#emc-devel | Logs for 2008-05-23

[01:07:54] <rayh> anyone else seeing ugly stuff in the ladder rung display?
[01:10:54] <SWPadnos> I always see the CL display as ugly, but that's just me
[01:12:49] <rayh> http://imagebin.ca/view/MLoTFV.html
[01:13:14] <rayh> The input names run in the symbols
[01:13:37] <SWPadnos> I wonder if there's some word wrapping thing with changed defaults
[01:13:51] <SWPadnos> is that on 8.04?
[01:16:38] <rayh> Yes. 8.04 and no I don't see any config but will look in the real file.
[01:17:03] <SWPadnos> I'm thinking in the text widget or something like that, not something specific in the CL source
[01:20:28] <SWPadnos> there was a recent change to the way text is drawn - adding color to it
[01:21:13] <SWPadnos> but it seems to use all the same parameters otherwise
[01:21:22] <rayh> Let me try it from cvs today.
[01:21:30] <SWPadnos> tbhis change is from April 12
[01:21:33] <SWPadnos> -b
[01:23:38] <SWPadnos> hmmm. do you have red and blue name text in any CL window?
[01:24:26] <SWPadnos> I love this one: http://imagebin.ca/view/gtgzRHfB.html
[01:26:00] <rayh> The current cvs is different from the post earlier. It does not display the names at all.
[01:26:14] <SWPadnos> hmmm. that is odd
[01:26:29] <SWPadnos> you sure that isn't a v2_2_branch checkout?
[01:26:39] <rayh> Okay.
[01:26:54] <rayh> looking
[01:27:07] <SWPadnos> CVS/Entries, I think
[01:28:28] <rayh> Looks like the tag is only the date.
[01:28:47] <rayh> rational eh?
[01:28:51] <SWPadnos> heh
[01:29:28] <SWPadnos> that may mean that cvs up doesn't change anything, since there have been no changes before that date after the checkout (if that made any sense)
[01:29:59] <SWPadnos> you may need cvs up -dPC to clear the sticky tag (I think it's C - not positive though)
[01:30:04] <SWPadnos> could be A, not C
[01:30:29] <SWPadnos> oh, it is A
[01:33:09] <rayh_> okay no tag
[01:33:13] <rayh_> thanks
[01:33:23] <SWPadnos> sure
[01:33:34] <rayh_> Nothing changed in cl or a relative.
[01:33:52] <SWPadnos> I guess I could have asked the date :)
[01:35:43] <rayh_> yesterday
[01:36:05] <rayh_> I did see the change to directory.map go by.
[01:36:11] <SWPadnos> ok
[01:36:13] <rayh_> so I've got the current thing now.
[01:36:22] <rayh_> makling it.
[01:36:30] <rayh_> making even
[01:36:39] <rayh_> mangling it now
[01:36:45] <SWPadnos> I didn't look at exactly where the color text was added - I was looking at the diff in the commit mesasge
[01:37:11] <SWPadnos> bu tit looks like it's supposed to color labels red or blue depending on whether they're input or output
[01:37:16] <SWPadnos> but it
[01:38:14] <rayh_> I did see that with the %I0 and %O0
[01:38:20] <SWPadnos> ah, ok
[01:38:34] <SWPadnos> they're all black in the screenshot
[01:38:37] <rayh_> and the spacing looked okay but then it always did with the % tags
[01:39:49] <rayh_> One more hardy install on the ITX mobo. Was playing with several of the lite desktops.
[01:40:38] <rayh_> That was from a 2.2.5 run
[01:40:57] <rayh_> Installed package rather than compiled.
[01:41:14] <SWPadnos> ah, ok. the color change is on trunk
[01:47:25] <rayh_> Same thing no names only %Ix
[01:47:51] <rayh_> I'll try changing the names of the signals from xxx-xxxx to remove the dash.
[01:47:56] <SWPadnos> hmmm. the latest checkout should show names, I'd think
[01:48:10] <SWPadnos> in colors even
[01:48:12] <rayh_> and see if that fixes the 2.2.5
[01:49:43] <rayh_> * rayh_ gets a push broom to clean the desks
[01:50:18] <SWPadnos> I could use a flamethrower, but I think there's tax info under here somewhere
[01:50:28] <SWPadnos> (plus I'd have to damage the monitors :) )
[01:50:30] <SWPadnos> er - hate
[01:51:47] <cradek> there's a checkbox thing that turns the signal names on an doff
[01:51:49] <cradek> and off
[01:52:12] <cradek> I forget what it says, but it's there because you still have to use the % names to do the editing (I think)
[01:54:01] <rayh> I see that but it made no difference.
[01:55:34] <cradek> I have 2.2.2 here. I ran demo_sim_cl and I get signal names as long as the "Display symbols" check is on
[01:55:46] <rayh> I see that display button works on 2.2.5
[01:56:00] <cradek> oh it's broken on trunk?
[01:56:17] <rayh> I get the display names mashed all up with the contacts and such.
[01:58:28] <cradek> http://timeguy.com/cradek-files/emc/ladder.png
[01:58:42] <cradek> I had to enlarge the window a bit to get this nice display
[02:04:37] <rayh> Looks nice.
[02:05:08] <cradek> I agree it is kind of ugly at the default window size
[02:05:49] <rayh> I went to full screen and user-request-enable is still all over the place as are a couple others.
[02:06:26] <cradek> I'm pretty clueless. maybe cmorley can help.
[02:07:25] <rayh> I'll do a bit of playing and get with him.
[02:09:16] <Roguish> hey, CL is on a pretty good road. lotssss better than just not too long ago. i like it. ugly and all. cudos to Morley
[02:13:43] <rayh> speaking of ugly what's this basic-ugly.dinero.dreamhost.com
[02:14:43] <jepler> that's the name dreamhost chose for the machine which hosts www.linuxcnc.org. read whatever you like into it
[02:15:10] <rayh> I was just pinging from a new box and saw it.
[02:31:55] <jmkasunich> aram thinks that 17,500 in harmonic gear reducers is a good idea
[02:32:02] <jmkasunich> $17,500 that is
[02:32:28] <jmkasunich> MMTB
[02:39:41] <rayh> That would be one way to make it.
[02:40:02] <rayh> Evening jmkasunich
[02:42:59] <jmkasunich> hi rayh
[02:44:03] <rayh> Get the remodel done?
[02:44:08] <jmkasunich> if he gears it down enough, he won't need servos anymore
[02:44:17] <jmkasunich> the wall is done, the bench is installed
[02:44:35] <rayh> Great.
[02:44:36] <jmkasunich> one section of shelves is made and hung, the other is about half done
[02:44:58] <rayh> That is a big improvement.
[02:44:58] <jmkasunich> I'm enjoying the woodworking - I forgot how satisfying it is to use a plane
[02:46:45] <rayh> gotta run.
[03:22:11] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[13:00:32] <rayh> Any clue why the Axis display has a dead spot in the middle. http://imagebin.ca/view/1wOQSL0i.html
[13:02:30] <rayh> This is last night's install on MiniITX 1gig via processor, 256 Meg RAM, onboard Via Unichrome.
[13:04:47] <jepler> huh that's interesting.
[13:05:33] <jepler> how does it behave after you rotate, pan or zoom the preview? what if you resize the window?
[13:06:53] <alex_joni> rayh: can you confirm that you have desktop effects switched off?
[13:08:05] <jepler> oops, bbl
[13:14:42] <rayh> I'll look.
[13:16:43] <rayh> It says no visual effects in the preferences
[13:18:04] <rayh> The image seems to slide under the blanked out area.
[13:21:26] <rayh> This is dual boot with an older 2.1.7 and that seems to work fine.
[13:39:41] <rayh> It appears to be the box the position display is in. My other installs allow you to push stuff up to the right of that box. This install does not allow that no matter how large I make the graphical part of the window.
[13:51:01] <jepler> that area is also supposed to be alpha-blended, so you should see a darkened version of the program when it's under that rectangle..
[13:52:06] <fenn> the rectangle isnt supposed to be that big? only just around the numbers
[13:52:15] <jepler> that too
[13:53:27] <jepler> are you using the via driver? if so, try vesa. there's a note in the via(4) manpage that there is "direct rendering 3d [] using experimental dirvers", so maybe axis is triggering a bug in that experimental code..
[13:56:10] <rayh> Okay.
[13:56:40] <rayh> I don't get the blending on either install here.
[13:59:26] <rayh> bbiab
[14:01:46] <cradek> jepler: we could put hardy on install2.
[14:03:45] <jepler> cradek: I'd need to look and make sure I didn't have any important uncommitted code on there
[14:05:11] <cradek> I'm looking too. I'm just saying we can do it if we want. I even have a cd.
[14:21:49] <CIA-32> EMC: 03jepler 07TRUNK * 10emc2/src/hal/drivers/hal_ax5214h.c:
[14:21:49] <CIA-32> EMC: from the better-late-than-never files: Dave Brink's changes to fix the ax5214 driver.
[14:21:49] <CIA-32> EMC: The control registers were improperly programmed and the outputs were all
[14:21:49] <CIA-32> EMC: incorrectly activated for a short time during initialization.
[14:24:43] <CIA-32> EMC: 03jepler 07v2_2_branch * 10emc2/debian/changelog: from TRUNK: dave brink's fix for hal_ax5214h
[14:24:48] <CIA-32> EMC: 03jepler 07v2_2_branch * 10emc2/src/hal/drivers/hal_ax5214h.c: from TRUNK: dave brink's fix for hal_ax5214h
[17:45:18] <rayh> How do I find and change the display driver in hardy? I'm lost.
[17:50:40] <skunkworks_> isn't it still /etc/X11/xorg.conf
[17:51:15] <rayh> That file doesn't show anything about the display driver other than configured driver.
[17:51:22] <SWPadnos> ;it's just that everything is automatic now, so there's nothing to "change"
[17:51:30] <SWPadnos> you have to add it
[17:52:01] <rayh> add it to xorg.conf
[17:52:08] <SWPadnos> yes
[17:52:13] <rayh> Okay
[17:52:36] <SWPadnos> I don't know the specifics either, and I was similarly dismayed when I saw nothing useful in xorg.conf
[18:17:05] <rayh> k here goes vesa
[18:19:08] <rayh> Yep that does the trick. Thanks guys.
[18:20:53] <SWPadnos> kewl
[18:21:57] <rayh> I added the line Driver "vesa" to xorg.conf in the Device section under configured video device
[18:22:51] <rayh> I suppose that it would be good to put that info in the wiki in relation to hardy.
[18:22:54] <SWPadnos> heh
[18:44:45] <CIA-32> EMC: 03swpadnos 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/stepconf.py: Fix incorrect signal name for spindle index input
[18:47:54] <skunkworks_> SWPadnos: is that it?
[18:47:58] <skunkworks_> is he threading?
[18:48:14] <SWPadnos> I still need to give him a workaround for custom.hal
[18:52:03] <rayh> Added a section to wiki in troubleshooting on the via display driver thing.
[18:52:44] <alex_joni> rayh: does that fix the AXIS display?
[18:58:53] <rayh> Yep
[18:59:33] <rayh> Well I don't see the path through the display numbers but it does show properly around them.
[19:01:59] <alex_joni> hrmm
[19:23:36] <rayh> Oh. After further testing, I can see just a bit of the tool behind the display number box.
[19:24:28] <rayh> So the vesa driver does give the proper functionality to boards that use the via unichrome display hardware.
[19:25:43] <jepler> it's fairly faint (25% if I recall correctly)
[19:30:56] <jepler> SWPadnos: do both of your recent stepconf fixes need to be backported to 2.2 as well?
[19:31:40] <SWPadnos> yes
[19:32:03] <SWPadnos> I may wait until it works for this guy before doing that
[19:32:08] <jepler> ok
[19:33:06] <SWPadnos> here are his files, in case you're interested: http://pastebin.ca/1027234 http://pastebin.ca/1027235 http://pastebin.ca/1027239
[19:45:16] <jepler> rayh: if paul_c actually sent you some kind of network diagnostic information, please forward it on to me, or to the board.
[19:45:26] <alex_joni> or pastebin it
[19:45:46] <alex_joni> (hi paul)
[20:09:25] <rayh> You bet. It dies between denver and unpythonic.
[20:11:14] <rayh> But that location is the same as my trace so I don't know why his would not get through.
[20:12:46] <alex_joni> can you pastebin it? would be best
[20:12:47] <SWPadnos> what's the packet size of his pings?
[20:12:54] <SWPadnos> or traceroutes or whatever
[20:14:13] <cradek> when I traceroute paul's last known IP address from the header of his emails FROM the cvs server, I get through London2.Level3.net to bddsl.net and it stops there. This may be because of changing static IPs, firewalling out there, etc.
[20:16:29] <cradek> I see that bddsl.net is an ISP in the UK called bulldog. I suspect my trace gets within a hop or two of his machine.
[20:17:01] <cradek> rayh: I'd also like to see the 'diagnostic' output
[20:17:23] <rayh> http://www.pastebin.ca/1027364 but get it quick cause it expires in 15 min.
[20:19:45] <cradek> maybe he has a misconfigured TTL set to 20. Possibly that's on whatever is.
[20:20:06] <cradek> I agree with rayh that my trace goes through that last address without a problem.
[20:23:30] <cradek> when I trace back the other way I get through all those machines right up to where I'm guessing his stuff starts (right after 83.146.22).
[20:23:48] <SWPadnos> that is the last address before it hits inetnebr
[20:23:52] <cradek> paul_c: check the TTL on your internal machines and whatever is doing NAT at
[20:25:23] <jepler> my traceroute got all the way to 83..123.
[20:26:08] <cradek> especially if it's one of those router-and-nat-in-a-box things, it would be easy to have it misconfigured and not know about it
[20:26:31] <jepler> cradek: was 15:22 when you ran a trace starting at cvs.linuxcnc.org?
[20:26:54] <cradek> yes
[20:27:01] <jepler> OK