00:03:32 denshi: it's ok if you have to dig into dissertations to prove your ideas unoriginal. now, when you start seeing your "original" ides being implemented on msft's front page, well ... 00:03:42 rolf 00:03:51 i mean, rotflmao 00:04:01 no, you mean rolf 00:04:09 Rolf speaks to us through the subconcious of others 00:04:16 Give In to your Inner Rolf 00:04:16 haha 00:04:17 and gitdown! 00:04:17 gitfunky 00:04:36 you might be right... I work near wolf street, and every time I see the sign I read it as 'rzolf'. 00:05:25 http://www.theonion.com/onion3539/national_funk_congress.html 00:05:26 A: http://www.theonion.com/onion3539/national_funk_congress.html from denshi 00:05:43 so, does the onion write serious stuff on apirl 1? 00:05:52 A: National Funk Congress deadlocked on Get Up/Get Down Issue 00:05:52 commented item A 00:05:54 haha 00:06:37 A: "Until our country's funky leaders can resolve this deadlock, U.S. funk leadership, and the booties of all Americans, will remain immobilized," 00:06:37 commented item A 00:34:26 markd2 has quit ("Ni Pop Nooo!!!!!") 00:36:26 davb has quit (Remote closed the connection) 00:36:46 davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs 00:36:55 don't restart gpm while X is running 00:37:53 argh. 00:38:08 I though I had fixed the bash_profile not executed in X problem 00:38:27 When I ran gnome it used to "just work" 00:38:35 night guys 00:38:36 denshi has quit () 00:38:39 bye 00:38:52 anyone loads tcl modules in AOLserver here? 00:39:00 I have done it before. 00:39:07 tcl modules written in tcl? 00:39:22 Yes 00:39:36 I want it so I don't have to set tcl_pkgPath, etc. in every script. 00:39:50 you have to put the files in a directory under the tcllibdir 00:40:10 then do in the modules section of the config file ns_param module packagename tcl 00:40:13 I thought that if I included the "global tcl_pkgPath", etc. in $AOLSERVERDIR/modules/tcl/init.tcl that it'd get taken care of 00:40:28 where packagename is the directory name where you stashed the files. 00:40:36 you might have to rename them so they load in order. 00:40:38 I don't want the modules to be loaded in _every_ new thread. I want to be able to do "package require foo" 00:40:47 oh. 00:41:21 your idea should work for tcl files, but adp files need that tcl_pkgPath in every file if I am not mistaken. 00:41:29 oh really? 00:41:42 I'm using ACS 4.2 (classic) 00:42:00 either that or its the other way around. I poked around in the aolserver source, ADP and Tcl are setup different when an interp is called. 00:43:25 there was a discussion about it in the aolserver list a while back, before I converted tclxmlrpc to use nsxml. 00:43:47 markd2 (~markd2@r-41.194.alltel.net) has joined #openacs 00:44:55 brb 00:44:58 davb has quit ("Client Exiting") 00:45:20 davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs 00:45:35 aha. now I remember, I put the stuff in the xsession file and it loads. 00:45:51 hmm? 00:46:41 So I have to source init.tcl in every adp? 00:46:57 not quite... 00:47:02 let me look 00:47:49 Looks like setting up tcl_pkgPath in init.tcl doesn't work from an ADP. It can't find the var. 00:48:10 right, you need to do a little trick. 00:49:22 set tcl_library [file join $tcl_pkgPath tcl${tcl_version}] 00:49:22 source [file join $tcl_library init.tcl] 00:49:35 heh, that is an a message you repsonded to on the list :) 00:50:28 doing "global tcl_pkgPath; global tcl_library" makes the var available, but still no package require 00:50:46 ack. yeah you do need to source that init.tcl 00:50:48 davb: I know that, but I don't think I tried with package require when I posted that 00:50:57 davb: what's the url? 00:50:57 package require will work with that. 00:51:12 http://listserv.aol.com/cgi-bin/wa?A2=ind0006&L=aolserver&P=R8022&X=6221793B1E69696698&Y=dave@deepskydesign.com 00:51:12 B: http://listserv.aol.com/cgi-bin/wa?A2=ind0006&L=aolserver&P=R8022&X=6221793B1E69696698&Y=dave@deepskydesign.com from davb 00:51:15 ick 00:51:30 I have done it. 00:51:41 I really wish they would just make it work correctly. maybe in 4.0 00:52:51 davb: You need the Community Uber-Search: 00:52:52 http://michael.cleverly.com/aolserver/search/ 00:52:52 C: http://michael.cleverly.com/aolserver/search/ from rbm 00:53:11 that indexes the aolserver list? cool. 00:53:12 davb: you found the 404 problem? 00:53:12 C:|Michael Cleverly's AOLserver/OpenACS Community Uber-Search 00:53:13 titled item C 00:54:23 jim. testing now. 00:54:51 So how come 'package require' still fails for me? 00:55:32 hmmm. 00:55:34 not sure. 00:56:36 aolserver somehow breaks tcl by not setting the environment correctly when it creates and interpreter 00:57:35 jim: using mozilla, when I click on the offending page (not a 404 error) it says connection refused almost instantly. 00:58:06 davb: interesting... 00:58:17 what shell are you using? 00:58:19 bash? 00:58:21 the log shows a bunch of queries then bang it stops, no errors or anything. 00:58:23 yeah. 00:58:39 I dunno even how to begin figuring this out... 00:58:48 Ack! doing the same thing in tclsh works! 00:59:04 davb: can we take a right turn for a sec/ 00:59:05 ? 00:59:26 sure, but I have to be careful I don't crash into the wife... not alot of room in this office. 00:59:36 oh :) 00:59:48 ok :) 00:59:59 jim: Did you see that I was assigned an AM? 01:00:11 We even exchanged a couple mails already 01:00:23 davb: umm, earlier when you downloaded aolserver from me, you registered on my site... yes? 01:00:32 rbm: excellent! 01:00:43 how'd they go? 01:00:56 jim: pretty good so far. 01:01:01 that is correct. 01:01:14 ok, did I require an email? 01:01:29 no. it just let me right in. 01:01:44 ok, sheesh... lemme check my settings 01:03:39 * rbm ponders why package require doesn't work from his .adp's 01:04:48 no error message about a missing variable? 01:05:14 hmm, ok... 01:06:09 can't find package TGDChart 01:06:09 while executing 01:06:09 "package require TGDChart" 01:06:09 invoked from within 01:06:31 where did you put the package? 01:06:53 $AOLSERVER/lib 01:07:36 I think it actually looks in the tcl package directory /usr/lib/tcl or something 01:08:01 davb: It looks in $tcl_pkgPath/$tcl_library 01:08:30 I wish I could remember how I got it to work. 01:11:42 * rbm tries something else 01:12:11 rbm: I'm trying to test my site; could you try registering? 01:12:39 hey: "I've been working on a WebDAV module for AOLserver all night long." 01:14:04 davb: who's that? 01:14:04 rumour has it that is not true 01:14:29 Dossy, one of the people still hanging around on the aolserver list. 01:16:03 * davb emails 01:22:43 vinod has quit (Read error: 104 (Connection reset by peer)) 01:23:54 darn 01:24:07 emacs/gdb isn't finding the source files anymore... 01:25:09 davb: must compile each .c file with -g 01:25:53 ah. it "just worked" with aolserver 3.3 01:26:37 darn gvd debian pacakge is broken 01:31:00 davb: if the sources are moved around after compiling against them with -g, the symbols that get added to the binary no longer refer to the right dir tree 01:31:13 ddd is useless, it displays white text on a white background 01:31:48 davb: if you strip the binary, that gets rid of everything that -g put there plus some 01:32:02 ok. 01:32:34 davb: finally, if you subsequently compile without -g, then the binary won't refer to any source 01:32:37 at aol we ran aolserver -g 01:32:46 makes diagnosing production problems much much easier 01:32:49 brb 01:33:18 markd2: so then the source was locked into a certain relative dir tree position? 01:33:35 nah 01:33:46 if we needed source, we could load it then symlink as necessary 01:33:53 or get a corefile from the prod machine and look at it on a dev machine 01:33:53 ahh 01:34:05 sometimes just the stack trace was necessary 01:34:39 did you guys use plain gdb? another debugger? front ends? 01:34:46 depends on the platform 01:34:48 usually gdb 01:34:51 'wildebeest' for hp-ux 01:35:08 dbx was preferred on irix, and digital unix had 'ladebug', which was wonderful for threaded apps 01:35:21 cute 01:37:19 markd2: could you do me a favor and try to register on 12.233.187.5:9000? 01:39:25 jim: request error on http://12.233.187.5:9000/register/user-new-2 01:39:46 right as it would have sent out the email... 01:39:49 ok 01:40:50 is that an option I have to set in the makefile? 01:41:59 yeah... and it might already be set, but check... usually, you would see something like CFLAGS = "-g -I/some/where -I. -O" or the like 01:42:09 ok 01:42:48 and if CFLAGS doesnt' have -g, you would add it... also, while compiling, make sure each line that compiles the file has the -g 01:43:04 ok. 01:43:27 jim: sure 01:43:59 (that would be a good acceptance test for "adding debug support to aolserver") 01:44:16 request error 01:44:44 one sec, still determining problem 01:45:08 * rbm goes home 01:50:16 hmmm gbd with aolserver only works with an -x gdbinitfile command line. 01:54:53 thought so. qmail error. 02:02:26 where does acs store a user's email addr? 02:03:55 parties? 02:04:14 hmmm. 02:05:34 darn, gvd is not picking up ld_library_path from the environment 02:06:48 nm, 02:07:05 gotta upcase the var name 02:07:16 ahh 02:07:19 parties 02:07:56 right, i needed a set environment in gdb 02:10:01 oops, crashed the debugger 02:10:22 congratulations :) you just learned C :) 02:10:32 heh 02:13:21 gvd is written in Ada 02:13:38 but it uses gtk, which is C 02:15:16 now its running. 02:19:09 ok, anyone got a different email address to try registering again? 02:19:40 deleting a user from the database would be a major pain atm 02:20:34 I will. 02:20:52 cool :) thanks 02:21:27 it says email sent 02:21:34 nice :) 02:21:39 check your email :) 02:21:41 email received 02:21:56 nice :) 02:22:16 uhoh 02:22:39 I went to the link, and clicked continue. entered my password and clicked login and got this error: 02:22:41 Problem with login 02:22:41 There was a problem authenticating the account: 3138. Most likely, the database contains users with no user_state 02:22:54 URL : http://12-233-187-5.client.attbi.com:9000/register/user-login 02:22:56 hmm... 02:23:22 oh, see how the host name changed? 02:23:48 apparently my state has not been updated. 02:25:32 might have to build the site again from scratch... can't right now; gotta go 02:25:51 thank you all for your help 02:26:12 np 02:35:06 this is really frustrating. apparently everyone else who has had success with dotlrn does not have the aolserver problem I have. 02:36:38 * davb tests the only unknown nsxml with xslt support 02:48:36 nope, that didn't help. 03:02:58 I have pinpointed the error somewhat. it happens during the parsing of an adp file. 03:10:50 davb has quit ("Client Exiting") 03:54:34 markd2 has quit ("Bork") 04:55:22 skel (~areplogle@ip68-3-17-218.ph.ph.cox.net) has joined #openacs 04:55:38 :) 04:55:40 hello 04:56:31 hello 04:56:32 que tal, rbm 04:56:41 paje: how are you paje? 04:56:41 rbm: i haven't a clue 04:56:51 paje: slashdot 04:56:53 Slashdot - Updated 2002-04-02 04:21:55 | Distributed Computing Program Hidden in Kazaa | Konqueror's Javascript Continues To Improve | April Fools Wrap Up | Blizzard removes Orcs from Warcraft III 04:57:16 could anyone enlighten me on how to use postgresql to execute multiple 'where' statements in a select? 04:57:50 example. select field from table where table1 = this and table2 = this 04:58:15 or is that possible? 04:59:14 wait, I mean select field from table where field1 = this and field2 = this 04:59:22 not tables.. 05:02:03 rbm has quit (carter.openprojects.net irc.openprojects.net) 05:02:34 rbm (rmello@fslc.usu.edu) has joined #openacs 05:10:39 vinod (~vinod@207-172-216-238.s1254.apx1.sbo.ma.dialup.rcn.com) has joined #openacs 05:14:14 skel has quit ("BitchX: no this is NOT a cybersex client") 06:10:46 vinod has quit ("changing universes") 07:25:13 buenas notches para todos 12:08:28 k2pts (~nkd@213.149.180.151) has joined #openacs 12:08:33 k2pts has left #openacs 12:34:13 <_djg_> _djg_ (~dirk@212.84.246.68) has joined #openacs 13:13:39 Psychephylax (~nickb@pool197.eng.cv.net) has joined #openacs 13:34:09 davb (~dave@rrcs-nys-24-97-22-204.biz.rr.com) has joined #openacs 13:44:24 http://www.oreillynet.com/pub/a/webservices/2002/04/01/outlining.html 13:44:24 D: http://www.oreillynet.com/pub/a/webservices/2002/04/01/outlining.html from davb 13:44:32 D:|Jon Udell on Instant Outlining 13:44:32 titled item D 13:45:24 D: Instant Outlining is subscribing to an outline file maintained by another person. I am not sure yet how notification of changes occurs, or where old items fall off to. 13:45:24 commented item D 13:54:43 D: http://radio.outliners.com/beta 13:54:43 commented item D 13:55:07 D: it decides if an outline is updated by comparing the size. Why not the last-modified header? I have no clue, that makes much more sense. 13:55:07 commented item D 13:55:55 D: it would be interesting to have OpenACS output outlines of the site or user activities as outlines in OPML format 13:55:56 commented item D 13:56:22 D: except we don't have anything remotely approaching a user interface that would work for this. 13:56:22 commented item D 13:56:37 D: maybe a little Mozilla XUL thing would work. 13:56:38 commented item D 14:26:36 hmmmm, I can't get the datamodel to load at all with postgresql 7.2 14:28:30 markd2 (~markd2@r-41.182.alltel.net) has joined #openacs 14:30:06 weird, there is nothing in the databse at all... 14:32:46 my god.... it's full of cars... 14:32:57 * davb misses the reference 14:33:11 actually I was pointing at the nonexistent old pgsql install dir. 14:33:18 now its not sourcing the tcl files. 14:35:37 ever seen the 2001 movie? 14:35:46 ah. 14:37:21 oops, I deleted the database and forgot to recreate it. 14:37:41 ah, there it goes. 14:38:01 I am trying to recreate my crashing aolserver with edit-this-page 2.0 on postgresql. 14:41:43 * davb notes that the data model runs about twice as fast on postgresql rather than oracle 14:43:51 yeah, it crashes. 14:43:56 now to figure out why. 14:44:37 ok, it is at the exact same line, but the parameters have different values 14:48:53 ok weird. 14:49:12 it seems to be a problem with ns_log 14:49:41 the request processor is writing to the log and something in running a regex on the contents of the string 15:01:04 hmmm, the dotlrn one is parsing an adp file, and is re'ing a tag 15:01:41 doesn't seem like the kind of thing that crashes a program 15:04:10 yeah 15:04:16 crashes, as in physically nuking the server? 15:04:38 yes. it segfaults, and all nsd processes stop 15:05:48 I notice that the memory address for the re contains the same thing on both crashes. 15:07:15 can I save a backtrace from gdb? 15:07:45 cut-n-paste to another file :-) 15:08:08 ok :) 15:08:15 should have run it in emacs. 15:08:51 everything is better run in emacs 15:09:08 I like this, C programming and debugging: trial by fire 15:09:29 aigh, bt full is very descriptive 15:14:31 hmm.. 've never used bt full. just bt / where 15:15:39 you don't want to. I got 10 pages of stuff just for the first command :) 15:16:11 heh 15:20:40 markd2 has quit ("Bork") 15:53:11 http://backend.userland.com/publishSubscribeWalkthrough 15:53:11 E: http://backend.userland.com/publishSubscribeWalkthrough from davb 15:53:31 Jackal (~chatzilla@ECOLOGY.MIT.EDU) has joined #openacs 15:53:49 hello all 15:54:00 can someone help me to get nsxml installed? 15:54:05 hi, sure 15:54:20 i'm running rh 6.2 15:54:49 ok. what have you done so far? 15:55:26 i;ve downloaded nsxml 15:55:37 put in next to my aolserver folder 15:55:41 tried a make 15:55:49 get nasty errors 15:55:55 [root@lfee nsxml]# make 15:56:24 hmm. won't let me paste a big chunk 15:56:38 gcc -g -I../aolserver/include -D_REENTRANT -pipe -fPIC -Wall -Wno-unused -DHAVE_CMMSG=1 -DUSE_FIONREAD=1 -DHAVE_COND_EINTR=1 -c -o ns_xml.o ns_xml.c 15:56:43 nsxml should be at the same level at the other ns modules, nscache, nerewrite etc... 15:56:52 ns_xml.c:40: libxml/parser.h: No such file or directory 15:57:00 do you have libxml2-devel installed? 15:57:41 I do but its a XImian version 15:57:48 same for libxml 15:58:05 should be ok. 15:58:10 ok good 15:58:22 so I have a nsxml folder afer untarring 15:58:24 but it should be able to find it. 15:58:30 where should I put this? 15:58:50 in the aolserver source tree at the same level as aolserver/nscache/nsrewrute etc... 15:59:53 ok let me look 16:00:18 actually I think that I did the binary aD installation 16:00:26 I need to recompile? 16:00:33 ummm compile? 16:00:46 you need the aolserver source to compile nsxml also. 16:01:15 ok. thanks. I'll try that out. 16:01:38 do you generally use the aolserver.com version or some other variation? 16:02:13 nsxml? 16:02:19 get it at acs-misc.sourceforge.ent 16:02:22 get it at acs-misc.sourceforge.net 16:02:48 no - i mean do you download the aolserver from aolserver.com or use other versions like the openacs or aD dists? 16:03:02 or dl a binary of the whole thing at: http://uptime.openacs.org/aolserver-openacs/ (no guarantees :) 16:03:13 I use aolserver 3.3+ad13 mostly. 16:03:27 frm arsdigita.com, but 3.4.2 should also work. 16:05:11 ok - i think I'll stick with aD 16:05:23 do the recent ones include nsxml? 16:05:36 no. 16:06:02 they also don't inlcude a postgresql driver. 16:06:55 also this source includes everything you need: http://uptime.openacs.org/aolserver-openacs/aolserver3.4.2-oacs1-src.tar.gz 16:07:00 I installed it yesterday. 16:07:02 actually i'm still using oracle | acs 16:07:06 ah. 16:07:26 you might want to stick with aD's version then. 16:07:31 I'm a bit worried about making the transition. 16:07:37 but it is inevitable 16:08:07 yes. well, ybos.net has an updated ACS Tcl that is closer to the aD version. 16:08:16 cool 16:08:45 where can I get this? 16:10:36 Psychephylax has quit (Read error: 110 (Connection timed out)) 16:15:10 markd2 (~markd2@r-41.182.alltel.net) has joined #openacs 16:17:03 which this? 16:18:05 ybos.net version of ACS Tcl 16:18:19 not entirely clear on ybos.net site 16:19:07 here it is: http://developer.ybos.net/ 16:19:17 ok cool thanks alot 16:19:52 np. 16:20:00 * davb hopes talli doesn't find out 16:20:24 :) 16:20:33 :) 16:22:58 jkhong (~jkhong@202.151.220.11) has joined #openacs 16:25:31 bbl 16:25:34 davb has quit () 16:39:50 vinod (~vinod@216-164-248-232.s3216.apx1.sbo.ma.dialup.rcn.com) has joined #openacs 16:46:17 markd2 has quit ("exercise") 16:49:58 Hi everyone! Has anyone found a good ide that has strong support for tcl and sql (and stuff for OpenACS devel?) 16:50:55 hi jkhong 16:50:57 emacs? 16:50:59 i use emacs 16:50:59 vinod! 16:51:06 hey rbm! 16:51:31 hi vinod, rbm! Is emacs seriously THAT good? 16:51:38 emacs is great. 16:51:53 I've managed to escape learning emacs for years... Don't know if I can face it now. 16:51:59 haha 16:52:44 i've spent most of my life on macintoshes, so the power of emacs was a siren song for me :-) 16:54:01 but seriously, it does everything i need for the level of openacs dev that i do 16:54:06 I've tried Komodo -- nice syntax highlighting, MS-Word -like underlining of syntax errors for tcl, etc. But SLOW... and easier to learn than emacs ;) 16:55:58 One thing I'd like emacs to do (and it possibly does, but I don't know how to set it) is command completion. 16:56:45 rbm: yeah, that would be nice 16:57:38 komodo does it, though I don't find it very useful. Maybe coz it takes longer for it to suggest completion than it does for me to type it in. 16:57:58 (Looking at emacs websites now -- scary thought! :) 16:59:52 emacs does syntax highlighting (although it gets disabled in plgpsql since PG stored procs are within single quotes) 17:00:17 it doesn't do underlining/highlighting of syntax errors 17:01:32 I found a gui-xemacs at http://www.cs.indiana.edu/elisp/gui-xemacs/ !! It's still nothing near a cozy, eye-candy ide though. 17:03:08 The underlining of syntax errors is good for spotting typos without having to run the script. *VERY* helpful when I was learning tcl. 17:10:54 markd2 (~Snak@r-41.15.alltel.net) has joined #openacs 17:12:05 davb (~dave@rrcs-nys-24-97-22-204.biz.rr.com) has joined #openacs 17:17:30 hello! I'm trying to install acs with modules... 17:17:40 can somebody answer some questions? 17:18:52 acs or openacs? 17:19:17 actually sorry - I meant to say aolserver 17:19:35 I am trying to compile aolserver with modules 17:19:39 What's the problem you're facing? 17:19:45 I downloaded the src from aolserver.com 17:19:45 denshi (~chatzilla@adsl-216-62-223-193.dsl.rcsntx.swbell.net) has joined #openacs 17:20:13 I downloaded modules from arsdigita.com (XML modules, cache, oracle driver) 17:20:31 I save them in the aolserver source folder 17:20:45 I made wach one with "make" 17:20:54 *each 17:21:05 and a then made aolserver and make install 17:21:19 after adding the modules in the makefile. 17:21:25 am I doing something wrong? 17:21:32 I just do make install from each module directory. 17:21:53 don't use nsxml from aD.com 17:21:57 it is way old. 17:22:08 before or after I make aolserver? 17:22:16 oh ok 17:22:18 after. 17:22:29 make install for aolserver will put it in /usr/local/aolserver 17:22:38 then make install for each module copies it to /usr/local/aolserver/bin 17:22:47 unless you specify another target 17:23:08 yeah - ichanged the target 17:24:04 gcc -I..//include -D_REENTRANT=1 -DNDEBUG=1 -g -fPIC -Wall -Wno-unused -mcpu=i686 -DHAVE_CMMSG=1 -DUSE_FIONREAD=1 -DHAVE_COND_EINTR=1 -c -o ns_xml.o ns_xml.c 17:24:12 ns_xml.c:40: libxml/parser.h: No such file or directory 17:24:28 thats what I get when I try the nsxml from xmlsoft.org 17:24:30 should be libxml2? 17:24:45 make sure you have libxml2 not libxml 17:24:52 the nsxml makefile is very distro specific... 17:25:01 hi vinod. 17:25:02 oic 17:25:07 hey davb 17:25:10 another option would be to use the ready-made distro that Mat Kovach built 17:25:20 that is what I said :) 17:25:27 very handy. 17:25:48 http://uptime.openacs.org/aolserver-openacs/ 17:25:48 F: http://uptime.openacs.org/aolserver-openacs/ from vinod 17:25:51 I'm trying to actually come to an understanfing of how to get aolserver working with modules. 17:26:10 F:| Mat's awesome AOLServer distro specifically for OpenACS4 17:26:10 titled item F 17:26:19 directions on aolserver.com suck 17:26:26 Jackal: that's a good plan too 17:26:31 I've been wanting to autoconf'ize AOLserver for the longest time 17:26:42 do the other modules compile (besides nsxml)? 17:26:59 Makefile for nsxml says libxml2 not lbxml 17:27:05 install libxml2 17:27:34 yeah oracle-driver, permissions, cache and maybe one of the virtual hosting modules 17:28:01 I also want to know that I can reproduce this when I need another module 17:28:46 ok. then you need 2 steps. 1) install libxml2 (test this by doing 'xml2 --version') -should get 2.4.16 or something 17:28:51 ns_xml.c says include should I have to edit? 17:29:30 [root@lfee /tmp]# rpm -qa | grep libxml 17:29:38 libxml10-1.0.0-2 17:29:46 libxml2-2.4.12-1 17:29:52 libxml-devel-1.8.15-ximian.2 17:29:53 most modules will work fine. nsxml needs a fancier makefile. 17:30:01 libxml-1.8.15-ximian.2 17:30:19 do you have libxml2-devel? 17:30:21 can you do 'xml2-config --version'? 17:30:42 * rbm hugs Debian 17:30:50 yeah libxml-devel-1.8.15-ximian.2 17:31:02 Jackal: that's libxml1 not libxml2 17:31:07 can;t do libxml-devel-1.8.15-ximian.2 17:31:29 not sure i understance rbm 17:32:01 download the libxml2-devel rpm and install it. that is what is missing. there are two versions of libxml 1 and 2. nsxml only works with 2. 17:32:03 do you mean I should have libxml2-devel instead of libxml-devel 17:32:03 Jackal: "libxml2" == libxml version 2.x. "libxml1" == libxml version 1.x. You have the -devel for version 1.x, you need the -devel for 2.x 17:32:27 ok - i understand 17:32:35 man I hate packages 17:32:45 Packages are great. It's RPM that suck. 17:32:53 apt-get install libxml2-dev :-) 17:32:56 s/suck/sucks/ 17:33:00 debian bastards 17:33:02 :) 17:33:06 haha 17:33:26 Jackal: There's apt for RPM you know. 17:33:39 I've used it in several RH/Mandrake boxen. 17:33:56 rbm: fascinating 17:34:06 rbm: where can I get that? 17:34:06 i guess that is not true 17:34:15 I'm on RH 17:34:53 http://apt-rpm.tuxfamily.org/ 17:34:53 G: http://apt-rpm.tuxfamily.org/ from rbm 17:34:57 G:| APT for RPMs 17:34:57 titled item G 17:35:26 cool thanks/ does that work well? 17:35:52 apt can't do magic and overcome RPM's limitations and suckiness, but it helps a lot. 17:36:33 ic 17:36:42 thanks 17:36:44 once you get libxml2-dev installed, then you need to edit the Makefile. Add the following lines near the bottom: 17:36:55 CFLAGS = $(shell xml2-config --cflags) 17:37:03 MODLIBS = $(shell xml2-config --libs) 17:37:15 ok thanks 17:37:26 that will set the compiler flags and libs to be appropriate for your distro 17:37:28 sure, np 17:37:46 * rbm goes to work 17:40:06 ns_xml.c:38: ns.h: No such file or directory 17:40:16 after installing libxml2-devel 17:41:34 is nsxml underneath the aolserver directory? 17:41:54 it's not finding ns.h which should be in the aolserver sources 17:41:59 yes in the source aolserver directory 17:42:08 i haven't installed aolserver yet 17:42:15 should I have? 17:42:19 or just make? 17:42:45 you just need the headers to be there 17:42:48 it can't find the source for aolserver. 17:43:03 what is the gcc command that it is trying? 17:43:25 the easiest way is to put nsxml at the same level in the aolserver source tree as aolserver/nscache/nsrewrite,etc.. 17:44:35 cc -I/usr/include/libxml2/libxml -I/usr/include/libxml2 -c -o ns_xml.o ns_xml.c 17:44:40 *gcc 17:45:06 mine looks like this: 17:45:08 gcc -I/usr/include/libxml2/libxml -I/usr/include/libxml2 -I../aolserver/include -D_REENTRANT=1 -DNDEBUG=1 -g -fPIC -Wall -Wno-unused -mcpu=ppc -DHAVE_CMMSG=1 -DUSE_FIONREAD=1 -DHAVE_COND_EINTR=1 -c -o ns_xml.o ns_xml.c 17:45:35 wow 17:45:36 you're missing the -I../aolserver/include (plus alot of other stuff) 17:46:10 i don't think that I messed with the Makefile THAT much. 17:46:19 the last line in my Makefile says: 17:46:28 include $(NSHOME)/include/Makefile.module 17:46:45 that's where most of this stuff comes from (NSHOME is set near the top of the makefile) 17:48:08 hmm. I'm going to redownload nsxml 17:51:12 same error 17:51:29 Jackal: You should use the 3.3ad13 tree. 17:51:30 vinod - did you change your makefile alot? 17:51:35 s/tree/release/ 17:51:53 nope. just the change that i mentioned above (adding CFLAGS and MODLIBS) 17:52:21 WTF: CrashCmd 17:52:21 Label WTF not found. 17:52:34 "The name of a Tcl function which, when executed, will crash the server. 17:52:46 cool! 17:52:48 hmmm... maybe my aolserver is calling this command? 17:52:58 from: The name of a Tcl function which, when executed, will crash the server. 17:53:01 argh 17:53:04 rename CrashCmd ns_db gethandle 17:53:14 from: http://tcl.activestate.com/man/aolserver3.0/con-ch3.htm 17:53:57 markd2: didn't they have that option in aolserver 2? 17:55:17 Jackal: from with the nsxml dir, do: "ls -l ../aolserver/include/Makefile.global" 17:56:32 * NsTclCrashCmd -- 17:56:32 * 17:56:32 * Crash the server to test exception handling. 17:56:32 which nsxml are you compiling? 17:56:50 that doesn't work because nsxml is inside aolserver dir 17:56:58 ahhhhh 17:57:11 got the nsxml from http://acs-misc.sourceforge.net/nsxml.html 17:57:14 ok 17:57:31 markd2: I don't need no stinkin' command to crash my aolserver! 17:57:35 nsxml should be next to the aolserver source (not inside it, like i said) .... sorry (always listen to davb, not me) 17:57:48 and in my Makefile I changed NSHOME to NSHOME = ../ 17:57:54 :) 17:57:57 ok 17:58:46 I still can't believe I am segfaulting aolserver while it is parsing an adp file. 17:58:49 Somebody using Red Hat should just build 3.3ad13 RPMs 17:59:54 is this true? why can't the src of aolserver work? 18:00:15 ad13 has internationalization fixes 18:01:48 hazmat (~ender@adsl-66-123-57-58.dsl.lsan03.pacbell.net) has joined #openacs 18:01:51 sigh 18:02:49 is that so? 18:03:03 ok, my aolserver is crashing in the tcl part of the code. I wonder if I have some other tcl installed that is screwing it up 18:03:08 this is really frustrating 18:03:31 nsxml should build if you move it up one level. then you should be in business. 18:04:04 hey! 18:04:15 not in business 18:04:15 tcl 8.3 includes TclSOAP and XMLRPC 18:04:16 Jackal: What red hat are you on? 18:04:20 6.2 18:04:29 6.2 is an evil pain in the ass to get working with libxml2 18:04:32 Oh crap, that's older than my grandma 18:04:36 ns_xml.c:38: ns.h: No such file or directory 18:04:38 lol 18:05:16 but that is not the problem you are having. 18:05:17 7.x is a pain to get working with Oracle 18:05:26 what is nshome set to in the nsxml Makefile? 18:05:42 NSHOME = ../aolserver-3.4.2 18:05:59 you are too high. 18:06:09 am I ;) 18:06:11 in the tree. 18:06:44 find the Makefile.global file in the aolserver source tree. 18:06:51 point NSHOME at that directory. 18:06:56 aolserver-3.4.2 and nsxml are in the same dir 18:07:29 anyway, if TclSOAP is part of Tcl now, it might be possible to trick it into working in aolserver., but you still need expat and TclDOM. 18:07:53 Jackal: is there a aolserver dir underneath aolserver-3.4.2? 18:08:05 hey nsfts is in tcl8.3 also. 18:08:10 vinod : no 18:08:18 interesting. the debian packager is stuffing everything in there. 18:08:41 Jackal: Do you want a working 3.3ad13 tarball? 18:09:04 nsfts? Dan's module? 18:09:05 well - i'd like to try getting the vanilla intallation working 18:09:18 rbm: yeah. 18:09:25 Jackal: have you compiled stuff before? 18:09:28 well the version that works with plain tcl anyway. 18:09:30 yes 18:09:33 you are very close. 18:09:42 can you find Makefile.global? 18:09:52 yes its in include 18:10:07 hmmm. 18:10:26 what is in your aolserver3.4.2 directory? 18:10:51 aolserver-3.4.2 is just the untarred src 18:11:02 and there is no aolserver directory under that? 18:11:24 no 18:11:46 weird. 18:12:12 why is that wierd? 18:12:43 there should be an aolserver directory with the aolserver source in it, along with a directory for every ns* module 18:12:53 the tarballs that i have has aolserver-3.x/aolserver/include/Makefile.global 18:13:12 aolserver-3.4.2 IS the source 18:14:32 all I did was untar the aolserver src. - rename it to aolserver-3.4.2 and move it to /home/nsadmin 18:14:56 i don't want to overwrite my work ( sans xml) installation of aolserver 18:15:04 right. 18:15:07 which was a binary installation 18:15:24 ok, so /home/nsadmin/aolserver-3.4.2/include/Makefile.global exists 18:15:47 yep 18:15:50 point NSHOME to /home/nsadmin/aolserver-3.4.2 18:16:39 same error 18:18:58 ns_xml.c:40: libxml/parser.h: No such file or directory 18:19:06 oh - that error 18:19:13 the ns.h error is gone though, right? 18:19:31 looks like it 18:19:59 ok, good 18:20:10 now we need to get nsxml to find libxml2-devel 18:20:23 at the shell, type this: xml2-config --version 18:20:36 2.4.12 18:20:48 perfect 18:21:07 ok add these 2 lines just before the "include" line at the bottom of the makefile 18:21:26 MODLIBS = $(shell xml2-config --libs) 18:21:26 CFLAGS = $(shell xml2-config --cflags) 18:21:28 CFLAGS? 18:21:43 added them before but AFTER include 18:21:46 yup 18:22:05 yaay! 18:22:11 tada. 18:22:14 like magic - thank you 18:22:23 vinod, does that xml2-config work on all platforms? 18:22:29 i knew it was something stupid 18:23:06 i think so. it works on BSD, Linux 18:23:13 cool. 18:23:15 very good to know 18:23:33 I am working on ns_xml 2.0, so I will make sure that gets in the makefile. 18:23:48 cool 18:23:48 there's a similar utility for postgres: pg-config 18:23:56 I didn't write it, I am just looking at it. 18:24:07 also xslt-config :-) if you want to add xslt support to nsxml 18:24:21 yeah, I did that. works great! 18:25:11 rbm: can you send me a working 3.3+ad13 tarball? Maybe I can track down this crashing bug I am having. 18:25:33 hmmm... I wonder if it is something wacky in the kernel... 18:25:41 when is it crashing? 18:26:08 davb: just sources or binaries too? 18:27:36 anyone know off-hand which imagemagick utility shows the width and height of a jpeg? 18:27:49 oh I know that one... identify 18:27:52 so would this be the new line to get xslt working? 18:27:52 I think 18:28:00 MODLIBS = $(shell xml2-config --libs) $(shell xslt-config --libs) 18:28:01 sweet. thanks 18:28:18 Jackal: you don't need the xml2-config part 18:28:36 xslt-config will give you the proper libs for xml and xslt 18:28:44 ic - make sense 18:28:58 dude where the hell do you learn this? 18:29:21 i implanted an electrode in markd2's brain to transfer all his knowledge to me 18:29:35 lol 18:29:37 * markd2 ooooms 18:29:39 thing is - it comes over on dialup 56Kbps, so it's taking a while 18:29:47 vinod: get me a copy will ya? 18:29:49 hey! I've got isdn now 18:29:54 I figured the upload would go a little better 18:29:59 maybe it's *cough* problems on the receiving end 18:30:07 p2p markd2 18:30:09 haha 18:30:33 * rbm puts markd2's brain in morpheus 18:31:06 xslt-config doesn' work for me - is that another lib I need to download? 18:31:24 yeah - avail at xmlsoft.org (should be rpms somewhere) 18:31:45 libxslt-1.0.15-1.i386.rpm ? 18:32:17 that should be it - you need the devel package too 18:32:23 oh ok 18:33:24 plus nsxml defines a separate CFLAG in order to compile xslt support (-DDO_XSLT), so you need to add that to CFLAGS: 18:33:44 CFLAGS += -DDO_XSLT 18:33:50 rbm: nm. I have a working 3.3+ad13 on my web server. 18:33:54 I can just use that. 18:34:10 to lunch! 18:34:16 damn - seriously is there any documentation anywhere? 18:34:54 Jackal: there are lots of documentation in several places all over the web 18:35:05 ;) 18:35:07 thanks 18:39:42 CFLAGS = $(shell xml2-config --cflags) 18:39:45 MODLIBS = $(shell xslt-config --libs) 18:39:47 CFLAGS += -DDO_XSLT 18:39:49 CFLAGS = $(shell xml2-config --cflags) 18:39:50 so is this right? 18:40:45 no, 18:40:46 :) 18:40:51 :( 18:40:56 very close. 18:41:14 besides the extra CFLAGS at the bottom? 18:41:14 all the CFLAGS have to be on one line. 18:41:26 oh 18:41:29 put the CFLAGS += -DDO_XSLT last. 18:41:41 every time you do CFLAGS = you erase the previous one. 18:41:51 the += adds it to the end. 18:41:54 CFLAGS = $(shell xml2-config --cflags) 18:42:03 MODLIBS = $(shell xslt-config --libs) 18:42:09 CFLAGS += -DDO_XSLT 18:42:28 yeah - that happened when I was trying to paste 18:42:42 using mozzilla irc chat - not so great 18:43:01 you need CFLAGS = $(shell xslt-config --cflags) I think 18:43:22 or comment out the -DDO_XSLT line 18:43:24 instead of the xml-config right? 18:43:57 AAAGH! 18:44:06 Still can't get package require to work 18:44:12 I know I did it in the past 18:44:31 me too. 18:45:14 but that was before I reformatted my harddrive. 18:45:44 I will have to look when I get home, to see if I burned that stuff to a CD 18:47:58 jkhong has quit (Read error: 104 (Connection reset by peer)) 18:57:26 darn, it still crashes 19:01:16 are these crashes reproducible? or random 19:01:29 extremely reliable. 19:01:47 its in the tcl interpreter somewhere, but only very rarely does it occur. 19:01:57 one time is parsing an ADP file for dotLRN. 19:02:22 the other is when I request a missing page under an ETP 2.0 package instance. 19:02:30 one is on oracle, the other postgresql. 19:02:50 it happens at exactly the same place with different data. 19:03:38 hmm.. i don't have dotLRN or ETP 2 installed to test 19:04:16 right. those are the only times I can get it to crash. I copied over the aolserver on my web server and it did the same thing. it runs fine on the other machine. 19:04:58 what are the differences betw the 2 machines (kernel versions?) 19:05:24 oh, that is right. web server is 2.2.19, this one is 2.4.18, although I think it crashed before I switched kernels. 19:06:16 yeah, i wouldn't know if it's a kernel issue. just throwing things out there :-) 19:06:21 heh 19:06:35 I am copying the files to the other machine, I will set it up there and test it. 19:18:15 So what registrars do you recommend? I'm looking for a better value than network solutions. 19:18:21 my brasileiro.net is about to expire 19:18:53 my cololocation place does registrations (I'm not sure who they go through) 19:20:57 nsadmin@server1:/tmp$ tar -xzf oacs4.tar.gz 19:20:57 nsadmin@server1:/tmp$ createdb openacs4 19:20:57 CREATE DATABASE 19:20:59 argh 19:21:01 heh 19:21:16 http://www.domainmonger.com/index.shtml this one is reportedly good. 19:21:18 rbm: i've been using namesdirect.com because they're assoc with mydomain.com which is where i manage my domains 19:22:54 talilee_ (~talli@ip64-75-147-135.dial.maui.net) has joined #openacs 19:23:09 talilee_ is now known as talli 19:23:14 hey guys 19:23:18 hey talli 19:23:25 vinod, congrats on the new gig 19:23:28 and in manhattan, too! 19:23:31 thanks! 19:23:37 yup - i'm psyched 19:23:43 at least i know where not to go if i'm sick 19:23:57 domainmonger.com is $17. I like that. 19:24:19 vinod, i guess we'll have to have more socials in nyc... 19:24:26 talli: appreciate that - hate it when sick people come to see me 19:24:29 btw, how is april 23rd for a cambridge social? 19:24:47 looks good to me 19:25:08 vinod, sick people and beautiful super models with no shirt on, right? 19:25:26 I probably should ask him :) 19:25:42 * hazmat peeks in 19:25:58 hey hazmat 19:26:14 hi vinod, congrats 19:26:40 thanks! 19:26:51 talli: I found someone possibly willing to work on webdav. 19:26:55 whoa! 19:26:57 nice!!! 19:26:59 or help someone who is. 19:27:05 hey guys - whay is nssha1 19:27:13 *what 19:27:19 and where can I get it? 19:27:32 davb, who is it? 19:27:33 When are you going to get me a ticket to go to manhattan to participate in one of these socials? 19:27:40 its included with aolserver 3.3+ad13 19:28:10 or with Mat's version of 3.4.2 for openacs 19:28:21 rbm: when you get me a ticket to brazil to participate in one of those jungle orgies 19:29:23 talli: But I'm just a poor student. 19:29:56 talli, I understand that jungle orgies have stringent screening procedures... you won't be able to get in. 19:30:09 talli: You're the one with all the corporate contacts and money :-) 19:30:14 yeah, i'm "too big" as the girls say 19:30:34 so far, corporate contacts != moeny 19:30:37 money 19:33:56 you mean, (corporate contacts) : money ? moeny 19:34:38 talli: Oh, c'mon now. A ticket SLC -> NY costs what, $150? 19:35:07 i mean that you should stay away from those jungle orgies, denshi. the head shrinkers aren't going to be doing you any favors, if you know what i mean. 19:35:21 rbm: is that what it costs, really? 19:35:33 check how much a ticket from SLC to Boston costs 19:35:48 rbm: you have to get back too! 19:36:47 What's the airport in Boston? 19:36:49 the problem is a ticket from SLC to boston is cheaper than a ticket frm albany to boston 19:36:58 haha 19:37:01 logan airport 19:37:03 in boston 19:37:27 BOS 19:38:30 careful there, talli, I can defmacro your ass. 19:38:31 argh 19:39:00 I started an nsd on my web server without daemontools, now I don't know which one to kill! 19:39:40 if I kill the one listed in log/nspid.servername will that get all the others attached to that server? 19:39:52 oh yeah, I plugged about with libical this weekend 19:40:15 took awhile, since there are no docs worth anything 19:40:35 how does it look? 19:40:38 you done anything new on that? 19:40:49 ok, it crashes on the other machine also. 19:40:51 this sucks. 19:40:53 no, been thinking, but been busy this weekend 19:41:05 talli: can you find out what version of aolserver luke has tested etp 2.0 with? 19:41:15 or I can email him... 19:41:37 sure 19:41:51 it looks pretty good, but it's in a silly place and will be replaced as the spec evolves. 19:42:07 what do you mean in "a silly place"? 19:42:34 k2pts (~nkd@213.149.180.151) has joined #openacs 19:43:57 luke's coming here, davb 19:44:05 cool, thanks 19:44:41 lukep (~lukep@h0050da60b459.ne.client2.attbi.com) has joined #openacs 19:44:52 hey lukep 19:45:03 howdy 19:45:30 hi lukep. as I was saying, I have consistently crashed aolserver on two different machines with aolserver 3.3+ad13 and aolserver 3.4.2 19:46:07 when I try to access the /etp/ or any other page that does not exist under etp 2.0 19:46:14 talli: silly as in it's more in the place where I'd use a scripting language. For instance, it doesn't do sockets, it expects to live in a socketed application and be handed CAP statements pulled from a socket, then it does parsing and the like. 19:46:45 lukep: so I was wondering what version of AOLserver you used. 19:46:59 etp 2.0 means you're using the new code from the musea cvs? 19:47:01 so it gets you there, but unless you really like flex, you'll eventually write the new spec stuff in a higher language. 19:47:02 yes. 19:47:11 denshi, is it better to build sockets in C? 19:47:53 what's the difference between building the sockets in C or in python and use libical to do the parsting? 19:47:55 we're using 3.3ad13 19:48:01 lukep: ok. hmmmm 19:48:09 except on low-level issues, I'd say there's no difference. And I don't know low-level networking. 19:48:14 well that doesn't help :) it must be something wacky in my systems. 19:48:41 yeah, weren't you having the same problem with aolserver crashing a while ago? 19:49:12 yes, but I ignored it and did something else. now I can't use part of dotlrn because of it, so I was trying to diagnose it. 19:49:38 rbm: BTW, can you send me a binary of your working 3.3+ad13? 19:50:00 davb: you want that now? 19:50:08 if you have it handy :) 19:50:12 I can send it 19:50:27 cool. thanks. 19:51:03 I changed the only thing that is different about my mahcines, I turned of xslt support in nsxml. everything else should be the same as everyone else is using. 19:54:20 brb 19:56:54 ok dave, i sent it 19:57:39 k2pts has left #openacs 19:58:04 bye k2pts 19:58:07 thanks 19:59:04 paje has quit ("regrouping; bbiab") 19:59:26 oh yeah, we have to train paje with the answers to common questions 19:59:53 like, "why is denshi so lame?" 20:00:01 I can install some info paks 20:00:40 argh, mozilla crashed 20:00:52 darn, that nsd doesn't like my nssock. 20:01:07 well, if you'd wash them more often... 20:01:24 heh 20:01:46 [02/Apr/2002:15:41:26][958.1024][-main-] Warning: modload: failed to load '/home/nsadmin/aol342/bin/nsxml.so': '/lib/libc.so.6: version `GCC_3.0' not found (required by /usr/lib/libxslt.so.1)' 20:02:01 i seam to be have trouble getting that nsxml to work 20:02:20 compile it without xslt support. unless you need it, that is. 20:02:35 i do 20:02:50 oh, looks like you need to update your OS 20:02:57 hey, who's the webdav victim? 20:03:06 davb: really? 20:03:12 davb: which part? 20:03:41 well, updating gcc and lic.so will affect everything else in your system I suspect. 20:03:52 argh libc6.so that is 20:04:01 but I could be wrong. 20:04:22 is nsxml that new? 20:04:56 no, but libxslt is pretty new. 20:05:03 oh ic 20:05:25 wait, does anything in oacs4.5 use libxslt? 20:05:32 not yet. 20:05:33 :) 20:05:42 lukep has left #openacs 20:05:45 soon though 20:05:48 thanks lukep! 20:05:54 darn. 20:05:55 well I think I need xslt - I want to use a xml data use it to generate some pages with formatting 20:06:00 galeon is pretty nice 20:06:03 isn't xslt the way to go? 20:06:11 possbily. 20:06:25 what is alternative? 20:06:47 writing some tcl procs to parse the xml and turn it into html 20:07:12 hmm. but then what help is the xml module? 20:07:40 it can parse the xml into a tree. or create xml. 20:08:07 by a tree - do you meean as in (list ) 20:08:31 bookmarklets.com 20:10:36 no, its a tree. kinda hard to explain, I am hardly an expert programmer, I just screwed up with openacs often enough to learn some stuff. 20:11:46 Ok well it sounds like I won't have to waste alot of code describing how to parse XML and get out values 20:12:04 denshi has quit (Remote closed the connection) 20:12:06 abbaJ has quit (Connection timed out) 20:12:58 right. there are pretty good docs and examples for using nsxml: http://www.fifthgate.org/articles/aolserver/xml/ns_xml_doc.html 20:13:19 cool - just have to recompile ( now that I know how) : thanks 20:14:08 np 20:14:25 denshi (~chatzilla@adsl-216-62-223-193.dsl.rcsntx.swbell.net) has joined #openacs 20:14:35 http://www.fifthgate.org/articles/aolserver/xml/ns_xml_doc.html 20:14:36 H: http://www.fifthgate.org/articles/aolserver/xml/ns_xml_doc.html from rbm 20:14:54 H:|pretty good docs and examples for using nsxml 20:14:54 titled item H 20:15:03 :) 20:15:05 actually tcl/adp is probably more elegant that xslt anyway 20:15:28 its is alot easier to learn 20:16:26 Jackal, i started a thread on the oacs list called "Utility of XSLT" that was very instructive 20:16:54 two guys whose opinions i respect ALOT, John Sequiera and Henry Minsky, said that XSLT wasn't really worth the overhead 20:17:04 you may want to check that out. lemme get the thread 20:17:12 sure 20:17:32 overhead vs what? 20:17:57 http://openacs.org/bboard/q-and-a-fetch-msg.tcl?msg_id=0002RD&topic_id=11&topic=OpenACS 20:17:58 I: http://openacs.org/bboard/q-and-a-fetch-msg.tcl?msg_id=0002RD&topic_id=11&topic=OpenACS from talli 20:18:23 I: thread on OACS.org: "Utility of XSLT?" 20:18:23 commented item I 20:18:28 great - thanks 20:19:12 this from John Sequiera (a hacker who we've worked with and is a GOD): " It is *so* easy to write different templates that run off the same application logic to target different devices, simply using procedural abstractions for code re-use, that I don't understand why people keep bringing this up as an argument in favor of XML. 20:19:12 " 20:19:35 modload: failed to load '/home/nsadmin/aol342/bin/nsxml.so': '/home/nsadmin/aol342/bin/nsxml.so: undefined symbol: xmlInitMemory' 20:19:56 still can't get nsxml to work 20:20:56 ever seen this? 20:21:53 that thread is also a nice artifact of aD vs OACS history 20:22:52 boy, *there's* something worth archiving... 20:23:05 Jackal: i haven't see that error... 20:23:12 crap 20:23:12 loosk liek it can't find the libxml2.so 20:23:25 ack, looks like I can't type 20:23:37 I recall having this problem on redhat also... 20:24:15 saw it fine while compiling 20:24:21 but now now? 20:25:05 you need to check with ldd 20:25:11 try ldd path to nsxml.so 20:25:21 denshi: what? the XML/XSLT discussion or the aD vs OACS war? 20:25:25 If only I could understand why is it that package require is failing... 20:25:39 what's interesting about that thread is that aD tried to friggin' hard to validate their approach 20:25:41 not sure what you mean about ldd 20:25:41 it is failing because aol doesn't care if it works or not. 20:25:43 which never seemed to friggin work 20:25:51 ldd will tell you where it is finding the so files. 20:26:00 mine looks like this: 20:26:09 dave@leo:/usr/local/aolserver$ ldd /usr/local/aolserver/bin/nsxml.so 20:26:09 libxml2.so.2 => /usr/lib/libxml2.so.2 (0x40006000) 20:26:09 libxslt.so.1 => /usr/lib/libxslt.so.1 (0x400a3000) 20:26:09 libc.so.6 => /lib/libc.so.6 (0x400cc000) 20:26:09 libz.so.1 => /usr/lib/libz.so.1 (0x401e9000) 20:26:10 libm.so.6 => /lib/libm.so.6 (0x401f8000) 20:26:14 sorry... 20:26:58 I know we went over this on the bboard way back... 20:27:15 talli: the aD vs OACS war. I suddenly had the dark vision of a hardcover textbook being published under the title "How to Ruin your Developer Community and go Bankrupt" 20:27:33 Where does Tcl store this package database? 20:27:36 question about databases, oacs and extended tables... every row of a typical acs table has an ID number that probably corresponds to object_id... if I want to have a table, then extend each row using a second table, I would establish a 1-1 relationship between rows, probably by making all the ids the same in the "composite row"... yes? 20:27:40 ldd /home/nsadmin/aol342/bin/nsxml.so 20:27:43 denshi :) 20:27:52 libxml.so.1 => /usr/lib/libxml.so.1 (0x40011000) 20:27:58 come to think of it, I should get on that. 20:28:01 libz.so.1 => /usr/lib/libz.so.1 (0x4007f000) 20:28:09 libc.so.6 => /lib/libc.so.6 (0x4008e000) 20:28:11 jim: Sounds like a good plan 20:28:44 "/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x80000000)" 20:29:21 ok, specifically about users (this goes to "I can't delete a user, or I'm not sure what's happening when I do so, and it says 'I did it'") 20:29:48 Jackal: its finding libxml1 instead of libxml2 20:30:06 we have users with a user_id, which must match a person_id in persons 20:30:17 I delete a user... 20:30:26 I try to create the same user again... 20:30:32 brb 20:30:58 it says "can't, found row in database matching the email addr you're trying to use" 20:31:52 I know I'm wandering around here, and being sorta vague about my questions and such... but my understanding in this area is a little vague too... please bear with me 20:32:28 so if I delete a user, how would I make it so the email address is available again? 20:33:05 (btw, is this only happening to me? only happening on 4.5 with pg 7.2?) 20:33:34 popping out a level, momentarily... 20:33:58 how are you deleting the user? 20:34:03 I set my server to send out an email with a link to confirm registration... 20:34:49 rbm: one sec please :) (congrats on getting the AM, btw... I'll make myself available to him for questions) 20:38:24 before I got things together with my mail server (configuring relaying), it didn't yet work when I tried to get some users added... so, they would try to add themselves, the oacs would try to send the mail, qmail replied "domain not in rcpthosts (aka relaying denied from 127.0.0.1)" 20:38:59 and my guess is the user record was not completely filled out in the db table... 20:40:18 so then I tried to approve them, and thereby created a situation where the user tables weren't working because not all the attributes were filled in (upon someone trying to log in) 20:40:49 so that's why I now want to delete the rows representing the users from the database 20:41:18 ok, going back in a level (to process of deleting a user) 20:41:19 errr. I'm not sure what to tell you. Perhaps it's be best to just reinstall or delete all person entries (except for the admin) 20:41:42 I was thinking that would be the case myself... 20:43:28 but I wanted to learn something about the users mechanism... but in more general terms, Here I am, I can install an acs and an oacs,,, but when the first problem comes up, I have NO idea how to deal with it... 20:43:56 does anybody know what I can do about /home/nsadmin/aol342/bin/nsxml.so: /lib/libc.so.6: version `GCC_3.0' not found (required by /usr/lib/libxml2.so.2) 20:44:25 in acss that are used in companies, it's nice for them to have someone who can take care of problems without having to reinstall, taking all data with them :) 20:44:53 hazmat has left #openacs 20:44:58 I am GREEN in that respect. neuoobeeeieeee. 20:45:05 err, newbie 20:46:19 so, I have some users, and rather than dumping them, I want to try fixing... or one step less good, I want to delete the users without reinstalling whole system. 20:47:55 jim: Check the PL/{pg}SQL procs in the parties datamodel 20:48:03 ok. 20:51:10 community-core-create.sql 21:01:28 question about create_type (will it create a table if not exist before?) and create_attribute (will it extend existing table to add attribute column if not exist?) 21:02:25 jim: there are procs that will build the table and add columns for attributes 21:03:32 acs_attribute__something 21:03:33 I think 21:07:31 ok 21:09:52 * rbm learns how to have his package have parameters 21:10:28 * til estimates rbm will have learnt everything about it in 10 seconds 21:10:56 ad_parameter is really convenient 21:13:27 davb: is there a way to compile libxml2-devel myself 21:13:42 I don't think thats the problem. 21:13:43 try running ldconfig 21:13:58 and then do ldd path-to-nsxml.so again 21:15:05 ok 21:17:12 ok - that worked : will try to load it on server 21:19:27 * rbm scratches head 21:19:58 crikey! it worked 21:20:05 I thought so. 21:20:11 To register a parameter you do apm.register_parameter(...) right? 21:20:22 should I try xslt? 21:20:30 you need to set the parameters in the package .info file 21:20:44 or use the apm interface to add them. 21:21:56 which will add them to the info file. 21:21:58 Ah... it's in the .ingo 21:22:23 crap. I thought I could add a parameter without reloading the package 21:22:24 I think the apm uses that function when it loads a package. 21:22:37 reload as in reinstall? 21:22:49 it should work if you use restart the server I think... 21:25:24 I think I'll insert that param on the .info and then on the table, restart the server and see if that works :) 21:25:33 apm_parameters should do it 21:27:26 ohh, /acs-admin/apm/version-parameters should do it all for me 21:27:28 i think using the webinterface for adding a parameter is the easiest way ... afaik it should be available right after adding it, without restart/load 21:30:00 bye 21:30:04 davb has quit () 21:30:41 Looks like it. That's nice 21:34:02 * til pulls his hair b/c dotlrn still won't work 21:40:47 * til mounts a _second_ instance of dotlrn and hides ... 21:44:36 vinod: how is the 23rd of april for an oacs social? did i ask you this already? 21:45:38 does anybody have some example code for how to use ns_xml? 21:47:49 Jackal: i have an old xmlrpctest package, but i am not sure if it still works 21:48:22 i think I got it installed - i just don't understand how to use nsxml 21:49:03 there should be some examples in the toolkit too, e.g. where the .info files are parsed 21:49:11 hmm 21:49:17 ok - i'll look 21:49:18 thansk 21:49:19 Jackal: Ask oacs-chomp for the URL. /m oacs-chomp view 21:53:37 talli: april 23 sounds good to me 21:53:47 ok, cool 21:54:11 i'm on my way out to catch my flight 21:54:33 i'll talk to you guys in a couple weeks! 21:54:43 vinod has left #openacs 21:54:44 later vinod 22:00:05 talli has quit ("Trillian (http://www.ceruleanstudios.com)") 22:02:04 denshi has quit () 22:28:32 Psychephylax (~nickb@pool197.eng.cv.net) has joined #openacs 22:28:39 * Psychephylax goes home 22:28:56 * markd2 goes crazy 22:37:39 Hi... could someone try registering on 12.233.187.5:9000/? 22:37:55 i'll try 22:37:59 thanks... 22:38:06 want em to try too? 22:38:07 er, me 22:38:11 sure :) 22:38:25 I reinstalled; no time to try fixing... 22:38:28 * til is a jazz lover too 22:38:50 * jim puts off learning to fix things without reinstalling everything for another day :) 22:38:58 there's nothing there yet... 22:39:06 just trying reg for now 22:39:54 * til reinstalls oacs/oracle for the third time to get dotlrn running 22:40:17 progress? :) 22:40:38 To confirm your registration, please go to 22:40:38 well, outside of the confirmation having 'yourdomain.com in there', seems to work 22:40:57 ... that's what i wanted to say too 22:41:13 I had to manually construct the confirm url 22:41:16 no biggie 22:41:21 jim, that is an acs-kernel parameter 22:42:27 i think setting that one should be included in the installation process 22:42:51 ok, could one of you give it another shot with a different email? 22:43:09 til: agreed 22:44:12 sweet 22:44:17 sorks 22:44:32 nice :) 22:45:25 thanks :) 22:46:38 I'll put stuff on the site (forums, chord lessons, key signature drills) tonite, and mail out to people to try signing up again 22:46:45 cool 22:46:55 I'm really square. reading this stuff will be good for me :-) 22:47:06 do you play? 22:47:07 when I see chord progressions on a chart, my brain goes numb 22:47:09 trombone 22:47:12 (and bassoon) 22:47:20 hmmm, I can -help- you :) 22:48:01 I need all the help I can get 22:48:22 know your key sigs? can identify intervals? 22:49:28 I know the key sigs. I can't identify them on sight 22:49:37 I have to count on my fingers to tell if something is a 4th or 5th, ferinstance 22:50:05 sorry, what's a key sig? are those the tones that one can play without moving the handle? 22:51:04 like c, g, c, e, g... on the trumpet? 22:53:03 you're thinking of the harmonic series 22:53:30 the key sig(nature) are the sharps or flats at the beginning of each line to tell you what's sharped/flatted in the piece 22:53:49 ah, ok 22:54:44 wow, having online lessons for that would be cool! 22:56:03 you could put up some mp3's as listening exercises too ... 23:04:01 i am entering the need for SystemUrl parameter on installation in the sdm 23:15:02 guys: I'm having some trouble with aol3.10 23:15:24 for some reason my home page is displaying the headers 23:15:41 HTTP/1.0 200 OK MIME-Version: 1.0 Date: Wed, 03 Apr 2002 00:01:57 GMT Server: AOLserver/3.4.2 Content-Type: text/html Content-Length: 8353 Connection: close Content-Type: text/html Content-Length: 6684 23:16:34 anybody know how to resolve? 23:16:34 sorry ACS 3.10 23:19:09 looks like something else has been returned before the headers. did you modify the index.tcl(adp) or ad_header proc? 23:20:29 no but I just updated my aolserver to latest stable version 23:20:54 it is a index.tcl which is using doc_serve_template 23:21:12 te pages not using doc_serve_template is fine 23:21:25 however I use a different template for my root/home page 23:21:44 so doc_serve_template allows me to specify master_home.adp 23:22:27 sorry, i've never seen this proc 23:23:16 oh ok 23:24:53 I'm also having other problems 23:25:06 could well be that never tested that aolserver with oacs 3.x 23:25:13 for example typing filename doesn't redirect to filename.tcl 23:25:16 it used to 23:25:32 i'm sure 23:25:37 s/never/no one ever/ 23:31:42 markd2 has quit ("dinner") 23:57:18 Jackal has quit ("ChatZilla 0.8.6 [Mozilla rv:0.9.9/20020311]")