00:04:46 <bakataro> bakataro has quit ("daabaa chizubaagaa setto onegaishimas")
00:06:32 <jim> davb: see: bug#1343
00:07:56 <jim> bbiaw
00:40:45 <denshi> denshi has quit ()
01:37:03 <Psychephylax> Psychephylax has quit ("BitchX-1.0c17 -- just do it.")
02:03:58 <jim> back
02:04:17 <jim> * jim will try to solve bug#1343
03:09:53 <thom> thom (~Snak@12-248-24-199.client.attbi.com) has joined #openacs
03:10:11 <thom> thom is now known as _thom_
03:11:36 <_thom_> _thom_ has left #openacs
03:23:14 <rbm> rbm has quit (Read error: 113 (No route to host))
03:26:37 <paje> paje has quit (Read error: 110 (Connection timed out))
03:26:48 <jim> hi, I'm having problems remembering how to order the ands for the where clause to make things most efficient
03:33:44 <jim> but more basic, I'm editing wp-slim/www/index-postgresql.xql and I'm realizing I don't know how to get the current package instance id
03:33:52 <jim> in a query
03:36:18 <hazmat> bind variable from whats stuffed in ad_conn ?
03:36:44 <hazmat> maybe ad_conn package_id ? (its been a while for me ;)
03:38:24 <jim> hazmat: Hi
03:38:32 <jim>         and   ao.context_id = :wp_instance_id
03:38:41 <jim> I added that to each query...
03:39:04 <hazmat> cool.
03:39:20 <jim> so now I make sure $wp_instance_id is set before any of those queries?
03:39:58 <hazmat> yup
03:40:08 <hazmat> set wp_instance_id [ad_conn package_id] ?
03:40:45 <jim> ooh, it is already, to a different variable name
03:40:54 <jim> which I'll change to
03:51:06 <jim> maan, the pg port is really a mess (seems like, of -all- of oacs)
03:51:30 <jim> deletes of packages or package instances never work
03:52:00 <jim> I have a strong suspicion the 0or1row doesn't work in pg
03:53:17 <jim> I have a somewhat less strong suspicion that fixing that might fix a lot of problems
03:56:29 <docwolf> docwolf (~docwolf@adsl-34-207-203.bct.bellsouth.net) has joined #openacs
03:57:56 <jim> ok, that query worked!
03:58:25 <jim> ls
03:59:56 <hazmat> cool
04:09:53 <jim> now I guess I gotta find everywhere else it wants a list of presentations
04:30:19 <jim> "show me the authors" is one
04:31:28 <jim> try to delete a presentation... boom... 0or1row problem
04:34:19 <hazmat> hmm?
04:34:57 <hazmat> 0or1row doesn't work for pg? i kinda of doubt that....
04:35:20 <jim> I would too, if I didn't run into it over and over again :)
04:35:46 <jim> ok, check this out:
04:35:54 <jim> I jsut made a new user, 
04:36:01 <jim> and tried to create a presentation
04:36:03 <jim> boom
04:36:09 <jim> look at the logs...
04:36:12 <jim> sure enuf
04:36:21 <jim> 0or1row problem
04:37:05 <jim> does this necessarily mean 0or1row is bad in the driver? -no-, of course not, because we haven't found all the conditions
04:37:26 <jim> but the magnitude of the indicating vector is growing :)
04:37:47 <jim> if we could just see exactly where it's pointing...
04:37:56 <hazmat> ;), except that its used all over the toolkit, but your powers of deduction are persuasive.
04:38:36 <jim> well, a really simple test is in order
04:39:13 <jim> here's my present theory, just thrown out here without looking at anything yet:
04:39:32 <jim> 0or1row called, found one row, works -fine-
04:40:12 <jim> 0or1row called, found zero rows, bzzt, "statement was not a query returning rows"
04:40:33 <hazmat> hmm...
04:42:39 <hazmat> consider also that login uses it to check passwords in ad_check_password...
04:43:41 <jim> well, see, if the zero-row condition works there, then my theory is dead...
04:45:29 <jim> and in my simple test, I'm factoring out -all- of acs, leaving just aolserver api
04:45:54 <jim> well, I'm not doing that completely unless my theory holds under that condition
04:48:24 <hazmat> hmm...
04:48:36 <hazmat> i don't have a test server setup at the moment. so i can't confirm.
04:52:21 <talilee> talilee (~talli@talli.dialup.access.net) has joined #openacs
04:58:03 <hazmat> hi talilee: is that you or a ghost?
04:58:08 <talilee> hey hazmat
04:58:10 <talilee> que pasa?
04:58:23 <hazmat> you  never know what ghost lives in your machine...
05:00:53 <talilee> hopefully, it's casper
05:01:21 <hazmat> so hows nyc around midnight?
05:01:59 <talilee> ugh, don't get me started again
05:02:17 <hazmat> wow, that good.
05:02:42 <talilee> i just spent got back from my semi-annual test of NYC nite life and confirmed that it sucks
05:03:00 <talilee> at least, i can no longer stand the people
05:03:01 <hazmat> depends on where you go i suppose...
05:03:10 <talilee> true enough
05:03:55 <hazmat> i had fun for two nites in the city, although bar hopping was probably a low point, and heading into the catskills was a high point :)
05:04:42 <talilee> yeah, i was doing the bar hopping thing with newcomers to NYC. also, fake hipsters. or real hipsters. dunno which is sworse
05:05:45 <hazmat> * hazmat is still trying to quit smoking.. as he chews another piece of gum and fiends..
05:06:17 <talilee> stay strong, miman
05:06:19 <talilee> my man
05:06:44 <talilee> are you cold turkeying or quiting slowly?
05:07:02 <hazmat> cold turkey from a pack a day (well a little over)...
05:07:11 <talilee> whoa
05:07:15 <talilee> impressive
05:07:25 <talilee> good luck! what was the impetus?
05:07:43 <hazmat> wanted to start living again.
05:07:57 <talilee> :)
05:07:58 <docwolf> hazmat: don't be a hero.  get some gum or the patch.
05:08:08 <docwolf> seriously, you're gonna go postal
05:08:11 <hazmat> i'm chewing gum at the moment.
05:08:16 <docwolf> cool
05:08:37 <talilee> hazmat, i found that i could quit much easier if i slowly lowered the number i would smoke over time
05:08:47 <talilee> but that was my appraoch
05:08:57 <talilee> for others it's one way or the other. smoke or no smoke
05:09:03 <hazmat> doesn't work for me, if i have them i smoke them... addictive personality or some such.
05:09:12 <talilee> yeah, i understand
05:09:38 <talilee> i just got back from a couple smoky bars though. the stuff is truly disgusting. it's fun, but not worth all of its aggravation
05:10:38 <docwolf> i don't understand why it's so fashionable again to smoke
05:10:45 <docwolf> it seems like every young lady does it.
05:11:10 <hazmat> something about not having morals... too much consumer society sucks the marrow of life.
05:12:03 <docwolf> what's odd though is that women are usually concerned about their looks
05:12:14 <docwolf> ... smoking will turn you into a leathery prune faster than almost anything else
05:12:30 <talilee> yeah, that's the real issue
05:13:03 <hazmat> most young women are more concerned with fitting and being cool, same with young men.
05:13:04 <talilee> not to mention that it destroys your taste buds, your sense of smell, the ability to do things like, uhm, walk and heal...
05:24:20 <jim> Nope, the driver works
05:24:36 <talilee> talk to you guys later... gonna crash. have a good night all
05:24:38 <talilee> talilee has left #openacs
05:26:16 <jim> so, let's see, what does that mean...
05:26:26 <jim> ok, I used this script...
05:26:48 <jim> set h [ns_db gethandle]
05:26:48 <jim> set out "before<ul>"
05:26:48 <jim> set result [ns_db 0or1row $h {select i from x}]
05:26:56 <jim> if {$result == ""} {
05:26:56 <jim>     append out "<li> no rows"
05:26:56 <jim> } else {
05:26:56 <jim>     append out "<li> one row: [ns_set get $result {i}]"
05:26:56 <jim> }
05:27:04 <jim> append out "</ul>after"
05:27:04 <jim> ns_db releasehandle $h
05:27:04 <jim> ns_return 200 "text/html" $out
05:27:26 <jim> I tried 0 rows in x, 1 row, 2 rows
05:27:36 <jim> 2 rows gave an error
05:28:02 <jim> 1 row made 0or1row return an Ns_Set
05:28:26 <jim> 0 rows made 0or1row return ""
05:29:18 <hazmat> well that looks good
05:30:17 <jim> so: supposedly, this means I shouldn't go deeper into lower levels, cause I'll find the same thing for this script
05:32:32 <hazmat> next thing would be to to use the db api in a test script.
05:33:02 <jim> well, I just did that, right?
05:33:19 <jim> or you mean acs api?
05:34:16 <jim> the thing I just showed you, I put into a.tcl
05:34:23 <jim> and I browsed to a.tcl
05:34:36 <jim> before
05:34:36 <jim>     * no rows
05:34:36 <jim> after
05:35:07 <jim> that's after I did 'delete from x;'
05:36:47 <jim> I just now inserted a single row, i = 2
05:37:00 <jim> before
05:37:00 <jim>     * one row: 2
05:37:00 <jim> after
05:37:36 <jim> so that tests: the driver, and aolserver's database api
05:39:02 <jim> so next thing to test is openacs-4 database api?
05:39:04 <hazmat> i meant acsapi, although honestly i think looking over the original script might be more fruitful.
05:39:50 <jim> so you're suspecting many, many cases of sloppy portage
05:40:46 <hazmat> not really... i think there is something subtle probably happenig in the wp arena, thats causing the problem.
05:40:58 <jim> and by "the original script", you mean the one(s) that produced that error?
05:41:20 <hazmat> of course i could be wrong...  it happens alot :) .. yes the one that produced the error .
05:42:17 <jim> I'm still hoping to find "there's one single central problem causing all these symptoms"...
05:45:42 <jim> I'm hoping this because if it is, and it can be stamped in one blow and all those symptoms go away, openacs suddenly gets much more stable in many areas...
05:46:31 <jim> if not, it means we have to find what's wrong in a lot of code, pretty much over and over again
05:48:00 <jim> so I don't mind doing one or two more tests in that direction
05:48:26 <jim> except now, there's a bigger chance I'm wrong about it
05:59:33 <hazmat> why do you think there is a problem in lots of code and not localized to wp ?
07:04:19 <rbm> rbm (rmello@fslc.usu.edu) has joined #openacs
07:09:15 <paje> paje (~paje@slxwy.dorms.usu.edu) has joined #openacs
08:39:07 <andyn> andyn has quit ("leaving")
09:58:39 <andyn> andyn (~andy@12-254-190-230.client.attbi.com) has joined #openacs
10:43:15 <andyn> andyn has quit (Remote closed the connection)
10:55:46 <andyn> andyn (~andy@12-254-190-230.client.attbi.com) has joined #openacs
14:08:52 <talilee> talilee (~talli@talli.dialup.access.net) has joined #openacs
14:42:58 <til> til (~tils@port-212-202-128-195.reverse.qsc.de) has joined #openacs
16:20:55 <Psychephylax> Psychephylax (nblyumbe@ool-18bd7775.dyn.optonline.net) has joined #openacs
16:21:09 <Psychephylax> :-D
16:30:51 <jim> I was really hoping to find that the problem with all the places that use 0or1row were hosed in one place that could be fixed and then everything would be fine... however, I just ruled that out... 
16:31:13 <jim> (just in postgres, oracle is fine)
16:33:03 <jim> specificaly, there are a lot of queries that use 0or1row, and when they run, they say "query was not a statement returning rows"
16:39:49 <jim> jim has quit (Remote closed the connection)
16:44:12 <jim> jim (~jim@12-233-187-5.client.attbi.com) has joined #openacs
18:31:24 <djg> djg has quit ("Hey!  Where'd my controlling terminal go?")
19:01:46 <til> til has quit (Read error: 113 (No route to host))
19:05:03 <jim> hazmat: ayh?
19:05:09 <hazmat> hi
19:05:14 <jim> hi :)
19:05:18 <jim> you were right
19:05:21 <talilee> hey guys
19:05:30 <jim> there's no problem with the driver,
19:05:30 <hazmat> hi talilee
19:05:44 <jim> and no problem with db_0or1row
19:05:47 <hazmat> did you localize the problem to wp
19:05:56 <talilee> check out what this guy runs as his server: http://www.sixgirls.org
19:05:57 <hazmat> i didn't understand why you thought it wasn't localized?
19:07:25 <jim> no, there's definitely more than just wp
19:07:35 <hazmat> cool, amiga and netbsd... sounds like someone with too much time on there hands...
19:07:41 <jim> because I was getting the same message every time
19:08:08 <talilee> talk to you guys later
19:08:09 <hazmat> hmm.. have any openacs4 sites on pg been released todate ?
19:08:14 <jim> hey talli
19:08:15 <hazmat> latter talilee
19:08:18 <talilee> hazmat, yeah
19:08:20 <jim> see ya talli :)
19:08:21 <talilee> we've built some
19:08:36 <talilee> not entirely released, but they're running. some are internal web apps, though
19:08:42 <hazmat> ok, just wanted a reference point.
19:08:53 <talilee> also, openacs.museatech.net is OACS4, but it hasn't been "released" yet
19:08:58 <talilee> ok, gotta jet. bbl
19:09:21 <jim> talilee: do/did you sometimes get something like "0or1row: query was not a statement returning rows"?
19:09:28 <jim> dang :)
19:10:34 <jim> hazmat: and I was getting them -often-, usually when I was deleting package instances
19:10:47 <jim> but also when deleting other things
19:10:50 <hazmat> hmm...
19:10:57 <jim> (acs_object??)
19:11:34 <hazmat> is your server public?
19:12:05 <jim> temporary different subject: is anyone planning on approaching redhat and asking them to GPL acs 4.6?
19:13:15 <jim> I have 2 servers that are available (public is relatively not, but you can see them, and I can put up yet another one if desired)
19:13:27 <jim> but...
19:14:00 <jim> if you yourself are running a pg oacs, no reason why you couldn't see the messages there that I can think of
19:14:49 <jim> do you have about a half hour to engage in a hack session to find one instance of the problem and fix it?
19:27:59 <jim> the one I consider most public is http://12.233.187.5:9000/
19:31:09 <jim> here's a ferinstance
19:31:10 <Psychephylax> hmmm
19:31:26 <jim> I tried to delete a package instance... and I got this:
19:31:48 <jim> Database operation "0or1row" failed (exception NSDB, "Query was not a statement returning rows.")
19:31:48 <jim>     while executing
19:31:48 <jim> "ns_pg_bind 0or1row nsdb0 {
19:31:48 <jim>       
19:31:49 <jim> 	select case when count(*) = 0 then 0 else 1 end from apm_packages
19:31:51 <jim> 	where package_id = :package_id
19:31:53 <jim>     
19:31:55 <jim>       }"
19:40:09 <jim> hmm, wonder if it's in decode()
19:40:24 <jim> or db_string
19:41:05 <jim> or that either oracle or pg doesn't support things like case when count(*) = 0 then 0 else 1
19:52:59 <jim> ok, pg does accept that case
19:53:16 <jim> but decode seems hard to use
19:54:47 <jim> * jim looks for decode
19:58:28 <hazmat> sorry jim, i don't have the time at the moment. i just wanted to see the error and look at the function source.
19:59:37 <jim> ok, np
20:02:12 <jim> that's what I want to do too... but the api browser is broken, the message I get is "query was not a statement returning rows" but not from 0or1row
20:03:29 <jim> about seeing the error, I'm running mozilla which doesn't do well with showing source
20:03:57 <jim> but that's OK, I'll keep pluggin for awhile
20:24:42 <markd2> markd2 (~Snak@63.224.172.190) has joined #openacs
20:36:53 <Psychephylax> * Psychephylax thwapsc markd
20:36:55 <Psychephylax> doh
20:39:29 <markd2> heh
20:39:32 <markd2> thwap(tm)sc
20:40:03 <Psychephylax> no
20:40:18 <Psychephylax> stupid telnet in Winblows doesn't push the ascii chars well
20:40:35 <Psychephylax> * Psychephylax tries to benchmark his memory
20:40:42 <Psychephylax> i'll be back if my computer fudges up
20:47:25 <markd2> markd2 has quit ("wheeeee")
21:44:40 <k2pts> k2pts (~nkd@adsl-168-174.cytanet.com.cy) has joined #openacs
21:44:55 <k2pts> hey guys, is tilmann around?
21:45:09 <k2pts> paje: k2pts?
21:45:09 <paje> somebody said k2pts was blah blah blah blah :)
21:45:17 <k2pts> paje: hi
21:45:17 <paje> hello, k2pts
21:45:28 <k2pts> paje: seen til?
21:45:28 <paje> til was last seen on #openacs 1 days, 3 hours, 14 minutes and 19 seconds ago, saying: particularily nice i found the interactive features of wish, so that one can manipulate the running programm via a shell ... is python capable of something comparable? [Fri Feb 22 11:33:04 2002]
21:45:35 <denshi> denshi (toddg@linux126.ma.utexas.edu) has joined #openacs
21:45:39 <k2pts> hey denshi
21:45:57 <denshi> hey k2.
21:46:03 <k2pts> how's it going?
21:46:31 <denshi> pretty good.  There was a wealth of basketball that needed playing, so I obliged.  With bigor.
21:46:46 <denshi> s/bigor/vigor/g.  Damn mac keyboards.
21:47:25 <k2pts> paje: seen talli?
21:47:25 <paje> talli was last seen on #openacs 1 days, 21 hours, 17 minutes and 2 seconds ago, saying: cool [Thu Feb 21 17:32:18 2002]
21:47:37 <denshi> how's cypriot?
21:47:57 <k2pts> cyprus: nice weather as usual, too much work these days
21:49:58 <talilee> hey k2pts
21:50:01 <talilee> hey denshi
21:50:26 <denshi> word up, talli.
21:51:51 <talilee> how's everyone this afternoon?
21:52:26 <k2pts> hey talilee
21:52:42 <denshi> k2, I want to find a coffee place as fast as that.
21:52:50 <denshi> must be a greek thing.
21:52:53 <talilee> hey denshi, i'm goign to post some ideas for how to get new developers into the community
21:52:55 <k2pts> frape
21:53:12 <k2pts> shaken ice coffee (do it myself)
21:53:14 <talilee> one of the ideas is to start looking at building developer tools
21:53:34 <talilee> like aolserver debug tools, nsadmin and TOra stuff
22:01:05 <denshi> either I'm dead or my watch has stopped.  Or maybe talli stopped typing.
22:01:13 <talilee> you're dead
22:05:31 <jim> wow... decode() is everywhere in acs
22:06:23 <jim> does anyone know where it's defined? (tried recursive grep in an openacs dir, couldn't find the def)
22:06:53 <denshi> jim: decode() in SQL or Tcl?
22:07:05 <k2pts> jim: that's an oracle construct
22:07:09 <jim> tcl is ad_decode, yes?
22:08:02 <jim> k2pts: yeah, which is used everywhere in openacs... it -appears- to work, but I can't find the def
22:08:02 <k2pts> yes
22:08:28 <k2pts> it's everywhere in the openacs (oracle or pg). oracle I presume...
22:08:29 <jim> I have an oracle openacs4 and 3 pg ones
22:08:47 <denshi> bye guys.
22:08:50 <jim> and I get err msgs involving decode
22:08:52 <k2pts> bye denshi
22:08:55 <jim> heya denshi
22:08:56 <denshi> denshi has quit ("denshi has no reason")
22:09:09 <jim> or, well, sometimes...
22:09:19 <jim> so apparently it's defined somehow
22:09:30 <jim> on the sql side
22:09:49 <k2pts> jim: on postgresql queries or oracle?
22:10:01 <jim> postgres
22:10:29 <denshi> denshi (toddg@linux126.ma.utexas.edu) has joined #openacs
22:10:43 <denshi> much like MacArthur, I have returned.
22:10:52 <denshi> what's needed, jim?
22:11:04 <jim> hiya denshi
22:11:04 <k2pts> jim: could you point me to query (pg) that uses it
22:11:20 <jim> try deleting a wp package instance
22:12:06 <k2pts> there's no decode in the wp-slim-drop.sql script
22:12:17 <jim> denshi: I'm trying to locate what I see as a fairly widespread problem which I initially thought was related to 0or1row, 
22:12:28 <jim> k2pts: package -instance- :)
22:12:35 <k2pts> ok, moment
22:12:41 <jim> you tried to delete a package :)
22:13:16 <denshi> not that much time on my hands atm.  I'll read the logs and check this later.
22:13:21 <denshi> bye again.
22:13:22 <denshi> denshi has quit (Client Quit)
22:13:27 <jim> I'm still hooked on this idea that I can find the problem in a fairly localized mabnner
22:13:42 <jim> err, that didn't make sense
22:14:10 <jim> I'm still hooked on this idea that I can find the problem in one place, or a few places that are fairly close together
22:14:31 <k2pts> yes, it sounds strange because we ported oracle decode with "case when"
22:14:35 <k2pts> in postgresql
22:14:51 <jim> ohh ok
22:15:01 <jim> yes, I saw an instance of that
22:15:07 <jim> and I tried the case
22:15:12 <jim> by itself
22:15:13 <k2pts> aha
22:15:14 <jim> it worked
22:15:47 <jim> what kind of thing is decode?
22:16:35 <k2pts> think of it like the C switch statement (returns a value based on some conditions)
22:16:48 <jim> I think openacs could benefit from a general macro facility that could optionally be applied to queries
22:16:56 <k2pts> e.g. decode (value, 0, 'zero', 1, 'one', 'unknown')
22:17:07 <k2pts> CASE WHEN value=0 THEN 'zero' WHEN value=1 THEN 'one' ELSE 'unknown' END
22:17:14 <jim> ok, I know what decode does, but how/where is it defined?
22:17:25 <k2pts> it's an oracle construct
22:17:34 <jim> yes, I know that also
22:17:43 <jim> what about the pg version?
22:17:50 <k2pts> sorry, my english are not that good...don't get your question
22:17:59 <jim> is it tcl? plpgsql? something else?
22:18:08 <k2pts> we don't use decode in the postgresql version...we use case when as above
22:18:26 <jim> so you replace decode
22:18:29 <k2pts> yes
22:18:35 <k2pts> see example above
22:19:29 <jim> so anything that uses decode in the pg version hasn't been properly ported?
22:20:04 <k2pts> anything that uses decode in the oracle version has been ported to postgresql using "case when...then...else..end'
22:20:25 <jim> so you're saying the port is complete and working
22:20:27 <k2pts> they are "equivalent"
22:20:29 <k2pts> yes
22:20:34 <jim> I'm not sure that's true
22:20:45 <k2pts> if we missed any queries you can submit a patch
22:21:30 <k2pts> don't be confused by the queries in the tcl files, the query dispatcher will select the appropriate query based on the dbms you are using (ie. oracle or postgresql)
22:21:39 <k2pts> from the xql files
22:21:50 <jim> which are in .xql, yes, knew that too
22:21:59 <k2pts> ok
22:23:03 <jim> so, can you create a wimpy, create a presentation, unmount the wimpy and delete it?
22:23:25 <k2pts> haven't tried it...I'll have to do a fresh install to do that...sorry
22:24:08 <jim> (btw, davb and I are working on altering wp so that package instances only see the presentations created using that instance...
22:24:55 <k2pts> sounds good. I applied davb's patches a while ago. you might want to give it a whirl
22:25:05 <k2pts> "whirl" (I hope that's the right word)
22:25:23 <k2pts> is it? :)
22:25:41 <jim> he made a patch that puts the package instance ID in the context ID of the presentation's acs_object row
22:26:10 <jim> yes, "give it a whirl" is correct slang for "try it"
22:26:30 <k2pts> thanks, and yes a package_id is the right thing to do 
22:26:44 <jim> did you see my patch along those lines?
22:26:53 <k2pts> number?
22:27:10 <jim> wait, I just filed the bug, haven't genrated a patch
22:27:14 <jim> Jim Lynch
22:27:21 <k2pts> ah ok. i have only seen the patches...
22:27:30 <jim> lemme find it
22:27:32 <k2pts> got a question for you give a min
22:27:59 <hazmat> hello.
22:28:08 <jim> ok
22:28:17 <jim> hi hazmat
22:28:30 <k2pts> hey hazmat
22:28:38 <jim> I can shoot you a strangely made patch
22:28:48 <k2pts> jim: patch #123, is it to allow html entries
22:29:00 <jim> that's my other one, unrelated to this
22:29:02 <k2pts> please no strange patches, show some mercy :)
22:29:24 <jim> ok, lemme try to generate
22:29:30 <k2pts> jim: patch #123, is it for allowing html entries? (question)
22:29:35 <jim> yes
22:29:38 <k2pts> ok
22:29:51 <k2pts> I'll try to apply it tomorrow, I'm done for this evening 
22:33:03 <k2pts> ok, have to head out guys, later
22:33:14 <k2pts> k2pts has left #openacs