#linuxcnc-devel Logs

Apr 16 2018

09:43 AM seb_kuzminsky: skunkworks: TLO_IS_ALONG_W is not used any more
10:08 AM skunkworks: seb_kuzminsky: thanks - so really - it would be just using your W offset in the tool table?
10:08 AM skunkworks: for TLO
10:14 AM seb_kuzminsky: yeah
10:18 AM seb_kuzminsky: looks like TLO_IS_ALONG_W was removed in 2.4.0, in 2010
10:30 AM cradek: yay for removing an ugly hack by making the general case possible
10:30 AM seb_kuzminsky: :-)
10:55 AM rene-dev: seb_kuzminsky more and more pepole are using stmbl now, and it is annoying that they have to use master. Would it be possible to merge the bugfixes in the sserial driver into 2.7?
10:55 AM rene-dev: it should not affect any existing setups
10:56 AM seb_kuzminsky: bugfixes, if they're safe, are welcome in 2.7
10:56 AM seb_kuzminsky: please prepare a PR and I'll look at it
10:56 AM rene-dev: so I just cherrypick the commits in a branch, and make a PR to 2.7?
10:57 AM seb_kuzminsky: yeah
10:57 AM seb_kuzminsky: make a branch based on 2.7 for your pr
10:57 AM seb_kuzminsky: cherry pick the sserial fixes into that branch
10:57 AM seb_kuzminsky: build it, test it, etc
10:57 AM seb_kuzminsky: if possible get other sserial users to test it
10:57 AM seb_kuzminsky: then open the PR
10:58 AM rene-dev: they are only changes in sserial.c, so first of all it only affects pepole with mesa cards and sserial slaves, and the code changes should not do anything with mesa hardware
10:58 AM rene-dev: Im just guessing 2.8 is not anywhere close?
10:59 AM seb_kuzminsky: ok it makes it easier that mesa hardware is un-affected
10:59 AM seb_kuzminsky: i don't know the schedule for 2.8, mozmck volunteered to be Release Manager for that stable branch
11:00 AM seb_kuzminsky: but my impression is that 2.8 (or linuxcnc 3 or whatever it will be called, "2.7+1") is in pretty good shape
11:00 AM rene-dev: that depends if I pick the float only, or the params as well. parameters also affect other sserial slaves, but that has been tested by many pepole.
11:01 AM seb_kuzminsky: it's good that it's widely tested :-)
11:01 AM rene-dev: I seem to have this issue: https://github.com/LinuxCNC/linuxcnc/issues/300
11:01 AM rene-dev: but what fixed it for him didnt fix it for me
11:01 AM rene-dev: I need to try to reprodude that in the simulator
11:02 AM rene-dev: Its on a machine where I use A axis for the toolchanger, and I always get this error on the first toolchange
11:02 AM rene-dev: then I need to jog A, after that its ok
11:03 AM rene-dev: I do a g53 g0 z0 in the m6 remap, and that fails with Linear move on line X would exceed A's positive limit
11:09 AM rene-dev: and I think there should be some default way for trivkins machines, so you dont need to specify joint and axis limits