#emc-devel | Logs for 2006-11-13

[00:00:32] <jmkasunich> converters went from 56 to 30
[00:00:33] <SWPLinux> you know, a mux2 is a good conv_bit_float ...
[00:00:48] <jmkasunich> yep - and a lot more flexible
[00:00:50] <SWPLinux> and you even get to choose what thefloat values should be
[00:00:55] <SWPLinux> right
[00:01:16] <jmkasunich> and a comparator is the right way (IMO) to do float to bit
[00:01:25] <SWPLinux> yep
[00:01:56] <jepler> we should kill the ones that aren't that useful sooner instead of later
[00:02:02] <SWPLinux> probably
[00:02:03] <jepler> so by all means kill the bit<->float converters
[00:02:10] <jmkasunich> consider them dead
[00:02:49] <SWPLinux> did you get rid of hal_[su]8 as well?
[00:02:55] <jmkasunich> not yet
[00:02:58] <SWPLinux> ok
[00:03:05] <jmkasunich> there are a few more places where it is used
[00:05:21] <SWPLinux> I think the number of converters goes down to 4 without the 8-bit types
[00:05:30] <SWPLinux> possibly 6 if you also have signed/unsigned conversion
[00:22:05] <dlabriola> I've got to bug out! have a good evening!
[01:59:33] <cradek> jmkasunich: whatever happened with your investigation of Interp::init, M2, startup gcodes, etc?
[02:00:23] <jmkasunich> I forget what we were trying to do
[02:00:26] <jmkasunich> wasn't that a week ago?
[02:00:34] <cradek> yeah
[02:00:36] <jmkasunich> thats forever in dog years
[02:00:49] <cradek> I must have been busy with something else because I barely remember what was going on
[02:00:54] <jmkasunich> same here
[02:00:58] <jmkasunich> oh, I remember
[02:01:20] <jmkasunich> it was the new codes to turn on/off feedoverride, spindleoverride, feedhold, and adaptive feed
[02:01:26] <cradek> ok right
[02:01:48] <jmkasunich> they were being set to their initial state in the wrong place, which let the interp and the controller get out of sync
[02:02:12] <jmkasunich> we fixed it all - the initial values now come from the controller, and the interp can get or set them with canonicals
[02:02:23] <cradek> great
[02:02:47] <jmkasunich> at least I think its all fixed... that was strange territory for me to be poking around in
[02:02:50] <cradek> I'm still worried about M2/G92.3 which is one of the tasks
[02:03:02] <cradek> yeah same for all of us I think
[02:03:27] <cradek> putting g92.3 in the startup gcodes doesn't fix it because it doesn't get run before every program
[02:03:56] <jmkasunich> we have startup = "starting the program" codes
[02:04:13] <jmkasunich> s/program/emc/
[02:04:17] <cradek> that's not what my testing showed a few weeks ago - did you change it?
[02:04:22] <cradek> right, starting emc
[02:04:27] <jmkasunich> should we have additional codes for starting the g-code program?
[02:04:33] <cradek> I don't know
[02:05:03] <cradek> part of me wants to say that's dumb - put them in the program if you want them to run at the beginning of the program
[02:05:19] <jmkasunich> we looked into how the startup codes are implemneted a bit while working on the other issue, but wound up with another solution that was more correct - so we didn't change anything in that area
[02:05:32] <cradek> I understand
[02:05:51] <cradek> running them at every program run seems problematic too. you can't change the interp state in MDI and have it stick then.
[02:06:28] <jmkasunich> well, if you want to do that, don't put any "program-startup-codes" in the ini
[02:06:33] <cradek> the whole thing driving this problem is tkemc's use of G92. maybe I should ask ray how it's supposed to work and what he wants to do about it
[02:06:48] <jmkasunich> sounds good to me
[02:06:56] <cradek> ... make ray and jon E sort it out, and then we can fix it
[02:07:07] <jmkasunich> I have absolutely no clue about G92, and probalby never will
[02:07:10] <cradek> adding configurability is NOT a good way to fix a simple bug
[02:07:23] <jmkasunich> or at least, not until I do my conversion and actually find myself using EMC to make parts
[02:07:31] <jmkasunich> true
[02:07:37] <cradek> well I understand it completely from reading the spec.
[02:07:50] <cradek> emc *does* what the spec says
[02:08:03] <cradek> but tkemc seems to ?? expect something else to happen
[02:08:20] <jmkasunich> I guess there may be times when adding configurability is the way to fix a bug: when a large chunk of the user community doesn't consider it one, and another chunk does
[02:08:25] <cradek> and that bothers me (and jon E)
[02:09:03] <cradek> yes we may decide to do that, but we probably shouldn't do it first
[02:09:09] <jmkasunich> agreed
[02:12:45] <A-L-P-H-A> A-L-P-H-A is now known as a-l-p-h-a
[03:23:16] <jepler> has anybody written some text to explain why we're not making packages for edgy? Something that could go in the wiki, for instance.
[03:25:09] <cradek> I think I said something in an email, but that's it
[03:25:43] <cradek> btw: my update manager does NOT show an edgy upgrade thingy
[03:26:44] <jepler> yeah that was operator error
[03:26:51] <cradek> oh good
[03:26:55] <jepler> somehow I was running a "dapper" that was 6.04, not 6.06
[03:27:08] <cradek> that was a curse in breezy
[03:34:00] <jepler> I'm going to add something like this to the wiki "Installing EMC2" page:
[03:34:00] <jepler> === # Why aren't there packages for Ubuntu 6.10 "Edgy Eft"? ===
[03:34:00] <jepler> We have chosen to use 6.06 because it's a "long term support" (LTS) release. If you use it, you can be sure of support and security fixes from the Ubuntu team for 3-5 years. However, we would be happy to accept and host packages for 6.10 if a user/developer of emc2 wishes to contribute them.
[03:34:33] <cradek> sounds fine to me
[03:36:09] <jmkasunich> I like it
[03:36:50] <jmkasunich> I think I'm going to do something radical and go to bed before jepler does
[03:36:56] <jepler> see you jmkasunich
[03:37:04] <jmkasunich> goodnight
[03:37:54] <cradek> bye
[21:09:33] <cradek> the best way to ask for help on a mailing list is to say "I have limited patience for this"?
[21:10:20] <jepler> I dunno, but I just saw this in someone's signature:
[21:10:20] <jepler> Q: Gentoo is too hard to install and I feel like whining.
[21:10:20] <jepler> A: Please see /dev/null. (from the gentoo installer FAQ)
[21:11:10] <cradek> heh
[21:53:32] <skunkworks> sounds like it booted up without the monitor hooked up. defaulted to 640X480
[21:53:40] <skunkworks> that would be my thought
[21:54:40] <jepler> sure could be
[21:54:41] <SWPadnos> I thought about mentioning <CTL-ALT-+>, but didn't bother
[21:54:51] <jepler> I doubt cradek changed the monitor detection routines for the live cd
[21:55:29] <skunkworks> what? thats the first thing he does... :)
[23:48:22] <owhite> anybody home?