#emc-devel | Logs for 2007-01-29

Back
[01:28:21] <cradek> must be too late in the weekend for anyone to care...
[01:37:33] <jmkasunich> ?
[01:37:48] <cradek> heh
[01:37:57] <cradek> just no responses or anything yet
[01:37:59] <jmkasunich> oh
[01:38:10] <jmkasunich> I'm about to start the wiki stuff
[01:38:15] <cradek> cool
[01:38:19] <cradek> I'm working on the cd
[01:38:45] <jmkasunich> I think I have the needed access to change the install scripts
[01:39:01] <cradek> if you don't, I might
[01:40:41] <jmkasunich> good, because I don't
[01:40:44] <cradek> heh
[01:58:37] <cradek> cd is built - testing an install now
[01:58:55] <jmkasunich> cool
[02:05:49] <cradek> haha figured out why I couldn't load any web pages -- rhinoceros asleep on the escape key
[02:06:00] <jmkasunich> lol
[02:07:34] <jmkasunich> so has that become his(her) official name now?
[02:07:46] <cradek> nah but it's sure descriptive
[02:08:20] <cradek> she's very sweet but she does disrupt a lot of my work
[02:09:47] <cradek> hmm maybe I do need some more ram
[02:10:03] <cradek> I guess I don't do this kind of thing very often
[02:14:37] <jmkasunich> those instructions for updating that you had in a pastebin - where are they?
[02:15:02] <cradek> just in that email so far
[02:15:13] <jmkasunich> duh
[03:04:09] <fjungclaus_> fjungclaus_ is now known as fjungclaus
[03:15:27] <jepler> halcmd: linkpp => =>
[03:15:27] <jepler> Segmentation fault
[03:15:36] <jepler> oops, I'd better not check this halcmd refactor in just yet!
[03:15:53] <jmkasunich> heh
[03:16:39] <jmkasunich> I wonder - should arrow stripping be a function of the parser, so do_linkpp_cmd() and do_net_cmd() and such don't duplicate it?
[03:17:04] <jmkasunich> or do you think we might want the arrows to actually mean something someday
[03:17:29] <jepler> I did make arrow stripping a function of the parser
[03:17:40] <jepler> a command can say that it wants arrows stripped
[03:17:48] <jmkasunich> great minds, and all that crap... ;-)
[03:17:58] <jepler> e.g., {"linkpp", FUNCT(do_linkpp_cmd), A_TWO | A_REMOVE_ARROWS },
[03:21:36] <jepler> so that I didn't have to change the command functions much, the dispatch function supports a variety of calling conventions: 0, 1 or 2 arguments, with or without additional arguments in an argv[] array
[03:22:02] <jmkasunich> works for me
[03:22:20] <jmkasunich> I guess the alternative would have been argv style for everything
[03:22:28] <jepler> I thought about that too
[03:23:21] <jmkasunich> each cmd could either get argc and test it (I need to args dammit), or tell the parser how many it expects and let it do the testing
[03:24:31] <jepler> In the C API for python and tcl, there is basically one preferred signature for all functions, and it's up to the function to check the correctness of the arguments
[03:24:56] <jepler> (I should clarify and, C APIs -- they each have their own)
[03:25:10] <jmkasunich> so would argc/argv be more generic
[03:25:36] <jepler> yeah -- in fact, argc/argv is what tcl uses
[03:25:35] <jmkasunich> let each funct test ( argc != my_required ) or ( argc < my_min ) ?
[03:26:10] <jepler> yes
[03:26:37] <jmkasunich> there are a few that expect the parser to do some screwing around - linkps and linksp have the same handler, and the parser swaps around the args
[03:26:54] <jmkasunich> although you could use two handlers instead, one would swap args and call the other
[03:27:02] <jepler> yes that's what I did
[03:27:08] <jmkasunich> ;-)
[03:27:07] <jepler> the only special cases now are "quit" and "="
[03:27:53] <jepler> cradek: didn't build -sim packages?
[03:29:44] <cradek> no, maybe I should
[03:29:50] <cradek> I didn't even think about it
[03:29:54] <cradek> 20:20:26 < jmkasunic> same filename as before, just a different sum?
[03:30:06] <cradek> uhh
[03:31:04] <cradek> on a fresh install, apt-get build-dep emc2, apt-get -b source emc2, working package
[03:32:42] <jepler> cradek: neat
[03:32:44] <jepler> cradek: great
[03:37:04] <cradek> cd ..
[03:37:10] <cradek> crap
[03:39:25] <jmkasunich> bash: cd: crap: No such file or directory
[03:40:52] <cradek> irc on my saturated line and vmware are both so unresponsive that it's a guess which one I'm typing into
[03:41:31] <jepler> yuck, sounds terrible
[03:41:48] <jepler> it's a matter of RAM that vmware is unresponsive?
[03:41:54] <jepler> that can be bought
[03:42:08] <cradek> yeah, but I don't use it much
[10:27:58] <alex_jon1> alex_jon1 is now known as alex_joni
[15:53:05] <jepler> I've been thinking that "1-line handshaking" might be useful for streamer. When .data-transfer is TRUE, streamer sets it to FALSE and outputs an item from the FIFO. The other component sets .data-transfer back to TRUE when the next item is desired
[18:04:43] <alex_joni> Assertion failed: (prev != NULL), function add, file /usr/src/gnu/usr.bin/cvs/cvs/../../../../contrib/cvs/src/add.c, line 472.
[18:05:28] <alex_joni> cradek: I broke it :(
[19:07:21] <cradek> uhhh what did you do?
[19:07:49] <alex_joni> I cvs add'ed something
[19:08:16] <alex_joni> guess it wasn't good I was lazy.. and added a couple of folders and files at once
[19:08:51] <cradek> hmmmm
[19:09:54] <cradek> is it still broekn? do I have to do something?
[19:16:23] <alex_joni> I don't think so
[19:16:26] <alex_joni> I managed to add it again
[19:23:19] <jepler> comments, please. http://emergent.unpy.net/files/sandbox/tristate.9.txt
[19:24:52] <cradek> what's the application?
[19:25:51] <jepler> like the 'conv' components, it's to bridge between incompatible pins
[19:26:46] <cradek> I only know of one use of io pins (index) and this will break it, since it's one-way
[19:27:16] <cradek> or I am missing something (likely)
[19:27:46] <jepler> I don't have a particular place in mind
[19:27:59] <jepler> I thought that the old motion controller spindle synch was OUT only
[19:28:20] <cradek> yes but it was bogus, and fixed now
[19:29:48] <jepler> I guess this is a solution in search of a problem, then
[19:30:43] <cradek> I'd have to understand the use of IO pins better than I do to say whether I agree with that
[21:54:09] <alex_joni> haha.. jmk just cracked a pumpkin
[21:59:18] <alex_joni> anyone heard any result from 2.1.0 ?
[21:59:26] <jepler> not me
[21:59:37] <jepler> I guess no one cares
[21:59:40] <alex_joni> same here..
[23:53:10] <cradek> I added emc2-sim packages for breezy and dapper to the emc2.1 repositories
[23:53:25] <cradek> (for non-realtime systems)