00:07:14 <loggy> loggy has joined #openacs
00:07:14 <niven.openprojects.net> Users on #openacs: loggy davb markd2 rbm strainmaker abbaJ shagster AaronSw GEM Psychephylax chump 
00:07:14 <ChanServ> This channel is logged: http://www.blogspace.com/openacs/chatlogs/ and blogged: http://www.thedesignexperience.org/openacs/ircblog
00:12:53 <Psychephylax> hmmm
00:13:07 <Psychephylax> Anyone know what a majority function with an even number of inputs looks like?
00:13:44 <markd2> huh?
00:13:54 <Psychephylax> uh huh!
00:13:56 <Psychephylax> hehe
00:14:01 <Psychephylax> A majority function
00:14:14 <markd2> what's a majority function?
00:14:17 <Psychephylax> Do you know what the truth table for one with an even number of inputs would look like
00:14:30 <Psychephylax> if you have 3 inputs 
00:14:42 <Psychephylax> and there's 2 True then you don't care what the 3rd one is
00:18:00 <markd2> I'd imagine that a maj func with even # of inputs could be ambiguous
00:19:08 <Psychephylax> That's what I'm thining
00:22:36 <Psychephylax> what about this
00:22:52 <Psychephylax> since 2 and 2...there's not a majority then it's false
00:23:29 <markd2> 100% is a majoity
00:23:40 <Psychephylax> 51% is a majority
00:23:44 <Psychephylax> 50% is not a majority 
00:23:59 <markd2> when you say "2 and 2"
00:24:07 <markd2> is that "two arguments, a 2 and a 2"?
00:24:07 <Psychephylax> 50% and 50%
00:24:15 <Psychephylax> you have this:
00:24:16 <markd2> ah
00:24:18 <Psychephylax> T T T T
00:24:20 <Psychephylax> T T T F
00:24:24 <Psychephylax> T T F T
00:24:25 <Psychephylax> ...
00:24:32 <markd2> yea, TTFF is ambiguous
00:24:37 <Psychephylax> obviously those cases are all T 
00:24:44 <markd2> but I think returning false is wrong
00:24:51 <Psychephylax> right, T F T F, F F T T, T T F F are ambiguous
00:25:02 <Psychephylax> But it's not a majority
00:25:22 <markd2> so, you're returning False, thereby implying that false is in the majority
00:25:37 <Psychephylax> hmmm
00:25:51 <markd2> can you use three-valued logic? 
00:26:00 <Psychephylax> no
00:26:01 <markd2> truse/false/null :-)
00:26:12 <Psychephylax> :)
00:29:07 <markd2> or if you're in java/C++ land, true/false/exception
00:32:04 <Psychephylax> heh
00:39:18 <Psychephylax> gah
00:52:37 <Psychephylax> well chi
00:52:38 <Psychephylax> t
00:52:44 <markd2> f
00:53:24 <Psychephylax> bah
00:53:32 <Psychephylax> It is not in the book or on google
00:54:31 <Psychephylax> How am i supposed to do the prelab if this is not even on google
00:57:52 <markd2> what class?
00:59:58 <Psychephylax> Some bullchit engineering class that has no relevance to my coding abilities
01:07:07 <Psychephylax> well
01:07:11 <Psychephylax> I emailed the professor
01:18:50 <markd2> hopefully this isn't due in the morning
01:31:05 <Psychephylax> no it's due tomorrow at 7pm
01:31:20 <Psychephylax> and my boss won't reply to my e-mail
01:31:33 <markd2> maybe his hair needs sharpening
01:32:56 <Psychephylax> maybe he's just trying to ignore it so I can come in to work tomorrow sick
01:33:01 <Psychephylax> I hope I get him sick back 
01:33:12 <markd2> germ warfare :-)
01:33:46 <Psychephylax> on a tiny scale
01:33:47 <Psychephylax> yes
01:33:51 <Psychephylax> one e-mail down
01:33:53 <Psychephylax> one to go
01:34:02 <Psychephylax> TTFF would go to 0
01:34:42 <Psychephylax> now I can od the prelab
01:34:46 <Psychephylax> at least part of it
01:35:57 <Psychephylax> I wonder if my RF key will still work to get into the building
01:36:21 <Psychephylax> my mom had the benefit of putting it through the laundry
01:36:47 <markd2> probably
01:36:55 <Psychephylax> i hope so
01:36:56 <Psychephylax> lol
01:37:15 <markd2> I've had all sorts of keys and badgers and whatnot in my career, and they stood up to incredible amounts of abuse
01:37:24 <Psychephylax> badgers?
01:37:25 <Psychephylax> rofl
01:37:34 <Psychephylax> Mark had badgers in his pockets!
01:37:48 <Psychephylax> So THAT's where my cookies go
01:40:22 <markd2> heh
01:40:28 <markd2> I'm armed and badgerous
01:42:17 <Psychephylax> rofl
01:42:39 <Psychephylax> I can't wait to see Undeclared
01:42:49 <Psychephylax> from commercials it looks just like college
01:44:40 <markd2> the tv here isn't hooked up anything
01:44:42 <markd2> WHat's Undeclared?
01:46:15 <Psychephylax> It's supposed to be a  "real" look at college
01:46:57 <Psychephylax> You know
01:47:20 <Psychephylax> they don't actually have to add too many jokes to what college life really is like
01:47:23 <Psychephylax> it's already funny
01:47:29 <markd2> yeah
01:47:44 <markd2> off to get some sleep before tomorrow's toil-fest
01:48:07 <Psychephylax> heh
01:48:10 <markd2> markd2 has quit
01:57:32 <davb> davb has quit
07:57:50 <djg> djg has joined #openacs
10:36:33 <PascalS> PascalS has joined #openacs
10:45:10 <ola> ola has joined #openacs
12:05:10 <Psychephylax> yay
12:05:13 <Psychephylax> I'm staying home
12:05:49 <PascalS> eh?
12:07:45 <Psychephylax> i got sick
12:07:52 <Psychephylax> and my boss finally replied
12:08:16 <PascalS> U in the US?
12:09:02 <Psychephylax> yes
12:09:21 <Psychephylax> * Psychephylax goes back to bed
12:09:33 <PascalS> Ah, that explains it then :) When I get sick I just call to say I'm sick and that's it, basically.
12:10:18 <PascalS> * PascalS wishes Psycho to get well soon.
13:02:44 <AaronSw> AaronSw has quit
13:02:59 <AaronSw> AaronSw has joined #openacs
13:03:22 <davb> davb has joined #openacs
13:03:26 <davb> Hi
13:03:36 <PascalS> hi
13:12:54 <AaronSw> hi
13:53:59 <davb> What is the best way to compare two documents, say XML files using AOLserver?
13:54:22 <PascalS> text or content?
13:54:30 <PascalS> I mean tag content?
13:55:01 <PascalS> XML files can be logically the same, but have different whitespace, ordering, etc.
13:55:05 <davb> The whole thing I guess. I want to see if a file has changed. I don't necessarily care what the difference is.
13:55:07 <davb> Aha.
13:55:16 <PascalS> exec diff?
13:55:21 <davb> Well I wouldn't care if the whitespace has changed.
13:55:22 <davb> Ok.
13:55:33 <davb> thanks
13:55:44 <PascalS> or 'if { ![string equal $s1 $s2] } {...}
13:56:18 <PascalS> use diff on files in the filesystem, string equal on tcl strings.
13:57:17 <davb> cool. I am working of hooking up to weblogs.com new update interface. You have to notify them via XMLRPC that your site has been updated.
13:57:35 <davb> And I want to include the chump which is just XML files. No database.
13:58:34 <davb> Or I could hack it into the chump itself. Now thats a cool idea. 
13:59:08 <PascalS> chump thet the thing that goes 'Chump url grabbed!' on this channel?
14:00:10 <davb> yes.
14:00:24 <davb> Its a bot written in python. I got the idea from #rdfig
14:00:46 <davb> I figure its good OpenACS advertising.
14:00:57 <PascalS> * PascalS is checking out weblogs.com
14:05:51 <markd2> markd2 has joined #openacs
14:06:51 <davb> wait... I just rememberd. I want to compare a file with itself. That is the tricky part.
14:07:01 <PascalS> oh.
14:07:04 <PascalS> that's simple
14:07:20 <PascalS> use md5sum to store the previous versions md5 hash
14:07:32 <davb> Aha. Thanks!
14:07:57 <markd2> is there something like pgp for tcl?
14:08:08 <markd2> say I want to encrypt credit card numbers with a public key when going into the db
14:08:20 <markd2> then use my private key to decrypt them before fulfillment
14:08:28 <PascalS> spawning gpg is prolly the simplest way.
14:08:53 <PascalS> read from it's stdout for the result...
14:09:07 <PascalS> gpglib isn't ready yet, I believe.
14:09:39 <markd2> ok
14:09:41 <markd2> thanks
14:09:52 <PascalS> but don't keep the privatekey online on the server :)
14:10:02 <markd2> heh
14:10:24 <markd2> we're going to be doing manual fulfillment, and the plan is to slurp the encrypted keys locally and then turn them into real numbers here
14:10:32 <markd2> the thought of storing raw credit card #'s makes me nervous
14:10:35 <PascalS> okay, sounds good.
14:11:24 <PascalS> you could also attach a line printer :)
14:11:40 <markd2> heh
14:11:46 <markd2> the colo place might object to the noise
14:11:54 <markd2> a mom-and-pop.  the pet turtles there might complain
14:12:14 <PascalS> <grin>
14:13:02 <davb> I thought there was an MD5 module for aolserver around somewhere
14:13:12 <PascalS> probably
14:13:27 <PascalS> it's a really simple algorithm.
14:14:10 <PascalS> Infact, use the SHA module.
14:14:19 <davb> Aha. Cool. 
14:14:20 <davb> I have that.
14:14:31 <PascalS> Like MD5 but the hash values are bigger
14:14:42 <davb> Probably not a problem for me.
14:14:55 <PascalS> They're not that big.
14:19:36 <davb> anyone have rbm's script to make an aolserver 3.3 install for OpenACS 4? 
15:01:49 <PascalS> PascalS has quit
15:28:11 <Psychephylax> * Psychephylax climbs out of bed to peek 
15:28:20 <Psychephylax> I feel worse than death
15:28:26 <Psychephylax> i hate my boss
15:28:49 <markd2> * markd2 hads Psyche some dayquill
15:29:13 <Psychephylax> hmmmg
15:29:15 <Psychephylax> good idea
15:29:22 <Psychephylax> i'll go pick that up when I go get a money order
15:34:47 <Psychephylax> my boss let me stay home
15:34:56 <Psychephylax> Now I have to focus on being sick on Wed so I get him back
15:36:19 <markd2> breathe into a lot of plastic bags
15:39:45 <Psychephylax> Eh?
15:40:58 <markd2> keep the germs
15:41:03 <Psychephylax> I don't want to keep them
15:41:05 <markd2> bring them in of wednesday even if you're feeling better
15:41:11 <Psychephylax> you never know when they'll get mad at you and lash out
15:41:21 <markd2> by that time you'll be immune
15:41:28 <markd2> ... unless they mutate ...
15:41:33 <Psychephylax> They always mutate
15:41:36 <markd2> then you can hit your boss with mutagerms!
15:41:51 <Psychephylax> hmm
15:42:00 <Psychephylax> I wonder if they will ever release SNL on DVD
15:46:21 <davb> That would be cool.
15:48:16 <Psychephylax> actually it seems as they did
15:48:31 <Psychephylax> at least parts of it...mostly titled "SNL: Best of ____"
15:48:40 <Psychephylax> like Mike Meyers and stuff
15:51:33 <Psychephylax> http://www.amazon.com/exec/obidos/tg/stores/detail/-/dvd/B00004ZEU5/tech-info/104-4433847-7857567
15:51:33 <chump> A: http://www.amazon.com/exec/obidos/tg/stores/detail/-/dvd/B00004ZEU5/tech-info/104-4433847-7857567 from Psychephylax
15:51:47 <Psychephylax> A:| Monty Python DVD Collection - a must get!
15:51:48 <chump> titled item A
15:52:12 <Psychephylax> A: Comedy, comedy and more comedy...
15:52:12 <chump> commented item A
15:52:18 <Psychephylax> Ministry of Silly walks!
15:56:32 <markd2> oh no.  I might have to get a dvd player now.
15:56:37 <markd2> complete set of The Prisoner. http://www.amazon.com/exec/obidos/ASIN/B00005NKCQ/qid=1001951824/sr=2-2/ref=sr_2_7_2/107-5916530-0810938
15:56:47 <Psychephylax> heh
15:56:57 <Psychephylax> I got a DVD player for the 1st season of the simpsons
15:58:32 <markd2> I was waiting for B5 to come out on dvd
15:58:43 <Psychephylax> i think it's going to or it might have
15:59:06 <markd2> it's not out yet
15:59:13 <markd2> the "in the beginning" move might be out in december
16:00:05 <Psychephylax> heh
16:00:10 <Psychephylax> * Psychephylax is not a big B5 fan
16:00:21 <markd2> bummer
16:01:03 <Psychephylax> nothing beats Star Trek classic
16:01:24 <markd2> true
16:01:36 <Psychephylax> damn it
16:01:37 <markd2> but it was cool seeing a coherent story unfold over 5 years
16:01:45 <Psychephylax> I want my name to time stamp too ;(
16:01:54 <Psychephylax> heh
16:02:01 <Psychephylax> Star Trek ruled though
16:02:11 <Psychephylax> 1 hour stories are much easier to follow than a 5 year one 
16:02:13 <markd2> * markd2 hands psyche a roll of time stamps
16:02:20 <Psychephylax> oh boy!
16:04:39 <Psychephylax> what to do...what to do...
16:07:20 <Psychephylax> D'oh!
16:07:27 <Psychephylax> SNL tickets are only sold in August
16:07:52 <Psychephylax> i hope they have a 28th season
16:08:11 <Psychephylax> i wanna go 
16:08:46 <Psychephylax> * Psychephylax crawls in his bed again...I'll be back later
16:16:58 <markd2> markd2 is now known as seen
16:17:03 <seen> seen is now known as markd2
16:29:13 <shagster> shagster has quit
16:29:54 <shagster> shagster has joined #openacs
16:34:25 <markd2> markd2 has quit
16:47:22 <davb> why would an update on one row in a tiny little table take say, forever, on postgresql?
17:21:51 <rbm> something wrong?
17:24:07 <rbm> * rbm nods
17:28:08 <davb> hi
17:30:41 <davb> this is my query: update cr_keywords set heading='News', description='The New News Description'
17:30:48 <davb> where keyword_id = '2457'; 
17:31:12 <davb> When I try that postgresql just hangs. I tried in psql and AOLserver.
17:31:41 <rbm> have you checked your processes and available RAM to see if PG has room to di its job?
17:32:05 <davb> Nope. I'll check.
17:33:14 <davb> plenty of RAM
17:35:08 <rbm> define "plenty"
17:35:34 <rbm> You might want to do some tunning in your PG startup options
17:35:37 <davb> I am using 83m 100 free
17:36:02 <davb> Probably.
17:36:21 <rbm> WAIT! Why are you putting "keyword_id='2457'"? PG is probably doing a cast for every single row. You should do "keyword_id=2457" (no quotes)
17:36:38 <rbm> "83m 100 free" makes no sense to me
17:37:10 <davb> sorry. I have 192MB of RAM in this machine. 83MB is used. 100MB or so is free. So I am using less than half.
17:37:23 <davb> Aha. That is the bind variable substituion doing that.
17:37:30 <davb> (not is psql I just copied the query)
17:39:10 <davb> I took out the quotes, but it didn't seem to help. There are only 5 rows in the table anyway.
17:45:01 <davb> I think this is it: 
17:45:01 <davb> create trigger cr_keywords_update_tr after update on cr_keywords
17:45:01 <davb> for each row execute procedure cr_keywords_update_tr ();
17:45:09 <rbm> * rbm goes out for a bit
17:45:17 <davb> davb has quit
17:47:56 <davb> davb has joined #openacs
17:52:34 <davb> Here is the offending trigger:
17:52:35 <davb> create function cr_keywords_update_tr () returns opaque as '
17:52:35 <davb> begin
17:52:35 <davb>   if old.parent_id <> new.parent_id then
17:52:35 <davb>     update acs_objects set context_id = new.parent_id
17:52:35 <davb>       where object_id = new.keyword_id;
17:52:37 <davb>   end if;
17:52:39 <davb>   return new;
17:52:41 <davb> end;' language 'plpgsql';
18:02:30 <markd2> markd2 has joined #openacs
18:06:37 <davb> I removed the trigger and it still is stuck.
18:06:56 <davb> maybe it's just REALLY slow.
18:07:43 <markd2> what's up?
18:08:11 <davb> I am trying to update one row in a 5 row table and postgresql is just sitting there. It never returns.
18:08:38 <davb> It had a trigger after update so I got rid of it, but that didn't help.
18:12:41 <markd2> any other session have a lock on that row?
18:12:46 <markd2> (not sure about how PG does locking)
18:14:50 <davb> I can't imagine. Noone else is using the server.
18:15:03 <davb> I wonder if there is a way to check...
18:15:26 <davb> I rebotted the server just in case.
18:17:03 <markd2> in oracle-land there's all sorts of data dictionary tables to see who has what locked and why
18:17:08 <markd2> PG probably has something similar
18:18:30 <talli> talli has joined #openacs
18:18:35 <talli> hey guys
18:19:00 <markd2> yo
18:19:58 <davb> hi
18:23:27 <AaronSw> yoyo
18:25:03 <davb> Here is my query plan
18:25:04 <davb> openacs-4=> EXPLAIN update cr_keywords set description='Description' where keyword_id=2457;
18:25:04 <davb> NOTICE:  QUERY PLAN:
18:25:04 <davb> Seq Scan on cr_keywords  (cost=0.00..0.00 rows=1 width=39)
18:25:04 <chump> Label NOTICE not found.
18:25:11 <davb> Sounds pretty easy to me.
18:28:02 <davb> OK. I have an idea.
18:29:24 <davb> A PLSQL block is a transaction. The CR_KEYWORDS API has a seperate call for set_description and set_heading. If I call both of them inside a PLPGSQL block it will never complete.... thats my theory anyway. But it doesn't explain why it doesn't work when I do it all in one update.
18:31:35 <davb> weird. Its just that one ROW that is causing the problem.
18:32:52 <davb> Ok. Here is the REAL problem. I have keyword_id and parent_id. If the keyword_id of a keyword is in the parent_id column of another keyword, the update hangs. Otherwise it works.
18:34:13 <davb> gotta go.
18:34:20 <davb> davb has quit
19:57:07 <davb> davb has joined #OpenACS
20:41:07 <djg> djg has quit
21:32:27 <ola> ola has quit
21:37:55 <talli> talli has quit
21:40:28 <markd2> markd2 has quit
23:13:54 <GEM> GEM has quit