**** BEGIN LOGGING AT Sat Dec 23 02:59:56 2006
Dec 23 03:16:56 <eno> right, this as a stop gap method
Dec 23 03:39:08 <chacko> Sinclair73de, by any chance are you a radio amateur?
**** BEGIN LOGGING AT Sat Dec 23 04:30:59 2006
Dec 23 07:04:17 <CIA-8> 03bzhou * r4816 10optware/trunk/make/ (5 files): made cherokee built on slugosbe
Dec 23 07:22:15 <CIA-8> 03bzhou * r4817 10optware/trunk/make/lighttpd.mk: lighttpd: made to work on slugosbe
Dec 23 07:36:16 <CIA-8> 03bzhou * r4818 10optware/trunk/make/apache.mk: apache: check dir existence before cd into it
Dec 23 07:36:50 <eno> g'nite
Dec 23 11:26:31 <CIA-8> 03rwhitby * r84 10slugimage/trunk/slugimage: Make allowance for the header when creating sized empty partitions.
Dec 23 15:59:54 <CIA-8> 03blaster8 * r639 10kernel/trunk/patches/2.6.20/ (KERNEL defconfig squashfs/squashfs-Kconfig.patch): Update to 2.6.20-git11
Dec 23 17:46:54 <jelle> does anybody know things about update-rc.d and initialisation and things like that?
Dec 23 18:23:53 <likewise> jelle: yes, although every distro is a bit different in that respect.
Dec 23 19:16:25 <jelle> likewise: 
Dec 23 19:16:28 <jelle> update-rc.d created a S20dnsmasq and a S20cherokee for me, but only S20cherokee will get started...
Dec 23 19:16:52 <jelle> is that by design?
Dec 23 19:38:42 <likewise> jelle: in what dir are they?"
Dec 23 19:41:50 <jelle> the script itself is in /etc/init.d
Dec 23 19:42:24 <jelle> the links in /etc/rc?.d (? = 0-6)
Dec 23 19:43:59 <jelle> I changed the S20dnsmasq links to S21dnsmasq, which seemed to have solved the problem.
Dec 23 19:44:13 <jelle> (the same for K*)
Dec 23 19:44:22 <jelle> but I felt a little strange
Dec 23 19:46:55 <jelle> there might also be another problem, I saw the led didn't turn green, but was still flashing greem/yellow. maybe I never got into the right runlevel
Dec 23 19:47:06 <jelle> odd
Dec 23 19:55:27 <jelle> how do I find out what runlevel I'm running in?
Dec 23 19:59:42 <likewise> sorry dunno
Dec 23 20:02:28 <HopsNBarley>  jelle: just a sec...
Dec 23 20:08:58 <HopsNBarley> on the first issue, the like stages within a run level - that should work (by looking at the scripts).
Dec 23 20:09:24 <HopsNBarley> i have, for example, two S10's in my rcS.d, and a verbose reboot verifies that both get run.  so something odd is going on.
Dec 23 20:09:41 <HopsNBarley> you could instrument /etc/init.d/rcS, or just work around it as you have  (-;
Dec 23 20:10:00 <HopsNBarley> on the second issue, "last -x" will show the run level changes.  you could grep it out there.
Dec 23 20:10:01 <jelle> right
Dec 23 20:10:27 <HopsNBarley> or, you could write a tiny program that reads the /var/run/utmp records, where runlevel changes are logged, (and read by last)
Dec 23 20:11:51 <jelle> how do I get a verbose boot?
Dec 23 20:12:48 <HopsNBarley> edit /etc/default/rcS - do you have a console?
Dec 23 20:13:06 <jelle> ssh terminal yes
Dec 23 20:14:16 <HopsNBarley> i mean a hard wired terminal.  a console (or at least netconsole).  otherwise ,you have to wait until you log in to do a dmesg.  i haven't tried that 
Dec 23 20:14:36 <jelle> no, nothing like that.
Dec 23 20:14:46 <jelle> hold, I remember one more thing
Dec 23 20:15:43 <jelle> I copied the dnsmasq startup script to create a cherokee startup script, but I think cherokee doesn't go into the background, the way dnsmasq does when you start it up... maybe my runlevel gets stuck by this startup command...
Dec 23 20:16:36 <HopsNBarley> sounds reasonable.  you can always test this *after* you have a sucessful boot without these things.  either the script exits or it doesn't (-;
Dec 23 20:17:34 <HopsNBarley> (he says grossly assuming the halting problem is solvable)
Dec 23 20:17:57 <jelle> I'll look into that
Dec 23 20:18:02 <jelle> last -x gives me this:
Dec 23 20:18:04 <jelle> root@catcher:~# last -x
Dec 23 20:18:04 <jelle> root     pts/0        jellealten.nl    Sat Dec 23 21:43   still logged in   
Dec 23 20:18:04 <jelle> runlevel (to lvl 3)   2.6.16           Sat Dec 23 21:39 - 21:43  (00:03)    
Dec 23 20:18:04 <jelle> reboot   system boot  2.6.16           Sat Dec 23 21:39          (00:03)    
Dec 23 20:18:04 <jelle> wtmp begins Sat Dec 23 21:39:58 2006
Dec 23 20:18:52 <HopsNBarley> i'd say you were in level three, which i also happen to know is the default multi-user mode for slugos.  so all is well.
Dec 23 20:20:24 <jelle> alright, but the amber/green flashing is suspicious
Dec 23 20:20:31 <jelle> I'll check the cherokee startup script
Dec 23 20:20:49 <chacko> Hello :)
Dec 23 20:22:03 <HopsNBarley> hey chacko, did you say yesterday you were bay area?
Dec 23 20:22:22 <chacko> Yes, Sunnyvale
Dec 23 20:22:44 <HopsNBarley> i'm Los Gatos.  hi neighbor!
Dec 23 20:23:52 <chacko> Hello Neighbor!  I have been in the B'Area since '67, and old f.rt :)
Dec 23 20:28:21 <HopsNBarley> chacko, i opened a direct window, see it?
Dec 23 20:30:53 <chacko> HopsNB: already replied there.
Dec 23 20:32:16 <jelle> HopsNBarley: should I use start-stop-daemon to start cherokee? or just do a "cherokee&" ?
Dec 23 20:34:23 <HopsNBarley> chacko, wierd, i don't see it.
Dec 23 20:34:33 <HopsNBarley> jelle, depends on cherokee and i don't know, sorry.
Dec 23 20:35:40 <jelle> I'll dig into cherokee, thanks
Dec 23 20:35:44 <chacko> HopsNBarley: well I selected your direct window, replied, and send another line there too
Dec 23 20:36:32 <HopsNBarley> i started a new one.
Dec 23 20:37:56 <chacko> Does shutdown including powering down work on SlubOS/BE?  I have tried many combinations of parameters without luck :(
Dec 23 20:40:26 <chacko> HopsNBarley: When you started the new win, it came here on the same old win of yours.  I replied to that too... I am using mIRC, only the second session, and have not used irc in almost 15-20 years :(
Dec 23 20:40:58 <HopsNBarley> i've never used mIRC, so i don't know what to tell you.
Dec 23 20:41:06 <HopsNBarley> who do you work for out here?
Dec 23 20:42:55 <chacko> Andy, I have worked for 3 successfule startups Amdahl,Zilog,Altera.  Have been doing independent consulting for the last 20+ years.
Dec 23 20:43:45 <HopsNBarley> nice.  i'm on my 3rd startup.  
Dec 23 20:46:15 <chacko> Andy, know how to powerdown SlugOS/BE?
Dec 23 20:47:35 <HopsNBarley> no sorry.
Dec 23 20:48:26 <HopsNBarley> chacko, lets get lunch sometime!  i'll past my email over in the other window.
Dec 23 20:51:39 <chacko> Andy, I tried to open another chat window to H...NB..., but it is sitting there for acknowledgement ??
Dec 23 20:52:17 <chacko> Andy,  I heard you made a new php-apache version with the accelerator?
Dec 23 20:52:34 <HopsNBarley> 'tis true.  the slug is even "snappy" with gallery.
Dec 23 20:53:14 <chacko> Was that for slugosbe or unslung?
Dec 23 20:54:25 <HopsNBarley> i guess slugosbe - openslug.  i've not been paying much attention to the byte order naming thing.
Dec 23 20:59:02 <chacko> Andy, I was running php-apache on unSlung, and just now got it running on slugosbe just before eno is doing the new slugosbe feed.  How much does you accelerator increase the memory foot print?
Dec 23 20:59:09 <rwhitby> HopsNBarley: want to help out on the nslu2-linux booth at SCALE?
Dec 23 20:59:45 <chacko> SCALE?
Dec 23 21:00:32 <rwhitby> http://www.socallinuxexpo.org/scale5x/exhibitions/nslu2linux.php
Dec 23 21:00:49 <HopsNBarley> since mem is so tight, i did some tuning to apache and php and apc to get it to work without too much swapping.  so i don't have a pure number for you.  apc by default creates a 8M shared mem segment though - as i recall.
Dec 23 21:00:56 <rwhitby> we have a booth
Dec 23 21:01:46 <HopsNBarley> rwhitby, sounds cool!  i'd love to.
Dec 23 21:02:12 <rwhitby> since OpenProtium is based on SlugOS, you'd be welcome to have a box at the booth too.
Dec 23 21:02:58 <chacko> Rod, Thnks. It was at the SanFrancisco booth 1.5 years ago that I first got introudced to slug!!
Dec 23 21:03:17 <HopsNBarley> rwhitby, you going?
Dec 23 21:03:31 <rwhitby> nope, but ka6sox and dyoung will be there
Dec 23 21:05:47 <chacko> Hi ka6sox! R u a ham too?
Dec 23 21:06:02 <ka6sox-office> yep
Dec 23 21:06:38 <ka6sox-office> so who's going to SCALE???
Dec 23 21:06:56 <chacko> 40+ yrs ago I was a ham in India :)
Dec 23 21:07:08 <ka6sox-office> ah..fun
Dec 23 21:07:18 <ka6sox-office> what is the prefix there?
Dec 23 21:07:50 <rwhitby> ka6sox-office: HopsNBarley is going to help us out and demo his OpenProtium stuff
Dec 23 21:08:13 <ka6sox-office> excellent
Dec 23 21:08:57 <chacko> vu2xxx, I had my own initials too nc, there were very few hams there then.
Dec 23 21:09:31 <ka6sox-office> I'm in California hence the KA6
Dec 23 21:09:59 <chacko> I never renewed it, not knowing there was reciprocal, it is now gone to somebodyelse.
Dec 23 21:10:15 <ka6sox-office> bumemr
Dec 23 21:13:08 <chacko> Andy, with your tuned up php/apache, have you compared it to the one in the stream for unslung?  Is your php-apache available as a pkg yet?
Dec 23 21:14:03 <HopsNBarley> yes, modphp and apache2 are the openembedded package names.  i don't know if they are in the unslung package feed or not.
Dec 23 21:15:03 <HopsNBarley> i've not compared them.
Dec 23 21:16:30 <chacko> Does modphp replace the php-apache?  I have apache2 and php-apache running together now --- wondering what it would take to try yours.  Did you change apache2 also?
Dec 23 21:18:54 <dyoung> win 4
Dec 23 21:21:04 <chacko> Andy, I just checked.  There is nothing in the slugosbe feed with php: ipkg list | grep php
Dec 23 21:27:57 <HopsNBarley> chacko, i'm not familiar with the php-apache package.  i added apache2 and modphp to OE so that there would be a cross-compile capable source version.
Dec 23 21:28:20 <HopsNBarley> up until then, apache was a native compile only, and i don't think modphp existed either.
Dec 23 21:29:30 <HopsNBarley> there is nothing too special about my work on apache2 and modphp, other than that they cross compile from x86 to either slugos/be and ppc.
Dec 23 21:37:49 <chacko> Andy, php-apache I think is the official name they have been using for many years, for php compiled as a module for apache.
Dec 23 21:39:05 <chacko> I had used it under that name on RedHat many years ago.  Probably modphp is the same and that is how it is called in some other environments?
Dec 23 21:40:18 <chacko> Andy, there is alos a php standalone pkg, indepenedent of apache2, used for commandline.
Dec 23 21:42:07 <HopsNBarley> chacko, modphp is what i call it   (-;  it's a way in OpenEmbedded to say you want the module only, rather than the command line php utility also... since OpenEmbedded doesn't have USE flags (like gentoo).
Dec 23 21:44:32 <HopsNBarley> the current OpenEmbedded package database does not contain php-apache, so i'm a bit confused on where that is coming from.  it is probably someones kind native build of these tools.
Dec 23 21:46:13 <HopsNBarley> perhaps rwhitby or someone can enlighten us to the archaeology of php-apache?
Dec 23 21:50:07 <rwhitby> php-apache is an Optware package name?
Dec 23 21:50:46 <HopsNBarley> optware == slugos/be ?
Dec 23 21:56:27 <chacko> Andy, was that statically linked?  See:  http://ilia.ws/archives/28-PHPApache-Static-vs-DSO.html
Dec 23 21:57:58 <HopsNBarley> it's a DSO module for apache.
Dec 23 22:12:52 <ka6sox-office> HopsNBarley, you going to need wireless at SCALE?
Dec 23 22:14:52 <HopsNBarley> personally, or for any demo?
Dec 23 22:23:13 <ka6sox-office> demo
Dec 23 22:23:47 <HopsNBarley> i doubt we'll be doing a demo.  
Dec 23 22:24:02 <ka6sox-office> oh..okay
Dec 23 22:26:07 <chacko> So anybody here has any recommendation for how to power down a slugos/be from the commandline?
Dec 23 22:35:50 <ka6sox-office> halt -p doesn't work?
Dec 23 22:35:51 <ka6sox-office> :D
Dec 23 22:36:27 <HopsNBarley> is it a "soft power" issue on the slug?  (much like we were faced with on the storcenter)
Dec 23 22:36:56 <ka6sox-office> there is a chip that "toggles" the button (always runs)
Dec 23 22:37:30 <ka6sox-office> so the cpu can shut it self off iirc...gotta look a the GPIO stuff again.
Dec 23 22:37:33 <HopsNBarley> exactly. you have to tell the chip to turn the power off.
Dec 23 22:38:02 <ka6sox-office> GPIO[8]	W23	Power Off (1 = Turn Off)	R10	Output
Dec 23 22:38:32 <HopsNBarley> nice.  on the storcenter they don't bother to hook any gpio to this.  unless you count the serial port as gpio (-;
Dec 23 22:38:34 <ka6sox-office> I thought this was fixed a long time ago.
Dec 23 22:38:39 <chacko> I do not know enough about the issues.  shutdown -r now, does reboot correctly. but halt -p just behaves like it is hanging?  The lights are still on, no response to command line after that
Dec 23 22:39:00 <HopsNBarley> it's only a few watts  (-;
Dec 23 22:39:05 <ka6sox-office> realyl
Dec 23 22:39:13 <ka6sox-office> okay off to christmas with the family
Dec 23 22:39:34 <ka6sox-office> HopsNBarley, cya @ LAX!
Dec 23 22:39:51 <HopsNBarley> ka6sox-office, yeah, looking forward to meeting some fellow sluggers.
Dec 23 22:39:55 <ka6sox-office> (will be here off and on) mostly in the background.
Dec 23 22:41:27 <ka6sox-office> when I'm doing my job right nobody will notice :D
Dec 23 22:41:51 <chacko> Andy, it is not a power issue.  I need to frequently switch drives during experiments.  halt -p does not give any indication. After that even if you hit the button at the end no respnse, so I have to unplug the power cable, do not feel comfortable about dooing that repeatedly...
Dec 23 22:42:20 <chacko> I meant the button at the front...
Dec 23 22:42:34 <ka6sox-office> Chacko...try pushing the button before you try a halt -p
Dec 23 22:42:41 <ka6sox-office> see if it does a normal shutdown
Dec 23 22:42:54 <ka6sox-office> I thought this was resolved like 2 yrs ago.
Dec 23 22:43:38 <ka6sox-office> (once the gpio's got straightened out.
Dec 23 22:43:38 <HopsNBarley> ka6sox-office, you live in LA?
Dec 23 22:43:47 <ka6sox-office> Santa Barbara
Dec 23 22:43:56 <HopsNBarley> nice.  love the pinot  (-;
Dec 23 22:44:10 <ka6sox-office> I'm a Merlot Man myself
Dec 23 22:44:11 <ka6sox-office> :D
Dec 23 22:45:32 <chacko> ka6sox, I am rebooting now.  Will try your suggestion ...
Dec 23 22:45:38 * ka6sox-office awaits the report.
Dec 23 22:45:43 <ka6sox-office> okay np
Dec 23 22:46:13 <HopsNBarley> ka6sox - office, saturday before christmas?  ()-;l
Dec 23 22:46:58 <ka6sox-office> yep...got a Manual to get done, backups of the nslu2-linux infrastructure to do  and a proposal to write.
Dec 23 22:47:44 <HopsNBarley> who you work for?
Dec 23 22:47:54 <ka6sox-office> myself
Dec 23 22:48:03 <ka6sox-office> I have a slavedriver for a boss.
Dec 23 22:48:09 <chacko> ka6sox, just hitting the pwr button does not give any indication either. Lights remain on.  I think it halted, or is hung, because I cannot ssh into it hafter hitting the button :(
Dec 23 22:48:22 <HopsNBarley> hey, me too!  
Dec 23 22:48:35 <ka6sox-office> so you know the drill
Dec 23 22:48:39 <ka6sox-office> chacko, wait a bit.
Dec 23 22:48:41 <HopsNBarley> chacko, are you hooked into a USB hub which might be supplying power to the slug?
Dec 23 22:48:45 <ka6sox-office> see ifit stops.
Dec 23 22:49:47 <HopsNBarley> i've seen the same thing when wired this way.  it actually halts and removes power, but is still getting power off the +5v rail.
Dec 23 22:50:11 <HopsNBarley> ka6sox-office, consulting?
Dec 23 22:50:24 <ka6sox-office> GPIO[5]	Y25	Power Button (Pulse when state change)	Power Button via flipflop	Input
Dec 23 22:50:34 <ka6sox-office> HopsNBarley, yes..on High Power RF.
Dec 23 22:50:45 <ka6sox-office> (and embedded systems)
Dec 23 22:50:54 <chacko> Andy, no. Before both of these halt -p, or power button, I remove the the second pwr from usb drive.  I am quite aware of the possibility of the pwr going back to slug.  But not so in this case.
Dec 23 22:51:00 <HopsNBarley> cool...
Dec 23 22:51:38 <HopsNBarley> just an idea.  that tripped me up once a ways back...
Dec 23 22:52:16 <chacko> Andy, yes, understand 
Dec 23 22:52:36 <HopsNBarley> i'll see ya'll later.  chacko, love to meet up sometime.  i've got my share of grey.  (-;
Dec 23 22:52:54 <ka6sox-office> me too
Dec 23 22:53:04 <ka6sox-office> we can commiserate
Dec 23 22:53:11 <chacko> See you soon again
Dec 23 22:53:21 <ka6sox-office> time for me to head home.
Dec 23 22:53:46 <ka6sox-office> finished wrapping my wife's pressie...Opteron 165
Dec 23 22:55:02 <chacko> Happy holidays!!
Dec 23 22:55:24 <ka6sox-office> bye
Dec 23 23:34:06 <chacko> Rod, you still there?
Dec 23 23:49:02 <CIA-8> 03pawsa * r4819 10optware/trunk/make/sendmail.mk: sendmail: mark config files as such.
Dec 24 02:28:30 <tuv> i'm having problems transferring files to the slug over nfs. running debian/NSLU2. files copied over nfs have different md5sum from the originals
Dec 24 02:28:45 <tuv> nfs-kernel-server
Dec 24 02:29:26 <tuv> transferring from the slug seems to have no problems
**** ENDING LOGGING AT Sun Dec 24 02:59:59 2006