I couldn't
you mean, to make gtk scale the image? no, I doubt you can do that either
EMC: 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/stepconf.glade: use website colors on stepconf
EMC: 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/Submakefile: this makes the source tree's emc2-wizard take precedence over a run-intsalled emc2-wizard
EMC: 03jepler 07TRUNK * 10emc2/emc2-wizard.gif: a new day needs a new wizard
jepler: I take it the bugreport would be obsolete now?
was struggling with my network connection for a while
alex_joni: might still want to do something else on the branch, I dunno
* alex_joni checks TRUNK
EMC: 03jepler 07TRUNK * 10emc2/ (emc2.xcf emc2.gif): and a new splash screen
you broke it
what I broke?
buildbot is spewing fail messages
make: Entering directory `/home/juve/emc2.TRUNK/src'
../docs/src/Submakefile:392: *** Required image file ../share/emc/emc2-wizard.gif does not exist. Stop.
might be something I have in my tree.. am investigating now
no, I get it on a fresh tree too
I didn't have docs enabled on the system where I checked that in
odd, the farm failed too, but no emails and no CIA (yet)
seems like the docs *.gif rule catches what it shouldn't
EMC: 03jepler 07TRUNK * 10emc2/docs/src/Submakefile: unfortunately, this is now a false-positive error (why?)
EMC: 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/Submakefile: oops, need to actually copy the file!
eww.. pickconfig looks pretty crappy now, imo
it looks the same as ever for me
did you up & compile?
are you talking about trunk?
not so crazy about the splash either :)
stepconf is ok though
I kind of like them both - pickconfig has a lot of whitespace though.
I like the splash
EMC: 03tissf 07TRUNK * 10emc2/src/po/fr_axis.po: french translation update
well, if I've fixed the build error I'm going to make tracks
please continue arguing over the art in my absence
thanks for doing the release
I automatically defer on all art decisions
jepler: I notice you uploaded hardy packages
did you also rebuild the repo? or should I?
alex_joni: I am doing that now
had to change my script a bit, -all wasn't listed
alex_joni: you need to make these directories group-writable:
* alex_joni looks
rsync: send_files failed to open "/home/emcboard/www.linuxcnc.org/hardy/dists/hardy/emc2.3/binary-amd64/.emc2_2.3.0~beta2_amd64.deb.OW42WG": Permission denied (13)
they should be +w now
I removed that other file, I think it's an rsync dropping
I'm not sure if my rsync upload for the group write thing didn't overwrite some Release files
had you already pushed the rebuild?
EMC: 03flo-h 07v2_3_branch * 10emc2/src/po/de_axis.po: updated german translation
I ran it again just now to be sure
now it runs with no errors for me
looks ok here too
hmm.. I'm not getting the update
EMC: 03alex_joni 07TRUNK * 10emc2/emc2-wizard-pickconfig.gif: like this one better for pickconfig
EMC: 03alex_joni 07TRUNK * 10emc2/tcl/bin/pickconfig.tcl: use separate logo for pickconfig
last bug found for today: http://www.pastebin.ca/1368491
I think we should rather point out the config needs python-serial than have it as a dependency
err.. it's already pointed out in the README (pyserial)
alex_joni: I agree; that doesn't look like a bug to me
though python-serial is tiny
hmmm. I wonder what I'm doing wrong on this Jaunty build
I can run the latency test, but for some reason, the EMC modules (rtapi.ko ...) won't load
the RTAI latency test that is, latency-test doesn't work
if you were a newbie I'd guide you in copying the output of dmesg to pastebin at this point
I'll look at it later. I think we're about to watch a movie
it's something to do with module versioning I think