#emc | Logs for 2004-10-19

Back
[00:31:01] <CIA-4> 03paul_c 07auto_configure_0_1 * 10emc2/configure.in: Modify the way CFLAGS & co are set - Should work now..
[00:31:58] <trp> Paul_c, Ray told us that we may get something today that is finished. Could this be true?
[00:35:14] <paul_c> Need the rc12 image moved (or copied) onto the smithy box
[00:45:11] <alex_joni> paul_c: I was just doing some work to configure.in,
[00:45:16] <alex_joni> guess you were faster ;)
[00:45:19] <alex_joni> just one thing...
[00:45:33] <alex_joni> I have rtai 24.1.12 here
[00:46:07] <alex_joni> the ./configure finds /usr/src/rtai-24.1.12/scripts/realtime-config script
[00:46:22] <alex_joni> but... realtime-config --prefix outputs /usr/realtime
[00:46:28] <alex_joni> and there is no such thing...
[00:47:00] <paul_c> note in configure.in
[00:47:07] <paul_c> Use with caution !
[00:52:39] <paul_c> make install_rt_devel
[00:53:27] <paul_c> make install_runtime
[00:53:48] <trp> paul_c, If I load rc12 it will wipe out everything else. Is this ok?
[00:54:09] <paul_c> Noooo...
[00:54:28] <paul_c> Just copy the ISO image to /home/smithy
[00:54:29] <trp> How do you want me to do it. download...
[00:54:41] <trp> I can do that just fine
[00:54:44] <paul_c> Thought you had already downloaded it...
[00:54:55] <trp> I did from my laptop
[00:55:07] <paul_c> Burnt it to CD ?
[00:55:28] <trp> I cant get out of the office very long but will get it done immediately
[00:57:00] <alex_joni> hmmm... make install_rt_devel && install_runtime
[00:57:44] <alex_joni> only installed include in /usr/realtime
[00:58:41] <paul_c> install_rt_devel should also install scripts/realtime-config
[01:00:23] <paul_c> * paul_c is reading from the top level makefile in /usr/src/rtai
[01:09:08] <alex_joni> it does .. but into /usr/bin
[01:10:09] <alex_joni> install -m 755 scripts/realtime-config /usr/bin
[01:13:38] <paul_c> ${prefix}/bin to be exact....
[01:18:32] <alex_joni> strange... I get some strange errors ...
[01:18:34] <alex_joni> from awk
[01:19:26] <paul_c> The alternatif is to strip off /scripts/realtime-config from $RTS
[01:19:56] <paul_c> s/alternatf/alternative/
[01:23:21] <alex_joni> can you check if it works for you?
[01:25:21] <CIA-4> 03alex_joni 07auto_configure_0_1 * 10emc2/configure.in: some minor tweaks
[01:28:45] <paul_c> where were you getting awk errors ?
[01:31:17] <trp> paul_c, smithy rc12 is on the box in the home directory. same port as before
[01:31:33] <alex_joni> well.. thing is...
[01:31:43] <alex_joni> now that I installed rtai-24.1.x
[01:31:51] <alex_joni> it finds 2 realtime-config
[01:31:55] <alex_joni> one in /usr/bin
[01:32:03] <alex_joni> and one in /usr/src/rtai...
[01:32:10] <alex_joni> so search doesn't work
[01:32:23] <alex_joni> when I run --with-rtai=/usr/realtime
[01:32:49] <alex_joni> I get:
[01:32:50] <alex_joni> checking for RT dir... configure: specified rtai=/usr/realtime/
[01:32:50] <alex_joni> /usr/realtime set from command line
[01:32:50] <alex_joni> Usage: awk [POSIX or GNU style options] -f progfile [--] file ...
[01:32:50] <alex_joni> Usage: awk [POSIX or GNU style options] [--] 'program' file ...
[01:36:06] <alex_joni> problem is with : echo $RTS | awk -F /'{print $NF }'
[01:37:23] <paul_c> which works here...
[01:37:53] <alex_joni> well... actually it needs to be awk -F/ '{print $NF }'
[01:39:06] <alex_joni> still detects the system as RTAI-3
[01:40:49] <alex_joni> * alex_joni is digging ;)
[01:43:30] <paul_c> trp: Will dive in and start tweaking...
[01:44:46] <alex_joni> very strange....
[01:44:57] <alex_joni> if (test `echo $RTS | awk -F/ '{print $NF }'`=rtai-config); then echo "yes" ; else echo "no" ; fi
[01:45:02] <alex_joni> returns yes
[01:45:09] <alex_joni> when $RTS=realtime-config
[01:46:05] <alex_joni> if (test `echo $RTS | awk -F/ '{print $NF }'`="rtai-config"); then echo "yes" ; else echo "no" ; fi
[01:46:36] <alex_joni> but
[01:46:38] <paul_c> I'll leave it in your capable hands for a while..
[01:46:46] <alex_joni> if (test `echo $RTS | awk -F/ '{print $NF }'` = "rtai-config"); then echo "yes" ; else echo "no" ; fi
[01:46:49] <alex_joni> works :((
[01:47:33] <cradek> what do you want that to do?
[01:47:48] <alex_joni> hello...
[01:47:56] <cradek> hi!
[01:47:56] <alex_joni> I think I sorted it out...
[01:48:03] <cradek> to get the word after the last /, use ${RTS**#/}
[01:48:05] <alex_joni> check those 2 lines...
[01:48:06] <cradek> doesn'err
[01:48:10] <cradek> ${RTS##*/}
[01:48:27] <cradek> if [ ${RTS##*/} = rtai-config ]; then
[01:48:32] <alex_joni> ${RTS##*/} ????
[01:48:37] <alex_joni> nice....
[01:48:39] <cradek> doesn't need other programs
[01:49:08] <cradek> you should probably actually write
[01:49:13] <cradek> if [ x${RTS##*/} = xrtai-config ]; then
[01:49:19] <cradek> so it doesn't error if $RTS is unset
[01:49:39] <cradek> or you can use
[01:49:43] <cradek> case $RTS in
[01:49:50] <cradek> *rtai-config) echo yes;;
[01:49:52] <cradek> *) echo no;;
[01:49:53] <cradek> esac
[01:50:15] <cradek> which also uses only sh constructs and requires no subprograms
[01:50:32] <alex_joni> I can't use []
[01:50:40] <cradek> why not?
[01:50:50] <alex_joni> because it get's stripped away by autoconf
[01:50:54] <cradek> oh
[01:50:55] <alex_joni> must use test
[01:51:06] <cradek> ok, test is the same as [
[01:51:22] <alex_joni> yup
[01:51:28] <cradek> if test ${RTS##*/} = rtai-config; then echo yes; else echo no; fi
[01:51:28] <alex_joni> don't I need ""?
[01:51:40] <alex_joni> if test ${RTS##*/} = "rtai-config"; then echo yes; else echo no; fi
[01:51:51] <cradek> no
[01:51:57] <cradek> "" only keep multiple words together
[01:52:01] <cradek> rtai-config is one word
[01:52:37] <alex_joni> I see... well that clears stuff up ;)
[01:52:39] <alex_joni> let me check
[01:54:37] <cradek> you really should put in x or . before the ${} and rtai-config unless you know for sure that RTS is set to something
[01:54:49] <alex_joni> what if there is no /
[01:55:06] <alex_joni> I can have RTS=rtai-config
[01:55:10] <alex_joni> found from PATH
[01:55:16] <cradek> do you want it to say yes? it says yes
[01:55:33] <alex_joni> well.. something's still broken
[01:55:35] <cradek> "remove everything up to and including the leftmost /"
[01:55:38] <cradek> err
[01:55:39] <cradek> rightmost
[01:55:59] <alex_joni> config.status: creating Makefile.inc
[01:55:59] <alex_joni> sed: file ./confstatA6cJy9/subs-1.sed line 39: Unterminated `s' command
[01:56:19] <cradek> uh-oh
[01:56:36] <cradek> that probably means you have s/a/b with no terminating /
[01:56:45] <cradek> (or whatever the separator is)
[01:56:48] <alex_joni> well... in config.status
[01:56:55] <alex_joni> which is created by autoconf :(
[01:57:11] <alex_joni> don't think I wanna go there
[01:58:21] <cradek> ugh
[01:58:56] <cradek> I fear/hate/avoid autoconf but am happy to help with sh programming
[02:00:37] <alex_joni> I went back to awk -F/
[02:00:55] <alex_joni> I think autoconf is confused by ${RTS##*/}
[02:01:02] <cradek> oh
[02:01:14] <alex_joni> I guess I should comment that out from autoconf... gotta check the docs
[02:01:20] <alex_joni> but I'm too lazy right now
[02:01:35] <cradek> * cradek cuts off his left arm and beats autoconf with it
[02:03:07] <alex_joni> hihi
[02:06:17] <alex_joni> could you check that it works for you?
[02:06:49] <cradek> I don't even have the tree checked out
[02:07:11] <cradek> I just saw you struggling with three kinds of quotes in one line of sh that could be written more easily
[02:07:38] <cradek> but I guess it can't, because of autoconf!
[02:07:49] <alex_joni> well.. maybe it can...
[02:07:58] <alex_joni> same problem .. even after going back to awk
[02:08:08] <cradek> ah
[02:08:38] <alex_joni> so I guess that was ok...
[02:08:56] <alex_joni> if you want... do a checkout -r auto_configure_0_1 emc2
[02:09:35] <cradek> ok, just a minute
[02:12:56] <alex_joni> I just took out a big part, and the error goes away... :(
[02:13:09] <cradek> binary search!
[02:13:57] <cradek> kk
[02:13:57] <cradek> oops
[02:14:04] <alex_joni> is this ok: (test -r $($RTS --module-dir)/rtai_math.o ) ?
[02:14:34] <alex_joni> what if RTS doesn't have a --module-dir function ? .. the whole thing should fail.. right?
[02:14:48] <cradek> it will not fail elegantly
[02:15:28] <alex_joni> I saw that ;9
[02:15:29] <alex_joni> ;)
[02:15:31] <alex_joni> well..
[02:15:52] <alex_joni> checking yet another part...
[02:15:53] <alex_joni> :((
[02:16:50] <alex_joni> found the problem....
[02:16:53] <alex_joni> sed s/' = '/'='/
[02:17:33] <alex_joni> can you rewrite this more nicely?
[02:17:34] <alex_joni> KERNELDIR=$($RTS --dump | grep LINUXDIR | sed s/' = '/'='/ | awk '{split($1, A, "=")}{print A[2]}') 2>/dev/null
[02:18:18] <cradek> eeeekkk
[02:18:22] <cradek> tell me what it's supposed to do :-)
[02:18:26] <alex_joni> that's paul work
[02:18:27] <alex_joni> :D
[02:18:37] <paul_c> wasn't me....
[02:18:41] <alex_joni> no?
[02:19:00] <alex_joni> anyways... here $RTS --dump produces a linuxdir
[02:19:04] <alex_joni> not LINUXDIR
[02:19:12] <paul_c> trying to split a string on the =
[02:19:28] <paul_c> and ignore any white spaces either side
[02:19:39] <paul_c> if present
[02:19:46] <cradek> and get the thing after the =
[02:19:55] <paul_c> yes
[02:20:01] <alex_joni> linuxdir="/usr/src/linux"
[02:20:19] <alex_joni> you need to loose the "" too
[02:22:44] <cradek> echo 'a = "b"' |awk -F'[ ="]+' '{print $2}'
[02:22:48] <cradek> b
[02:23:42] <CIA-4> 03alex_joni 07auto_configure_0_1 * 10emc2/configure.in: fixed some awk and test bugs
[02:23:49] <cradek> KERNELDIR=$($RTS --dump |grep -i linuxdir |awk -F'[ ="]+' '{print $2}')
[02:23:56] <cradek> (untested)
[02:24:19] <alex_joni> something like that
[02:24:26] <alex_joni> but sometimes LINUXDIR...
[02:24:49] <alex_joni> let me check
[02:25:02] <cradek> grep -i fixes that
[02:25:29] <alex_joni> ohhh.. really ?
[02:26:00] <cradek> and it will work with and without whitespace around the =, and with and without " around the answer
[02:26:44] <cradek> (also untested, heh)
[02:27:54] <alex_joni> I added a few more tcl and tk libs to be searched for Tcl_Init
[02:28:43] <cradek> ok I finally got a checkout
[02:29:34] <alex_joni> hm... still sed complaining :(
[02:29:51] <cradek> how do I run this so I can see the error?
[02:30:08] <alex_joni> first an autoconf
[02:30:13] <alex_joni> then ./configure
[02:30:26] <cradek> checking for RT dir... configure: error: RT not found, try to specify one by --with-rtai=<path>, --with-rtlinux=<path>, --with-rtlinuxpro=<path>
[02:30:29] <cradek> heh
[02:30:31] <cradek> (on my work machine)
[02:30:35] <alex_joni> yup...
[02:30:41] <cradek> so ... that part works
[02:30:41] <alex_joni> got it
[02:30:44] <cradek> * cradek is very little help
[02:31:00] <alex_joni> awk -F'[] gets replaced by autoconf
[02:31:05] <alex_joni> [] go away ;)
[02:31:11] <cradek> oh good grief
[02:31:28] <cradek> and I only have one arm left
[02:32:07] <alex_joni> well... I shouldn't have any left ;)
[02:32:27] <cradek> how about if you put \ before the [ and ]?
[02:32:36] <alex_joni> I put [[ and ]]
[02:32:40] <cradek> aha
[02:33:14] <alex_joni> now the output is: KERNELDIR=$($RTS --dump | grep -i linuxdir | awk -F'[ ="]+' '{print $2}')
[02:33:51] <cradek> yay
[02:33:53] <cradek> does it work?
[02:33:53] <alex_joni> which seems ok
[02:34:23] <cradek> huuuungry
[02:34:42] <alex_joni> nope :(
[02:34:49] <alex_joni> sed: file ./confstatcvVVrU/subs-1.sed line 39: Unterminated `s' command
[02:35:55] <cradek> wish I could help but I don't think I can get that on my non-rt system
[02:36:30] <cradek> KERNELDIR=`$RTS --linux-dir`
[02:36:36] <cradek> I see this somewhere else in configure.in
[02:36:38] <cradek> doesn't it work?
[02:37:24] <alex_joni> well.. it depends on the type of rt installed
[02:37:32] <cradek> oh, I see
[02:37:35] <alex_joni> RTAI-3.x has $RTS --linux-dir
[02:37:45] <alex_joni> RTAI-24.1.x has no such thing
[02:39:03] <alex_joni> is this ok ?
[02:39:04] <alex_joni> RTFLAGS=`$RTS --cflags-fp`
[02:39:04] <alex_joni> # RTFLAGS="$RTFLAGS -DRTAI=1"
[02:39:05] <paul_c> which is why it needs to be grepped out from --dump
[02:39:09] <alex_joni> without the #
[02:40:50] <alex_joni> if I remove RTFLAGS from ./config.status it works
[02:41:00] <alex_joni> it seems the value in config.status is on two lines
[02:41:14] <alex_joni> * alex_joni checks where RTFLAGS get's assigned
[02:43:24] <alex_joni> RTFLAGS = -D__KERNEL__ -DMODULE -I/usr/src/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -pip
[02:43:24] <alex_joni> e -march=i586
[02:43:24] <alex_joni> -D__KERNEL__ -DMODULE -I/usr/src/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -pipe -march=i
[02:43:24] <alex_joni> 586 -fno-rtti -DRTAI=1
[02:43:51] <alex_joni> I don't see where -fno-rtti is coming from
[02:46:25] <alex_joni> so... RTFLAGS is defined twice...
[02:46:41] <alex_joni> on two lines, that's why sed thinks s/ is not ended
[02:49:51] <paul_c> commit your changes...
[02:51:26] <CIA-4> 03alex_joni 07auto_configure_0_1 * 10emc2/configure.in: somemore tweaks for rtai-24.1.x
[02:52:54] <alex_joni> if I comment out :
[02:53:03] <alex_joni> RTFLAGS=`$RTS --cflags-fp`
[02:53:04] <alex_joni> RTFLAGS="$RTFLAGS -DRTAI=1"
[02:53:06] <alex_joni> it works...
[02:54:41] <alex_joni> seems that realtime-config --cflags-fp
[02:54:44] <alex_joni> is the problem
[02:55:35] <alex_joni> * alex_joni is angry ;)
[02:55:48] <alex_joni> realtime-config --cflags-fp
[02:55:56] <alex_joni> outputs cflags-fp
[02:56:15] <alex_joni> AND cplusplusflags-fp
[02:57:01] <alex_joni> there is a BUG inside scripts/realtime-config ...
[02:57:18] <alex_joni> if test "$echo_cflags_fp" = "yes";then
[02:57:18] <alex_joni> echo $cflags_fp
[02:57:18] <alex_joni> fi
[02:57:18] <alex_joni> if test "$echo_cplusplusflags" = "yes";then
[02:57:18] <alex_joni> echo $cplusplusflags_nfp
[02:57:18] <alex_joni> fi
[02:57:20] <alex_joni> if test "$echo_cflags_fp" = "yes";then
[02:57:22] <alex_joni> echo $cplusplusflags_fp
[02:57:24] <alex_joni> fi
[03:00:13] <alex_joni> same BUG in all rtai's 24.1.10, 24.1.11, 24.1.12, 24.1.13
[03:00:35] <paul_c> Updating local copy on the BDI-TNG box...
[03:02:46] <paul_c> fsckit - need autoconf-2.57
[03:03:00] <paul_c> have 2.13 installed.
[03:03:07] <alex_joni> ;)
[03:03:57] <alex_joni> it works now
[03:04:03] <alex_joni> I added a head -n 1
[03:05:20] <CIA-4> 03alex_joni 07auto_configure_0_1 * 10emc2/configure.in: workaround for bug in rtai-24.1.x script
[03:05:24] <paul_c> suggestion: Commit the configure & configure.in files
[03:05:50] <alex_joni> I wanna fix the libm problem
[03:05:56] <alex_joni> then I'm gone
[03:06:12] <Imperator_> Hi Alex
[03:06:17] <alex_joni> hello
[03:06:23] <alex_joni> I got my own bot now :D
[03:06:30] <CIA-4> 03alex_joni 07auto_configure_0_1 * 10emc2/configure: workaround for bug in rtai-24.1.x script
[03:09:37] <alex_joni> short question
[03:09:47] <alex_joni> if I have more tests..
[03:09:52] <alex_joni> if (test1 && test2)
[03:10:04] <alex_joni> and if test1 is false, does test2 get evaluated?
[03:14:02] <paul_c> pass
[03:15:21] <alex_joni> I added a search for /lib/modules/`uname -r`/rtai/rtai_libm.o
[03:16:33] <paul_c> which may fail if you are compiling for another kernel
[03:16:43] <paul_c> than the one that is running
[03:16:50] <alex_joni> should I remove it?
[03:17:15] <alex_joni> as it is now... it's the only one found for my machine
[03:17:25] <alex_joni> in /usr/realtime there is no link to the modules
[03:17:26] <alex_joni> :(
[03:20:14] <CIA-4> 03alex_joni 07auto_configure_0_1 * 10emc2/configure.in: added libm support for rtai-24.1.x
[03:20:35] <alex_joni> if you think it's not ok.. I'll remove it...
[03:21:13] <paul_c> We can work on it...
[03:21:47] <alex_joni> guess I'm going home now...
[03:21:53] <alex_joni> it's been a looong day at work :D
[03:22:59] <CIA-4> 03alex_joni 07auto_configure_0_1 * 10emc2/configure: added libm support for rtai-24.1.x
[03:24:48] <alex_joni> paul_c: there is a check in autoconf whether we are cross-compiling or not...
[03:24:55] <alex_joni> maybe that one could be used :-?
[03:26:28] <paul_c> I think you'll find that is for arch cross-compiling
[03:26:59] <alex_joni> I see :(
[03:28:12] <alex_joni> * alex_joni is adding a check for kernel-cross-compiling to his TODO list
[03:29:00] <alex_joni> well... gotta go...
[03:29:49] <paul_c> I'll test on a BDI-2.xx box..
[03:32:00] <alex_joni> post your results to the channel
[03:33:33] <alex_joni> I have now a logger to see the results *g*
[03:33:53] <alex_joni> * alex_joni runs home to get something to eat...
[06:28:26] <CIA-4> 03zwisk 07auto_configure_0_1 * 10emc2/Makefile:
[06:28:26] <CIA-4> Changes to the makefile to install man pages in their proper DESTDIR
[06:28:26] <CIA-4> location, as well as creating that directory if it doesn't already exist.
[07:10:50] <trp> GOODNIGHT
[07:38:31] <CIA-4> 03paul_c 07auto_configure_0_1 * 10emc2/ (configure.in configure): Noted a few issues with tcl/tk headers and libs...
[07:56:35] <CIA-4> 03paul_c 07auto_configure_0_1 * 10emc2/ (configure configure.in): Tweaked the gcc section a little..
[13:28:15] <CIA-4> 03zwisk 07auto_configure_0_1 * 10emc2/ (6 files in 3 dirs):
[13:28:15] <CIA-4> Support for --with-module-dir= which sets the location of kernel modules
[13:28:15] <CIA-4> Support for --with-kernel which allows you to compile for a kernel other
[13:28:15] <CIA-4> than the one currently running (Replaces all "uname -r's" with kernelver)
[13:28:15] <CIA-4> Addition of install_init, which puts the realtime start/stop script in
[13:28:15] <CIA-4> /etc/rc.d/init.d/
[14:02:48] <zwisk> anybody around?
[14:35:39] <CIA-4> 03zwisk 07auto_configure_0_1 * 10emc2/Makefile:
[14:35:39] <CIA-4> Changed the install location from configs to $sysconfdir
[14:35:39] <CIA-4> This seems like the right variable, but the default of $prefix/etc
[14:35:39] <CIA-4> probably needs to be changed yet.
[16:46:54] <alex_joni> hello
[17:18:38] <zwisk> howdy...
[17:19:27] <alex_joni> hello
[17:19:53] <alex_joni> I've seen the work you've been doing ... great ;)
[17:19:57] <zwisk> looks like you've been busy on the autoconfig work.
[17:20:01] <zwisk> ditto. :)
[17:21:34] <alex_joni> how's the autoconfig stuff working for you?
[17:21:48] <alex_joni> btw... what rt do you have?
[17:21:52] <zwisk> so far so good. I've been working on install aspects, mostly.
[17:21:57] <zwisk> I'm using rtai-3.1
[17:22:02] <alex_joni> nice...
[17:22:15] <alex_joni> searching for rtai should work..:-?
[17:22:29] <alex_joni> doesn't it?
[17:22:35] <zwisk> seems to do quite well. yep, works fine...
[17:22:58] <zwisk> which rt are you using?
[17:23:08] <alex_joni> here I have rtai-24.1.12
[17:23:14] <alex_joni> and at home rtai-3.0r4
[17:23:42] <zwisk> cool.
[17:24:13] <alex_joni> I think searching is not really ok
[17:24:47] <alex_joni> the thing is .. it searches for rtai-config (RTAI-3.x) realtime-config (RTAI-24.1.x) rtl-config (RTLinux ???)
[17:25:05] <alex_joni> and it searches in /usr/*
[17:25:12] <zwisk> I haven't looked at RTLinux.
[17:25:21] <alex_joni> if it finds more than one it gives up...
[17:25:34] <zwisk> hmm... once I added /usr/realtime it seemed to work for 3.1
[17:25:53] <alex_joni> yup.. it should fin /usr/realtime/bin/rtai-config
[17:25:58] <zwisk> Paul wanted it to give up if it found more than one, right? To force a user choice.
[17:26:03] <alex_joni> exactly...
[17:26:15] <alex_joni> and that's what happens here
[17:26:24] <alex_joni> it finds one in /usr/src/rtai...
[17:26:34] <alex_joni> and one installed in /usr/realtime or whatever
[17:27:09] <alex_joni> it's good that it stops when more than one is found... but sometimes it's the same...
[17:27:23] <zwisk> hmm... good point.
[17:27:27] <alex_joni> I'm thinking maybe to compare the two...
[17:28:01] <alex_joni> well... maybe I'll give it a shot later ;)
[17:28:34] <alex_joni> it's pretty early for me right now (around 10 a.m., I'm at work for about 2 hours.. but still not 100% capable ;)
[17:28:52] <zwisk> :)
[17:29:12] <alex_joni> btw... I installed a logger yesterday
[17:29:21] <zwisk> It's 12:22am here, so I'm on the other end of my day :)
[17:29:30] <alex_joni> where are you from?
[17:29:31] <zwisk> a logger on #emc?
[17:29:37] <zwisk> California
[17:29:38] <alex_joni> yes..
[17:29:47] <alex_joni> ohh... pretty late ;)
[17:29:56] <alex_joni> I've got some friends up there...
[17:30:01] <alex_joni> logger_aj, bookmark
[17:30:01] <alex_joni> See http://193.226.12.129/irc/irc.freenode.net:6667/emc/2004-10-19#T17-30-01
[17:30:08] <zwisk> Where are you?
[17:30:27] <alex_joni> europe->romania->timisoara ;)
[17:31:03] <zwisk> Wow.. Nice place? Would I want to take a vacation there? :)
[17:31:40] <zwisk> cool logger. Pretty handy.
[17:32:42] <alex_joni> well.. I think you would like it here
[18:02:14] <alex_joni> still around?
[18:03:24] <zwisk> yup
[18:03:38] <zwisk> but fading fast..
[18:05:29] <alex_joni> if you care for some pictures from here
[18:05:35] <alex_joni> you could check my homepage ...
[18:05:51] <alex_joni> www.juve.ro
[20:30:11] <alex_joni> hi paul
[20:31:11] <paul_c> Hi Alex
[20:31:22] <alex_joni> small problem.. I'm about to fix
[20:31:33] <alex_joni> realtime-config -cc doesn't exist
[20:31:51] <paul_c> Yes...
[20:32:11] <paul_c> I think we need to grep .buildvars for it
[20:32:37] <paul_c> and another thing wrong with realtime-config
[20:32:41] <alex_joni> I was thinking abou tone thing...
[20:32:56] <alex_joni> when by search multiple RTS'es are found
[20:32:58] <paul_c> It doesn't return all the CFLAGS
[20:33:11] <alex_joni> :(
[20:33:19] <alex_joni> well.. so when multiple are found...
[20:33:26] <alex_joni> maybe we could check if the files are the same
[20:33:34] <alex_joni> that should work for symlinks too...
[20:33:45] <paul_c> agreed
[20:33:49] <alex_joni> on my machine it finds /usr/bin/realtime-config
[20:33:58] <alex_joni> and /usr/src/rtai/scripts/realtime-config
[20:34:08] <alex_joni> well... unfortunately .. I gotta run for today
[20:34:11] <paul_c> Found the symlink on the BDI-2.xx box caused the test to fail
[20:34:17] <alex_joni> gotta go to a customer about 60 km from here
[20:34:31] <paul_c> Who added the --with-kernel option ?
[20:34:39] <alex_joni> I think zwisk
[20:34:49] <alex_joni> don't you like that?
[20:34:53] <paul_c> OK - I'll take it up with him...
[20:35:04] <alex_joni> don't think he's around... anymore
[20:35:19] <paul_c> The kernel version is already defined in the RT-config
[20:35:42] <paul_c> to over-ride it would lead to BIG problems
[20:35:42] <alex_joni> I think he was meaning the kernel version emc2 is to be compiled for...
[20:36:03] <paul_c> That can be determined by rt-config
[20:36:17] <alex_joni> maybe a check if those two match...
[20:36:28] <alex_joni> let's say... rt has been configured for a kernel
[20:36:32] <alex_joni> or for more
[20:36:40] <paul_c> doesn't seem right to have to do that...
[20:36:40] <alex_joni> and user tries to compile for a specific kernel
[20:36:53] <alex_joni> well... just my 5 pence ;)
[20:37:03] <alex_joni> talk to him.. he should know why he did it ;)
[20:37:13] <alex_joni> gotta run now... got some robots waiting for me...
[20:37:22] <paul_c> OK - later.
[20:38:39] <alex_joni> interesting mail I just got on emc-developer
[20:38:46] <alex_joni> from him.. about make install..
[20:39:57] <alex_joni> bye...
[23:02:43] <trp> Good afternoon paul_c. How are things going today?