01:42:25 markd2 (~Snak@r-41.154.alltel.net) has joined #openacs 01:47:49 davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs 01:48:03 [to tils]: yes you can compile the need modules and use 3.4.2 but a few things will be slow 01:58:52 markd2 has quit ("Bork") 02:11:58 davb has quit ("Client Exiting") 02:27:27 tils (tils@B3c76.pppool.de) has joined #openacs 02:30:06 [to shagster] thanks, just read your reply in the log 02:31:14 you mean slow as a bit slower, or horribly slow? 02:36:28 davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs 02:36:49 * davb has to get cracking on the top-secret project 02:39:18 -:- davb [dave@alb-24-58-162-46.nycap.rr.com] has joined #openacs 02:39:18 * davb/#openacs has to get cracking on the top-secret project 02:39:18 [03:40am][tils(+iw)] [#openacs(+n)] 02:39:19 [Lag ??] [O/0 N/12 I/0 V/0 F/0] [U:a:S:b:h] 02:39:19 [#openacs] 02:39:26 aaargh 02:39:40 * tils tears out his middle mouse button 02:41:08 heh 02:43:15 anyone know the secret update the database so locate works command? 02:43:28 updatedb? 02:43:33 aha, thanks! 02:43:48 turbotax seems to have eaten my tax return from 2000 02:43:53 actually it's updatedb and then wait 30min 02:44:09 ah 02:45:02 i think it should be started as root, otherwise you get a lot of permission denied msgs 02:45:26 aha. another learning experience for me :) 02:45:29 baad turbotax 02:45:40 i wish i would get tax returns 02:48:22 argharghargh 02:48:47 * rbm invokes his supernatural powers to understand davb's language 02:49:01 :) 02:50:28 argh 02:51:37 xfce fails again. or progeny/gnome did it better. 02:51:56 X doesn't run the .bash_profile so the environment is all screwed up in an xterm 02:52:04 * rbm hands davb windowmaker 02:52:48 does it magically set the environment? 02:52:50 :) 02:54:49 davb: Is this Debian? 02:55:23 My environment seems to be set just fine. 02:55:44 * rbm leaves 02:55:54 rbm: yep. 02:55:56 oops too late 02:58:30 hmm, but shouldnt bash itself read .bash_profile, as soon as it is started, e.g. the xterm is invoked? 02:59:02 nope, not according the the XFree86 docs 03:00:04 maybe as a dirty workaround use .bashrc then? 03:00:36 looking in my .bashrc file i see that there are a lot of environment settings 03:00:56 yeah, I am sure there is a trick to it, I am just that lazy :) 03:01:27 i remember reading somewhere that the environment stuff just sucks on all unix systems 03:01:37 wow, I am almost getting the hang of this cvs local repository thing. 03:01:38 heh 03:01:57 I never had a problem until I setup my own window manager. 03:02:11 so one should give up trying to make sense of it ... 03:02:14 I think they probably copy the stuff into the window manager startup scripts. 03:02:52 .xession .xinitrc etc... 03:02:54 yeah - but why on earth should the windowmanager be responsible for a user's environment? 03:04:07 ah I did stuff pgsql/bin in the path in .xsession :) 03:07:05 so you have to restart X to make the database work? sounds like you are trying to emulate windows ;-) 03:07:43 well only sortof. I can reload the files I think if I read the docs :) 03:09:50 in a running bash you can say "source filename". sorry if that's obvious 03:10:43 that is handy, thanks 03:10:55 but doing that with .xsession might have unpredictable effects ;-) 03:11:45 could be, but according to the docs I read that is the way to do it. 03:12:39 if it only contains VARNAME=bla then it should be ok 03:15:08 cool thanks. 03:17:36 * tils has to go 03:22:11 bye, thanks! 03:22:23 thanks for helping, not for leaving :) 03:40:29 tils has quit (Read error: 110 (Connection timed out)) 03:50:23 davb has quit ("Client Exiting") 11:47:41 dlk is now known as dlk-MeatingMan 11:57:22 dlk-MeatingMan is now known as dlk-afk 13:57:15 markd2 (~markd2@r-41.112.alltel.net) has joined #openacs 13:57:15 chump has quit (Read error: 104 (Connection reset by peer)) 13:57:34 chump (~chump@alb-24-58-160-41.nycap.rr.com) has joined #openacs 15:12:47 * shagster bangs head against wall 15:12:55 * shagster . o O ( that feels better ) 15:22:53 the poor wall 15:23:06 hopefully it wasn't Larry Wall 15:23:08 he might object 15:23:17 moo 15:23:27 yo pops 15:23:52 how's it going markd2? 15:25:12 going good 15:40:14 davb (~dave@cm-208-136-23-203.nycap.rr.com) has joined #openacs 15:43:25 markd2 has quit ("Bork") 15:54:29 * davb wonders if ETP should use its own table for attributes instead of the generic one. 15:54:53 since I plan on using it for almost every application. 16:07:18 ola (~ola@as3-1-3.mal.s.bonet.se) has joined #openacs 16:11:06 hi ola! 16:13:22 hey davb! 16:13:49 how is it going? 16:14:35 I think I found a bug in acs_object__delete 16:16:08 quite good. I finally managed to get software RAID 1 to work on my system. 16:16:40 how bad is the bug? 16:17:10 wow cool! 16:17:25 the bug makes it impossible to use a custom content_type apparently 16:18:16 now we know why all the aD packages used 'content_revision' incorrectly :) 16:19:00 how weird. I'm going to install qmail and courier IMAP now. 16:19:18 good luck. there are debian packages for those. 16:19:38 well if you use a customer content_type you are not allowed to use the generic storage actually. 16:19:38 did you install from those? 16:19:48 ola: nope I could not get it to work :) 16:22:54 Also I am using vpopmail which you need to compile something special in courier-imap for. 16:23:11 hmm.. that's valuable info. guess I'll go with the source tarballs. the debian daemontols package is not the lates anyway. 16:23:22 s/lates/latest 16:23:34 yeah. qmail is no big deal, it isn't updated too often so no problem :) 16:24:06 oh, I figured something out. 16:24:25 davb: ok. cool. is pop better than imap? 16:24:34 if you use a custom content_type you can have a wide table with one column for each attribute instead of a table with one row for each item/attribute pair 16:24:40 no pop is icky 16:24:51 heh. ok 16:25:05 imap is much more poweful. Except I don't know if you can save the messages say on a latptop that is not connected all the time. 16:25:17 pop is for people who do not have a persistent connection. 16:25:36 I have both setup on my server, you can use pop or imap to the same mailboxes. 16:25:49 I see. thanks! 16:26:36 Hmmm. I wonder if I have say 10 or 15 etp applications with several uniqure attributes each if that would be a problem with a table with 100 columns or something... 16:27:37 uhoh, etp has some stuff hard coded in to use generic storage... 16:29:10 ah... 16:29:15 * davb learns more 16:34:59 I see, it looks like each ETP content application can have its own content_type 16:35:09 which can have its own attribute table. 16:35:25 I think this should be an option instead of using generic storage. 16:35:42 most of my applications will use the extended attributes more than the built in ones. 16:40:04 I wonder if I am trying to make things too complex and I will end up recreating a monster like CMS... 16:40:47 also I wonder if there is really a bug in acs_object__delete still. It works for static_pages.... 16:42:19 which of the etp packages are you going to fork? :-) 16:42:34 oh... static_pages content type refers back to static-pages table. 16:42:49 ola: heh. no forking involved. just getting ahead of luke. 16:42:53 dave: I wish I knew the answers to at least some of your problems;-) 16:43:07 ola: I am just thinking aloud to get my brain working. 16:43:08 cool. 16:43:29 I need to have a site that works. so I will make it work, then make it pretty when I get the chance. 16:43:30 the way infiniteinfo uses different "instances" of one and the same application for the purpose of enabling different templates is quite cool, I think. 16:43:45 ola: ? 16:43:53 I am not sure what you mean... 16:44:11 to the demo 16:44:12 ! 16:45:45 yes... I think they use copies of one self made application and point them to different templates... 16:46:03 ah spiffy. I need to get in touch with him. 16:46:28 I see that you can change templates. 16:47:18 dunno how they implemented it, though. just guessing... 16:48:39 also I see a new feature, upload logo and image for a section... 16:48:42 shouldn't the user get to change what app to be shown? 16:48:54 nifty. 16:49:14 you can upload files to any section. I am not sure how they are tied to the templates though. 16:49:19 jim (~jim@12-233-187-5.client.attbi.com) has joined #openacs 16:50:56 aha, he has it built in section_image is the name of the section_image. 16:53:08 hi... 16:53:12 hi jim 16:53:25 gonna install anothher openacs 4... 16:53:37 I got last nite's nightly tarball 16:53:40 postgresql or oracle? 16:53:45 postgres 16:53:55 I am planning on trying out the cvs from last night.... 16:53:56 trying it on a woody box 16:55:49 (debian pre-3.0 box, that is) 16:55:58 cool. 16:56:17 havent' decided if I'm compiling everything or not 16:56:24 hi jim 16:56:28 hi ola 16:56:32 err, 16:56:36 ola, ola :) 16:56:43 does ns_xml come with any debian package? 16:56:48 ola jim! 16:57:22 I recommened keeping the postgresql source around if you want to use openfts. 16:58:03 anyone know what not equals is in pl/pgsql? 16:58:17 jim: my name is pronounced more like: ooola 16:58:27 oh :) 16:58:41 how many ooos are in that :) 16:59:01 enough. :-) 16:59:02 aha != 16:59:13 we could add another la at the end :) 17:00:02 but of course that would be silly... not a good idea to mess around with someone's name too much :) 17:00:09 time to reload and test my data model changes :) 17:00:40 anyway, I'm trying to find those install docs for oacs 4 17:01:37 the new one? 17:02:07 yeah, is there a whole doc tree for oacs4 yet? 17:02:28 sure, its in the acs-core-docs package 17:02:40 that is the core-docs. each package hsa a www/doc directory also. 17:03:42 I don't think the new install guide is in there though... 17:03:59 jim, try 17:04:10 http://kurup.com/acs/openacs4/ 17:04:11 A: http://kurup.com/acs/openacs4/ from ola 17:04:18 thanks ola. 17:05:28 yes, thank you :) 17:06:13 ola: also I want to hook etp up with workflow. it wasn'y my idea orginally, but i like it :) 17:07:01 * davb vacuums 17:07:43 oops 17:07:58 davb: very cool. I think Don made some enhancements to workflow recently. 17:09:08 weird, my home page is broken... 17:11:34 (cough) it's been for some time. (on NN 4.77 Linux anyway) ;-) 17:11:55 workflow is the example to follow for documentation 17:12:07 ola: heh... I don't support obselete browsers :) 17:12:14 I also though I had fixed it... 17:13:47 its some sort of style sheet issue. 17:14:52 davb: it's more than obsolete, IMO. a piece of crap is an understatement. 17:15:04 it's a poor excuse for a browser... 17:15:20 ok. I'll stop. 17:15:22 I really like mozilla, except the mail is still flakey and slow 17:15:44 ok, why is my new install trying to run an oracle query... 17:16:37 davb: have you used Opera? 17:17:03 yeah. its ok, but I like mozilla better. I can't stand what it does when pop up windows appear. 17:17:32 What? Opera? 17:18:04 yeah. If a small window opens, it chnages the size of all the windows, very annoying. 17:18:06 hi rbm. 17:18:08 I like Mozilla, except that it takes 40 Mb of RAM and 15 seconds to render a page. If a page has any animated gifs, it takes 90% of your CPU. 17:18:15 opera is way fast! 17:18:17 davb: You can change that behaviour. 17:18:36 rbm: aha. cool. maybe I'll try it again. 17:18:51 Mozilla takes 200 Mb of disk space. Opera takes 5 or 6 with Qt _statically_ compiled. 17:18:59 hi ola 17:19:07 Mozilla is fast enough though at least 0.9.7 before that it was practically unusable. 17:19:40 opera doesn't support java (applets), does it? 17:19:58 I think the main reason mozilla is slow is because of the dynamic rendering thing. To enable theming crap they've compromised regular speed. 17:20:19 ah, could be. 17:20:21 ola: I just pointed opera to my netscape plugins directory and everything works fine. 17:21:48 anyone here used ino_chart? 17:22:12 not me. 17:22:40 the tcl graph program? nope. 17:22:41 I'm trying to modify it a bit, but not very successfully apparently. 17:22:53 ola: It's actually a module for AOLserver 17:23:11 oh. kewl. 17:23:21 kick butt! apt-listchanges is working for me now! 17:24:58 woo hoo! 17:35:58 hmmmm... all the existing ACS code assumes if you use a custom content_type then you MUST use a custom table for attributes. 17:36:14 this is contradictary to the documentation of course... :) 17:38:27 davb: Submit a patch! 17:38:32 Z) 17:38:36 s/Z/:/ 17:40:52 :) I am working on it. it is pervasive throughout the toolkit 17:42:08 I am just astounded how all the packages just used content_revision instead of fixing this. 17:46:20 hmmm. should we just require a table for a custom content type? 17:46:32 would an extra 10 or 15 tables hurt anything? 17:48:02 oh wait, nm. I read the docs wrong... 17:48:20 it says this: "design(# 'My Item', 17:48:23 oops 17:48:58 aha. the CR is requiring the specific table, not acs-objects. 17:49:03 ack, my head hurts... 17:58:38 tils (tils@port-212-202-128-208.reverse.qsc.de) has joined #openacs 18:07:49 oops. 18:08:16 table_name is supposed to be optional in acs_object_types, but it is not... 18:10:10 ok, this is just pointless. I do not have time to rewrite pratically every query, table, function, etc relating to object types. 18:10:29 it just needs its own table and that is that :) 18:36:10 dlk-afk is now known as dlk 18:58:34 higgins (~rolf@above.badgertronics.com) has joined #openacs 18:58:59 hello higgins 18:58:59 i'm trying to connect to postgres from my machine on 192.168.0.6 18:59:04 hello davb. 18:59:11 but postgres won't let me. 18:59:15 i'm running it with -i 18:59:20 on the same machine or a different one? 18:59:23 and pg_hba.conf has a line in it 18:59:27 on a different machine 18:59:38 postgres is on 192.168.0.4 18:59:43 ok 18:59:51 i have this set up in pg_hba.conf: 19:00:16 host all 192.168.0.0 255.255.255.255 trust 19:00:45 ARGH my netmask is wrong. 19:00:52 un momento. 19:02:36 * higgins lames 19:02:40 that was the problem! 19:02:46 cool. 19:02:53 thanks for your support. 19:02:56 np 19:03:01 higgins has quit (Client Quit) 19:03:24 higgins continues the record of the #openacs channel helping the asker answer his own question 19:04:07 just being a clear channel helps with that :) 19:04:37 ah. 19:22:40 ok. I get this error when I compile ucspi-tcp: 19:22:40 tcpserver.c:1: sys/types.h: No such file or directory 19:22:40 tcpserver.c:2: sys/param.h: No such file or directory 19:22:40 tcpserver.c:3: netdb.h: No such file or directory 19:22:40 make: *** [tcpserver.o] Error 1 19:22:40 hal:/usr/local/src/ucspi-tcp-0.88# 19:23:07 I bet davb or rbm recognize it... 19:25:16 ola: sorry, I never have had an error with any of djb's stuff :) 19:25:58 looks like you're in the wrong dir 19:26:05 s/^/ola: / 19:26:44 what? 19:28:11 in tcpserver.h, what does the include line that includes sys/types.h look like? 19:28:19 I meant tcpserver.c 19:31:11 the three first lines look different from the rest... 19:31:38 So you don't have the library that defines sys/types.h 19:31:59 no, I guess not. 19:32:36 I should come with the tarball, right? 19:32:38 vinod (~vinod@207-172-255-207.s1223.apx1.sbo.ma.dialup.rcn.com) has joined #openacs 19:32:40 Mine is on /usr/includes/sys/types.h 19:32:46 yeah me too. 19:32:48 s/I/it 19:32:49 hi vinod 19:32:52 Let me find out which debian package has it 19:33:06 hey davb, rbm, ola! 19:33:10 hey vinos 19:33:17 s/vinos/vinod/ 19:33:28 hey vinod! 19:35:57 ola: You need libc6-dev 19:36:23 aha! so the files between < and > are supposed to reside on the computer already? 19:36:31 thanks rbm! 19:37:17 Between < > means that the C pre processor will look in the standard include locations. #include "foo.h" will look for foo.h in the current directory 19:38:22 I see. well I learn something new every day. 19:42:41 and if you -I dirname on the gcc command line, you can add to the standard include locations 19:43:57 (which you might want to, for the X includes, or postgres includes, etc 19:51:44 jim: you mean like this: make -I /path/to/includefile.h ? 19:52:12 ola: more like this: 19:52:34 CFLAGS="-I /path/to/includefile.h" make 19:52:57 note I'm doing a shell assignment then running make 19:53:46 ah. thanks. I'll try to remember all these tricks. 19:53:52 reason, make has all these rules builtin, one of them says how to make a .o (object) file from a .c file 19:53:52 :-) 19:54:31 oh, don't remember that one; it's better to have a makefile and edit it 19:55:02 right 19:55:45 projectwise, if you're working on your own stuff, you'd have a directory for your project, and inside, a Makefile and some .c and .h files 19:56:28 the Makefile knows how to bring all the things in the directory uptodate (building what needs to be built, etc) 19:58:07 the rule that knows how to build a .o from a .c, uses a gcc command line... it automatically puts whatever is in CFLAGS variable in there 19:59:23 it's expected to be command line switches, like -g for make debugging convenient, or -O for optimize the code 19:59:47 anyway, those are fed to gcc by this make rule3 20:02:04 but if you don't want to use make, if you want to just use gcc directly or in a script, 20:02:11 say you had foo.c 20:02:15 you could say 20:03:03 gcc -I /path/to/includefile.h foo.c -o foo 20:03:28 err, small thing to clean up there...