00:06:07 <rayh> back
00:06:13 <SWPadnos> hiya
00:06:37 <rayh> I did set line buffering on several tests.
00:06:51 <rayh> Need to clean up this script a bit brb
00:06:56 <SWPadnos> ok
00:36:31 <SWPadnos_> SWPadnos_ has joined #emc-devel
00:42:52 <jmkasunich> jmkasunich has joined #emc-devel
04:30:43 <jmkasunich> jmkasunich has quit
06:21:39 <SWPadnos> SWPadnos is now known as SWP_Away
06:41:01 <LawrenceG> LawrenceG has quit
06:42:09 <LawrenceG> LawrenceG has joined #emc-devel
13:41:16 <rayh> rayh has quit
16:13:26 <SWP_Away> SWP_Away is now known as SWPadnos
16:55:10 <alex_joni> alex_joni has quit
16:55:37 <rayh> rayh has joined #emc-devel
16:57:03 <rayh> SWPadnos: You available for a couple minutes.  Got a thought on halcmd -skf.
16:57:35 <SWPadnos> sure
16:59:03 <rayh> I am able to loop the gets or read commands and see the output of a show xx
16:59:29 <SWPadnos> but you can't tell when it's done?
16:59:31 <rayh> What I can't do is quit the loop when it runs out of stuff.
16:59:58 <rayh> could we/you put an eof at the end of each reply
17:00:16 <SWPadnos> eof would be bad, I think.  the pipe should close on EOF
17:00:51 <rayh> We can define a different eof as far as the channel is concerned.
17:00:59 <SWPadnos> ah
17:01:19 <SWPadnos> there should be an option to print the prompt while in script mode - that was something that jmk and I had discussed
17:01:21 <rayh> something that a terminal might ignore
17:01:26 <SWPadnos> when you get the prompt, the output is over
17:02:01 <rayh> prompt? like halcmd:
17:02:05 <SWPadnos> yes
17:02:19 <SWPadnos> you still want the other features of sript mode, so you don't want to leave out the '-s'
17:02:37 <SWPadnos> it's aeasy to change so that the prompt is still printed in script mode though
17:02:57 <rayh> How about a % for -s prompt
17:03:03 <SWPadnos> I can tell you how to do it in 1 minute (can't do it myself, since the emc machine isn't on)
17:03:08 <SWPadnos> that would work as well
17:03:14 <rayh> k
17:04:00 <SWPadnos> load up src/hal.utils/halcmd.c
17:04:07 <alex_joni> alex_joni has joined #emc-devel
17:04:36 <SWPadnos> hi Alex
17:04:53 <rayh> k
17:05:12 <SWPadnos> two changes to make
17:05:28 <SWPadnos> the first is at lines 307-309
17:05:38 <SWPadnos> comment those lines out
17:05:58 <SWPadnos> (should be: if scriptmode !=0 promptmode=0)
17:06:31 <rayh> is // okay
17:06:35 <SWPadnos> yes
17:06:40 <rayh> done
17:07:10 <SWPadnos> ok.  go to line 352, and change it from this:
17:07:14 <SWPadnos> 	    rtapi_print("halcmd: ");
17:07:18 <SWPadnos> to this:
17:07:50 <SWPadnos> 	    rtapi_print(scriptmode ? "% " : "halcmd: ");
17:08:15 <SWPadnos> heh - that looks almost as bad as ASM in gcc ;)
17:09:16 <rayh> k
17:09:29 <SWPadnos> save, and recompile
17:10:53 <rayh> error 2
17:10:56 <rayh> phone
17:10:59 <SWPadnos> ok
17:11:10 <SWPadnos> error 2 is "phone"?
17:11:53 <rayh> halcmd.c:477: parse error before `if'
17:12:45 <SWPadnos> did you get all 3 lines (including the close curly brace) at 307-309 ?
17:13:26 <SWPadnos> you can just delete those lines instead
17:14:36 <SWPadnos> oops - the % probably needs a \ in front of it:
17:14:54 <SWPadnos> rtapi_print(scriptmode ? "\% " : "halcmd: ");
17:15:02 <SWPadnos> or %%
17:15:59 <rayh> k
17:17:25 <SWPadnos> I shouldn't code before coffee.  change it like so:
17:17:47 <SWPadnos> 	    rtapi_print(scriptmode ? "\%\n" : "halcmd: ");
17:18:01 <SWPadnos> that'll make it work with gets, since you need the newline
17:18:03 <rayh> It compiled with the previous.
17:18:10 <SWPadnos> compiled, but won't work ;)
17:19:05 <rayh> fixed
17:19:17 <SWPadnos> that last line will print a percent sign, then a newline as the "finished" signal (or prompt)
17:23:32 <rayh> compiled
17:28:14 <SWPadnos> cool
17:34:43 <rayh> rayh@ray64:~/emcdevelop/emc2$ bin/halcmd -skf
17:34:43 <rayh> %
17:34:43 <rayh>  show sig M
17:34:44 <rayh> bit    FALSE  MySig1
17:34:44 <rayh> bit    FALSE  MySig2
17:34:45 <rayh> halcmd:
17:35:02 <SWPadnos> oops - there are two prompt prints - hold on a sec.
17:36:02 <SWPadnos> line 467 (unless you deleted the other 3 lines)
17:36:20 <SWPadnos> make it the same as line 352
17:36:31 <rayh> k
17:37:03 <SWPadnos> (that last rtapi_print line I pasted, in case the line numbers don't match up)
17:38:36 <SWPadnos> you may not get a prompt after you send a "quit" line, by the way
17:38:50 <SWPadnos> not that you'd really need it in that instance
17:45:29 <rayh> No difference with the fconfigure  still locks up during a gets.
17:45:49 <SWPadnos> at the end of the output?
17:45:49 <rayh> I could use a string search command I suppose but that takes time.
17:47:50 <rayh> It does show the % as an eof
17:48:05 <rayh> * rayh goes of to play with the new toy.
17:48:13 <SWPadnos> ok.  does the fgets hang after the halcmd output?
17:50:48 <rayh>  fgets??
17:50:53 <SWPadnos> sorry - gets
17:51:07 <SWPadnos> fgets in c "get string from file"
17:51:17 <SWPadnos> rather than stdin
17:51:45 <rayh> I knew I'd heard of it.  Yes.  I am using blocking and line buffering.
17:53:21 <SWPadnos> when does the gets hang?  is it waiting for more input at the end of the halcmd output, or is there something else going on?
17:57:12 <rayh> still testing to see where.
17:57:16 <SWPadnos> ok
18:21:04 <rayh> If I don't use fconfigure  -eofchar {"%" {}}
18:21:31 <rayh> I can pick up the output until it runs out of data.
18:21:38 <SWPadnos> you don't want to treat it as an eof, since that would normally cause the file handle to be closed
18:22:00 <SWPadnos> you should read a line, and then stop reading if you see that the line returned is "%"
18:22:04 <rayh> If I use the % then it does not return anything.
18:22:25 <SWPadnos> hmmm
18:23:23 <rayh> Yes I can test for % but that will be a lot of checking.  I was hoping that there was a channel based approach.
18:23:41 <SWPadnos> yes - that would be better
18:23:59 <SWPadnos> I'm not sure how tcl uses the EOF char - there may be a mode to tell it not to close the file at eof
18:24:06 <SWPadnos> (though that would be weird)
18:25:57 <rayh> The % does not close the channel.
18:26:03 <SWPadnos> ok - that's good
18:26:32 <SWPadnos> what did you mean when you said "if I use % then it does not return anythinig"?
18:26:40 <SWPadnos> use % where, and what is "it"?
18:45:30 <rayh> as the fconfigure eol
18:45:39 <rayh> eof
18:46:21 <rayh> the command line is fconfigure $tmp -buffering line -blocking yes -eofchar {{%} {}}
18:46:40 <rayh> and with it I get nothing returned from halcmd.
18:46:48 <SWPadnos> what's the second curly brace in the eofchar param?
18:47:25 <SWPadnos> ok - outgoing
18:49:38 <rayh> right.
18:50:01 <rayh> I guess I've got to quit trying and go to something useful.
18:50:14 <SWPadnos> does the gets block?
18:50:29 <rayh> Yes most ways it will
18:50:30 <SWPadnos> ie, do you get no return from gets, or does gets return nothing?
18:50:58 <SWPadnos> can you pop the source into the dropbox or something - I may be able to take a look later today
18:51:04 <rayh> It will block with some combinations of fconfigure and it will return nothing on some.
18:51:33 <rayh> I can certainly send you the test file I'm using if that would help.
18:51:59 <rayh> dcc would be easy but i can email as well.
18:52:01 <SWPadnos> it wouldn't help this minute, since I'm not on a Linux machine (and I don't know what hairy translations will be done going from Windows -> linux)
18:52:08 <SWPadnos> email would be fine
18:52:16 <SWPadnos> spadnos at sover dot net
18:52:41 <rayh> k I'll get it to you.
18:53:00 <SWPadnos> thanks - I'll see if I can figure out anything (considering that I'm a TCL newbie ;) )
18:55:07 <rayh> Thanks.  I'll play for a bit yet.
18:55:10 <SWPadnos> ok
19:29:17 <rayh> borked emc reboot time
19:29:24 <rayh> rayh has quit
19:43:23 <alex_joni> hello
19:43:41 <SWPadnos> hi
19:43:48 <alex_joni> :P
19:43:52 <SWPadnos> :P
19:43:52 <alex_joni> what's up?
19:44:04 <jtr_> jtr_ has quit
19:44:25 <SWPadnos> not too much.  Ray's trying to get tcl to accept halcmd output reliably (piped)
19:44:38 <jtr_> jtr_ has joined #emc-devel
19:46:46 <alex_joni> aha
20:23:37 <SWPadnos> argh
20:25:01 <SWPadnos> this BDI vs. emc2 war is getting very annoying
20:25:25 <cradek> I agree
20:25:48 <cradek> I miss the time we were all working together
20:26:03 <alex_joni> SWPadnos: ditto
20:26:10 <SWPadnos> I'm having a hard time finding the right words to convey the message that it's FUCKING STUPID for everyone to keep acting like 5-year-olds
20:27:23 <cradek> surely not everyone is doing that
20:27:37 <SWPadnos> not everyone, but close, I'd say
20:28:05 <SWPadnos> not necessarily at present, but at some time in this "conflict"
20:48:46 <alex_joni> SWPadnos: what conflict?
20:48:46 <alex_joni> :D
20:49:23 <SWPadnos> yeah - har har ;)
20:51:44 <alex_joni> it's a small dispute at most
20:55:40 <SWPadnos> well, there's a general attitude that the BDI and emc2 projects are adversaries, and in "competition"
20:55:59 <alex_joni> probably "perception" is a better word
20:56:13 <alex_joni> I'm working on emc2, and I'm not trying to compete with anyone
20:56:22 <SWPadnos> I agree
20:56:49 <SWPadnos> on this side, there's an attitude that we can't ask Paul for anything (libraries, ideas for kernelspace issues, etc)
20:57:03 <SWPadnos> packaging ...
20:57:09 <alex_joni> well.. I did ask in the past, and always got a decent answer
20:58:14 <SWPadnos> me, too - I asked about the small libs for KDE-style config files, and got a prompt reply
20:58:28 <SWPadnos> (the library look unmaintained, so it's probably not a good thing for us to use)
20:59:21 <alex_joni> right, he can be a very good sport (sometime)
20:59:37 <SWPadnos> well, everyone can be an asshole sometimes :)
20:59:56 <alex_joni> it's just a matter of inner diameter
21:00:02 <alex_joni> ROFL
21:00:21 <SWPadnos> the list mails are just annoying.  "of course this is possible" (but I'm not going to tell you how I did it) ...
21:02:33 <alex_joni> finally getting cable at home (for internet access I mean)
21:02:42 <SWPadnos> that should be nice
21:02:49 <SWPadnos> high speed is so necessary these days ;)
21:02:54 <alex_joni> yeah, not that expensive either
21:03:01 <alex_joni> 1024kbps for 19EUR
21:03:16 <SWPadnos> not bad
21:03:31 <alex_joni> * alex_joni longs to move to taiwan
21:03:32 <SWPadnos> that's about the same as here - $30 for 2MBits/s
21:03:39 <SWPadnos> yeah -fiber 100mbit
21:03:45 <alex_joni> 54Mbit for 40$
21:05:28 <SWPadnos> there's only one *fast* service in the US at the moment, FIOS from Verizon (other than real connections, like DS3 and the like)
21:06:07 <SWPadnos> I think that's 15 mbits/s
21:06:15 <alex_joni> heh.. I'd like a NOC in my backyard
21:06:20 <alex_joni> btw, do you know 1&1 ?
21:06:25 <SWPadnos> 1
21:06:36 <alex_joni> big server farm in germany
21:06:44 <SWPadnos> ah - never heard of them
21:06:55 <SWPadnos> I've only heard of t-online (now onlinehome, I think)
21:08:26 <alex_joni> t-online is for access
21:08:31 <SWPadnos> yep
21:08:34 <alex_joni> 1&1 is for renting servers & webspae
21:08:36 <alex_joni> space
21:08:49 <SWPadnos> ah.  I may have heard of them
21:08:59 <SWPadnos> why do you ask?
21:09:01 <alex_joni> just read an article about them :D
21:09:05 <alex_joni> NICE
21:09:06 <SWPadnos> ah
21:09:14 <alex_joni> 2 million kW power consumption / month
21:09:20 <alex_joni> imagine that
21:09:30 <SWPadnos> 2M KW?
21:09:35 <alex_joni> yup
21:09:35 <SWPadnos> KWh?
21:09:39 <alex_joni> yup
21:09:45 <SWPadnos> that's a bunch
21:10:08 <alex_joni> they need to announce the power grid when they are doing tests to come offline
21:10:20 <alex_joni> because the excess in energy brings everything down in the area
21:10:21 <alex_joni> :D
21:10:24 <SWPadnos> heh
21:10:34 <SWPadnos> I've done some software for a local company like that
21:10:39 <alex_joni> they also have big diesel power gens.
21:10:50 <alex_joni> 2000 l/h diesel consumption
21:11:00 <SWPadnos> they make large power supplies, and tap directly to the 13.2kV lines
21:11:25 <SWPadnos> they had an oscillation in a 4MW supply, and it caused lights to flicker on the entire trunk
21:11:34 <SWPadnos> for about 5 miles
21:12:34 <alex_joni> heh.. I imagine
21:15:06 <alex_joni> care for a flash?
21:15:12 <alex_joni> http://index.hu/homar/media/honda_civic_choir.swf
21:17:51 <SWPadnos> heh - funny
21:18:05 <SWPadnos> I winder if the people actually made all those sounds
21:18:07 <SWPadnos> wonder
21:20:34 <alex_joni> probably not..
21:20:43 <alex_joni> omg, I'm reading bash.org again :D
21:21:02 <SWPadnos> heh - BOFH! (or are those the irc logs?)
21:22:03 <alex_joni> yeah..
21:22:17 <alex_joni> let me paste an example:)
21:22:26 <SWPadnos> I'm looking at the top 100 now ;)
21:22:36 <SWPadnos> <erno> hm. I've lost a machine.. literally _lost_. it responds to ping, it works completely, I just can't figure out where in my apartment it is.
21:22:38 <alex_joni> ok ;)
21:25:12 <alex_joni> http://www.bash.org/?515093 ROFLMAO
21:25:30 <SWPadnos> yep - that's an old one, but still funny
21:26:29 <SWPadnos> http://bash.org/?602224
21:28:24 <alex_joni> <@AntiHeiss> it was a girl cop, she was pretty cute too
21:28:26 <alex_joni> <@AntiHeiss> she said anything you say can and will be held against you....he sat there for a while and said 'tits'
21:28:34 <SWPadnos> heh
21:29:11 <SWPadnos> <Sceptic> I am not going to argue mideast history and politics here. I have better, more useful things to do--such as draining the ocean with a spoon.
21:29:13 <SWPadnos> <David> Just out of interest, where will you be putting the water?
22:15:52 <logger_devel> logger_devel has joined #emc-devel
22:15:52 <kornbluth.freenode.net> topic is: "Welcome to the Enhanced Machine Control development place. | Regular Developers' meetings 24/7 !"
22:15:52 <kornbluth.freenode.net> Users on #emc-devel: logger_devel jtr_ alex_joni LawrenceG SWPadnos_ SWPadnos jtr_away cradek @ChanServ