IRC log of openacs on 2002-03-05

Timestamps are in UTC.

00:00:14 [jim]

is --with-tcl necessary for oacs-4.5b?

00:01:13 [rbm]

I don't think so.

00:01:31 [jim]

cool...

00:02:02 [jim]

thanks :)

00:02:36 [til]

maybe search needs --with-tcl

00:03:00 [rbm]

Yeah, I was just thinking of that. OpenACS 3.x searching needs PL/Tcl, not sure about 4

00:03:34 [shagster]

well, I think 3 only needed pltcl for the bboard search

00:03:37 [til]

i think i heard something like "the new openfts needs --with-tcl" recently, dont know if "new" means the one that is currently used

00:05:27 [jim]

say I wanted to use 4.2.0 final with pg-7.2... anything like that possible?

00:05:50 [rbm]

jim: s/4.2/4.5/, yes.

00:06:17 [rbm]

jim: http://www.brasileiro.net:8000/ is running openacs 4.5 pre-beta on PG 7.2

00:06:50 [denshi]

denshi has quit (Remote closed the connection)

00:07:50 [jim]

oop, it made an error when I tried to look at your publ bookmarks :)

00:08:12 [talli]

what's the difference between pre-beta and alpha?

00:08:29 [shagster]

a few charactors?

00:08:31 [shagster]

:)

00:10:20 [rbm]

jim: Hmmm, maybe I forgot to restart after I enabled the modules

00:10:22 [davb]

talli: alpha was a release. we are installing on the CVS bleeding edge.

00:10:24 [rbm]

jim: hang on

00:10:55 [rbm]

talli: pre-beta is what's on CVS now and includes everything that has been changed since alpha was released

00:14:02 [denshi]

denshi (~chatzilla@adsl-216-62-223-193.dsl.rcsntx.swbell.net) has joined #openacs

00:19:37 [denshi]

talli, if I ever say anything about language design again, feel free to smack me.

00:24:08 [talli]

denshi, how come?

00:24:29 [talli]

were you put in your place by some 30 year hacker who told you about the culled features of fortran?

00:25:03 [talli]

"denshi, when will you learn? we dropped that from cobol cuz it kept us from drawing bunny rabbits on punch cards!"

00:25:08 [denshi]

start here:

00:25:09 [denshi]

http://www.paulgraham.com/thist.html

00:25:09 [chump]

A: http://www.paulgraham.com/thist.html from denshi

00:25:37 [denshi]

A: Olin Shiver's history of T, the first production-level lisp with lexical scoping.

00:25:37 [chump]

commented item A

00:26:00 [denshi]

then follow the relevant disserations produced by that.

00:26:40 [denshi]

at the same time, read what you can find about the implementation of lisp interpreters. Try not to freak out.

00:27:28 [denshi]

link A details a summer project that directly created *5* disserations and 1 master's thesis.

00:28:32 [denshi]

just reading this stuff is a fabulous lesson in recursion; given the number of phrases that are undefined to the layman.

00:29:35 [talli]

ok, i've hit my limit

00:29:38 [talli]

what the hell is the point?

00:29:53 [talli]

(and my mom is calling me to dinner ;))

00:30:09 [denshi]

the point is when you go far beyond that limit, and see what top-notch nerds can do.

00:30:14 [talli]

ah

00:30:15 [talli]

ok

00:30:24 [talli]

so?

00:30:30 [talli]

are you not a top notch nerd?

00:30:31 [denshi]

It's humbling, and grates at me how much extra education they have had available to them.

00:30:39 [denshi]

I'm just bitter today.

00:30:42 [talli]

ah, ok

00:30:56 [talli]

be back after dinner

00:30:58 [denshi]

maybe I'll be a top notch nerd in a few years.

00:31:01 [denshi]

have fun.

00:31:03 [talli]

i'll have a schnitzel for you denshi

00:31:19 [denshi]

save me some matzolah soup.

00:54:21 [rbm]

anybody tried to setup cvsweb with oacs 4?

00:54:40 [talli]

rbm, why not try setting up subversion?

00:54:55 [rbm]

talli: Because I don't have root on this machine

00:55:03 [talli]

it seems to be pretty stable now. at least, if you're used to playing oacs early, it might be cool

00:55:03 [talli]

oh

00:55:23 [rbm]

and because I need a web interface to CVS, not a whole other software management system :0

00:55:31 [talli]

ah

00:55:48 [talli]

rbm, do you have root access to teh OACS server?

00:56:15 [rbm]

and because I don't have python installed here, otherwise I'd use viewcvs, whose code I can actually read, unlike this piece of snot that cvsweb's perl code is.

00:56:38 [talli]

jim: did you hear that? rbm is dissing perl, AGAIN!

00:56:39 [talli]

go at him

00:56:42 [rbm]

talli: Through sudo, yes

00:56:58 [rbm]

talli: I said cvsweb's perl code, not perl code in general.

00:57:00 [talli]

who's got root?

00:57:31 [rbm]

ben, don and I, I think.

00:57:41 [talli]

ah, ok, so you do have it

00:58:03 [rbm]

yeah, through sudo.

00:58:04 [talli]

what is the difference between having root and having it through sudo?

00:58:15 [talli]

meaning you don't have a serial connectivity?

00:58:48 [rbm]

with sudo you can restrict which commands someone can execute as root. In our case, Ben just gave us full root access through sudo. The advantage is that he doesn't have to tell us the root password for us to become root.

00:59:00 [talli]

ah, ok

00:59:15 [talli]

well, the reason i ask is that the new openacs.org really shuold be sitting on the new box

00:59:28 [talli]

no reason to do dev on it on our box and hten have to move everything

00:59:33 [talli]

might as well move it now

00:59:41 [rbm]

Do you want to move the new oacs.org to openacs.org? I'm good with that.

01:00:00 [talli]

no, i think we should move it tentatively to something like new.openacs.org

01:00:12 [rbm]

We'd have to ask Ben to make a DNS entry for new.openacs.org or whatever we'll call it.

01:00:12 [talli]

then, when it's all good and ready just make it live

01:00:17 [talli]

ok

01:00:25 [talli]

whatever

01:01:10 [rbm]

There would be some logistical problems too because you'd need to be able to restart the aolserver process, but since it'll probably be running on port 8000 anyways, that shouldn't be much of a problem.

01:01:47 [til]

staging.openacs.org ? that way it could be useful later too ...

01:02:05 [talli]

rbm, do you think there would be interest in having OACS participate in the dotgnu project?

01:02:24 [rbm]

talli: I don't see why not

01:02:37 [rbm]

I haven't really looked at dotgnu yest

01:02:47 [talli]

there's not much to look at, really

01:02:55 [talli]

except portable.net, which is rocking the house

01:03:08 [talli]

it's one guy hacking in australia and he's keeping up with mono

01:03:26 [talli]

he had a six-month headstart, but he also bootstrapped totally on linux

01:03:32 [talli]

mono bootstrapped on win

01:03:45 [rbm]

Have you read "Hackers of Our Lives"?

01:03:46 [denshi]

cool.

01:03:48 [talli]

portable.net is written in C, mono is written in C#

01:04:01 [talli]

no, what's that rbm

01:04:26 [rbm]

http://www.advogato.org/article/430.html

01:04:26 [chump]

B: http://www.advogato.org/article/430.html from rbm

01:04:43 [talli]

the reason i think that being involved in dotgnu might be cool is to benefit from the community of similar minded projects trying to coordinate web services

01:04:44 [rbm]

B:|"Hacks of Our Lives" Soap-Opera

01:04:44 [chump]

titled item B

01:05:17 [talli]

i guess. but i'm kind of in the camp that doesn't trust miguel

01:05:22 [talli]

or, rather, ximian

01:06:09 [denshi]

what's the crossover between ximian & the original gnome guys?

01:07:12 [rbm]

talli: Have you read the above yet?

01:07:19 [talli]

i did read it

01:07:30 [talli]

before

01:08:06 [denshi]

because I don't know if any of you were punching out bugs in gnome back in '98, but gnome still remains in my memory as one of the most hideous pieces of software I have ever seen the source to.

01:09:12 [denshi]

oops.

01:09:16 [denshi]

s/(\w+\ )[5]^/to which I have seen the source/

01:09:32 [denshi]

I need little grammar regexes to follow me around.

01:10:32 [talli]

i don't know how much VC money ximian got, but they sure act like they got a lot

01:10:43 [talli]

at least, they had one of the biggest spaces at linuxworld

01:11:14 [talli]

and red carpet acts a lot like some of the other more obnoxious update tools

01:11:25 [rbm]

I don't know how ximian plans to make money

01:11:44 [talli]

good point

01:11:53 [rbm]

The only thing I know they sell is the adpator for exchange

01:11:58 [talli]

i think the VCs are going to decide that

01:12:02 [talli]

yeah :(

01:12:08 [rbm]

and the faster access to red carpet, but I don't anyone who likes it.

01:13:06 [talli]

i'm not into those pay-2-update services, at least i'm willing to take the extra time to learn debian (or BSD) to avoid them

01:13:29 [talli]

i was going to use up2date, until i was blocked by redhat because of "peak download time"

01:14:49 [talli]

i suppose it might make sense if i was a biznarse that had a bunch of servers and just HAD to pay someone

01:15:10 [denshi]

I just can't take Mono seriously.

01:16:05 [talli]

denshi, what're your feelings about it?

01:16:16 [rbm]

I still haven't drank the entire .NET kool-aid, so I feel misinformed to comment.

01:16:42 [denshi]

1) Gnome sucks. I mean really sucks.

01:17:22 [denshi]

1.a) they reinvented many things, and for specious reasons. It didn't really pay off.

01:18:12 [denshi]

1.b) it's still sucking at high speed, but after 5? years it's pretty decent. But that's no excuse for 5 years.

01:18:59 [denshi]

1.c) I'm still amazed that Gnome exists; that the OSS community could flag the Qt/KDE issue into such a flame fest that Gnome was taken seriously.

01:19:22 [denshi]

1.d) check that -- Gnome was started intentionally *because* of the Qt license issue.

01:19:24 [rbm]

I'm not really familiar with gnome internals, but I see that they reinvented a lot of stuff in the form of libraries instead of just doing "the unix way" of re-using little tools doing each a small job

01:19:39 [denshi]

rbm: you'd really rather not know.

01:20:16 [talli]

i think that's why they're so into .NET

01:20:18 [denshi]

So, in 1, my summary is that, in my experience, the Gnome guys are both incompetent and motivated by political reasons.

01:20:23 [rbm]

denshi: So what's with gnome internals? I know this one guy in Brazil who hangs out in #debian-br who keeps telling me Gnome is beautiful

01:20:58 [denshi]

yes, I think it's very cute that they're into .NET -- given that they release MS-quality stuff, they make an excellent match.

01:21:01 [rbm]

I don't know if the Gnome developers (besides Miguel) have anything to do with .NET and mono

01:21:35 [denshi]

rbm: that's why I was asking what the crossover is. It's a sincere question in my attempt to determine if Mono is for real.

01:21:38 [talli]

well, Mono is a ximian sponsored project

01:21:56 [denshi]

But I have more reasons:

01:22:53 [denshi]

2) after reading several disserations last month about optimizing lisp compilers, I have enormous respect for the supernerds who can code serious runtimes.

01:23:10 [denshi]

2.a) I don't think I'll find much in common between them and Gnome.

01:23:33 [rbm]

talli: But Gnome isn't

01:23:56 [talli]

but ximian-Gnome is

01:23:57 [denshi]

and finally, 3) Is it just me, or is the common runtime completely blown out of proportion?

01:24:12 [talli]

and ximian-gnome probably has a pretty strong influence on the rest of Gnome

01:24:42 [talli]

i have a feeling that a common runtime, if it ever really works, will lead to some of the worst, most unreadable code in history

01:25:11 [denshi]

talli may be right. But what I'm getting at is that language interoperability has been around for a long time.

01:26:58 [talli]

CLI might be cool. but i think you would need some really heavy, hardcore programmers to be able to build a system with it well

01:27:54 [denshi]

for instance, scheme48's module system could take an argument of the parser used to read in a file; so you could use several languages which would all be parsed into a common sexpre format internally.

01:28:14 [denshi]

and everyone in creation know how to link to libraries in different languages.

01:29:35 [talli]

MS has some really smart guys, so i imagine a CLI might be pretty cool for many applications.

01:29:56 [talli]

and denshi, does anyone use Scheme48 anymore?

01:30:07 [denshi]

I just don't know what you can do with CLI that you can't do now. Mostly I'm guessing that the win here is to pass around objects in a common binary standard, so that many languages can funge with the same spot of memory. I don't think that's a good thing.

01:31:08 [talli]

how come?

01:32:24 [denshi]

well, isn't that the opposite of OO design that the entire industry has bought into?

01:34:43 [talli]

denshi, if you're interested in playing around with the internals of a new language, check out the portable.net project

01:34:48 [talli]

some serious work is going into that

01:34:51 [talli]

they need help, too

01:35:07 [denshi]

url?

01:35:13 [talli]

but from what i hear, both mono and portable.net (perhaps even MS's CLR) are like two years from ready

01:35:19 [talli]

http://www.dotgnu.org

01:35:19 [chump]

C: http://www.dotgnu.org from talli

01:35:32 [talli]

C: the dotGNU project

01:35:32 [chump]

commented item C

01:35:46 [talli]

http://www.southern-storm.com.au/portable_net.html

01:35:46 [chump]

D: http://www.southern-storm.com.au/portable_net.html from talli

01:35:54 [talli]

D: Portable.NET project homepage

01:35:54 [chump]

commented item D

01:36:27 [talli]

oh, and the FAQ for pnet actually deals with technical matters, rather than spin

01:36:38 [denshi]

wtf is up with all these groups implementing C#?

01:36:49 [talli]

dunno

01:36:55 [talli]

but i do hear that C# is really cool

01:37:36 [talli]

i know someone that is working on an ASP.NET project, though, and he says that with the whole .net thing MS is just "catching up"

01:37:51 [talli]

i asked him was ASP.NET is, and apparently it's ASP but more "OO"

01:37:57 [talli]

which is, uh, gross

01:38:53 [denshi]

C# is tiny tweak off of Java. If all these people want a Java clone, but more open (since C# is being submitted to ISO), then I could get that. But it doesn't make any sense; that such a wide swatch of the OSS community would, for purely technical reasons, choose this same language as the focus point for a CLI.

01:39:10 [talli]

yeah, that's a good point

01:39:27 [talli]

i still buy rolf's point about if you want to use Java, why not use C++?

01:39:33 [talli]

at least that is proven tech

01:40:40 [denshi]

You'd expect to see people doing it in MLs, in Scheme, in Hackell/C++ for gods sake.

01:41:16 [talli]

the scary thing is that most uni's are now teaching java in the data structures classes

01:41:54 [talli]

now, i wasn't such a big fan of perl (but i'm not a big fan of programming) for my net programming class, but java seems a bit much

01:42:19 [talli]

maybe they do it so their students can put java on their resumes

01:42:29 [talli]

ten bucks that's why my school does it now

01:42:31 [denshi]

that and the professors.

01:42:41 [talli]

the profs being bought, you mean?

01:42:42 [denshi]

remember, not everyone gets tenure.

01:42:49 [talli]

oh yeah, good point

01:43:41 [talli]

but at tufts, the tenured professors teach the intro level classes as well.

01:44:01 [talli]

the non-tenured teachers are the ones in the backrooms doing research and slaving away

01:44:11 [talli]

s/teachers/professors

01:44:14 [denshi]

I don't understand undergraduate CS. What can you teach for data structures in Java that you can't pick up on the job?

01:44:35 [talli]

lots of people take data structures, at least at tufts, that are not CS majors

01:44:46 [talli]

so even though i bombed it, it was a good intro

01:45:26 [talli]

actually, i didn't bomb it. i just hated it so much i would fall asleep trying to study it and so could never absorb what the hell a stack was. i really hated that shit

01:46:02 [denshi]

oh, here's a good lesson in data structures:

01:46:03 [denshi]

http://www.paulgraham.com/onlisptext.html

01:46:03 [chump]

E: http://www.paulgraham.com/onlisptext.html from denshi

01:46:19 [denshi]

E: Paul Graham's "On LISP", online.

01:46:20 [chump]

commented item E

01:46:28 [denshi]

Read chapter 6 for an amazing insight.

01:47:04 [talli]

how's his new language coming along?

01:47:12 [talli]

have you ever taken a look at REBOL?

01:47:17 [talli]

http://www.rebol.com

01:47:17 [chump]

F: http://www.rebol.com from talli

01:47:31 [talli]

F: new language developed by Carl Sassenrath (sp?)

01:47:31 [chump]

commented item F

01:47:42 [talli]

F: unfortunately, not free (as in beer or speech)

01:47:42 [chump]

commented item F

01:48:12 [talli]

it's supposed to be mondo cool

01:48:16 [denshi]

rebol's been around for a while.

01:48:25 [talli]

and that carl dude is supposedly a genius, as he build the amigaOS

01:48:30 [denshi]

it's only cool b/c it has nice pronouns.

01:48:39 [talli]

oh yeah? you're not a dan?

01:48:40 [talli]

fan

01:48:44 [talli]

how about curl?

01:49:19 [denshi]

* denshi snickers

01:49:25 [talli]

whoops

01:49:29 [denshi]

they just went out of business, if you didn't notice.

01:49:35 [talli]

no way!

01:50:00 [talli]

uh, you sure?

01:50:12 [talli]

they just announced that they raised 7 million dollars

01:50:35 [talli]

of course, they sold their soul to do it

01:51:23 [talli]

where did you see they went out of business?

01:51:46 [denshi]

sorry, my bad; it was just the CEO leaving.

01:51:51 [denshi]

fuckedcompany.

01:51:54 [talli]

ah

01:52:05 [talli]

yeah, a whole bunch of them left

01:52:10 [talli]

and were replaced by VCs

01:52:16 [talli]

so... they're doomed

01:52:34 [denshi]

they're doomed because there's no way the language will catch on.

01:53:03 [talli]

maybe it's because while their language is supposed to deliver content, their site is written using JSPs

01:53:17 [denshi]

it's not open, there's no competition to push the language, and where's the incubation time? Languages take forever to catch on.

01:53:23 [talli]

yeah

01:53:27 [talli]

what about rebol?

01:53:51 [talli]

seems you might need the size and cache of MS or Sun to be able to build and make money off of a language

01:54:19 [talli]

and java's taken 5 years and billions of dollars to catch on

01:54:26 [talli]

and it still doesn't wrok

01:54:27 [talli]

work

01:55:11 [denshi]

it's been a while, but I thought the win with rebol was it has a small api crowded with networking features.

01:55:27 [talli]

yeah, i think that's it

01:58:39 [davb]

davb has quit (Remote closed the connection)

02:00:10 [davb]

davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs

02:03:20 [davb]

denshi do you read: http://www.treedragon.com/ged/map/ti/new.htm

02:03:31 [davb]

he has some stuff to say about .NET and CLR

02:03:36 [denshi]

never heard of it.

02:03:47 [davb]

and he is writing his own language :)

02:04:01 [davb]

let me find the CLR stuff

02:04:02 [denshi]

do you read narbonic.com?

02:04:06 [davb]

nope.

02:04:09 [talli]

don't look at this guys blog!

02:04:17 [talli]

he links to aaron swartz!!!

02:04:20 [talli]

stay away!!!

02:04:47 [davb]

starting here can't hurt: http://www.treedragon.com/ged/map/ti/newFeb02.htm#27feb02

02:04:51 [denshi]

talli has a point.

02:04:57 [talli]

haha

02:05:15 [talli]

where is that little bugger?

02:05:20 [talli]

he never comes around anymore

02:05:22 [davb]

cool!

02:05:29 [davb]

he is busy changing the world without us!

02:05:43 [talli]

bastard

02:06:42 [talli]

paje: aaronsw

02:06:43 [paje]

talli: i'm not following you...

02:06:51 [talli]

paje: seen aaronsw?

02:06:51 [paje]

I haven't seen 'aaronsw', talli

02:07:49 [denshi]

I think he's busy working on taking over his school.

02:08:34 [talli]

that kid needs some more homework

02:08:43 [davb]

heh

02:08:49 [denshi]

...and a concealed weapons permit.

02:08:55 [denshi]

funny!

02:09:13 [talli]

haha

02:09:29 [davb]

anyway, denshi, that guy has some great ideas on programming and langauages.

02:09:33 [talli]

and some platform shoes

02:09:39 [davb]

heh

02:10:52 [talli]

denshi, you have any experience with smalltalk?

02:11:15 [denshi]

none at all.

02:11:45 [talli]

oh. i thought that was the ultimate OOP language

02:12:37 [davb]

it depends on who you ask I think :)

02:12:46 [talli]

i ask denshi!!!!

02:13:22 [davb]

so today we found out which all the stuff we neglected to prepare for the data coversion and had a panicked afternoon.

02:13:27 [talli]

haha

02:13:39 [talli]

you mean, like how you neglected to PLAN

02:13:40 [talli]

??

02:13:43 [davb]

heh

02:14:09 [davb]

not my fault, noone listens to me. we did recover and made it back to where we thought we were at 10am when I said, oh shit.

02:15:31 [davb]

thing is, noone there understands how a relational database works, not even the alleged consultant.

02:15:40 [talli]

nice.

02:15:52 [talli]

you're using access, though, right?

02:15:53 [davb]

he's a nice guy anyway.

02:15:55 [davb]

yes.

02:16:04 [denshi]

orbitz.com sucks.

02:16:10 [davb]

so we don't actually _have_ a real relational database... :)

02:20:01 [davb]

btw, an excellent web based application is the US Dept. of Ed FAFSA form. it is easy to use and hard to screw up.

02:20:44 [davb]

uhoh: Title: Java Applet Can Redirect Browser Traffic

02:20:44 [davb]

Date: 04 March 2002

02:20:44 [davb]

Software: Microsoft Virtual Machine

02:20:44 [davb]

Impact: Information Disclosure

02:20:44 [davb]

Max Risk: Critical

02:20:45 [davb]

Bulletin: MS02-013

02:23:54 [shagster]

What, can error in a Microsoft product? Nah...don't believe it :)

02:24:04 [talli]

shagster!

02:24:11 [shagster]

Not after they spent the month of Febuary fixing it!

02:24:15 [shagster]

talli!

02:24:16 [talli]

you lousy bastard

02:24:22 [denshi]

l8r guys

02:24:24 [shagster]

I'm still writing the specs

02:24:30 [talli]

ah, neat

02:24:31 [talli]

:)

02:24:32 [talli]

later denshi

02:24:37 [denshi]

denshi has quit ()

02:24:47 [shagster]

talli, you don't know me well enough to call me that -yet- :P

02:25:19 [talli]

i'm trying to get in as many shots as i can before you take my balls

02:25:28 [shagster]

Heh :)

02:26:42 [shagster]

Okay, I think either the end of March of start of April I'm going to try and have a Cleveland, Ohio OpenACS social

02:27:45 [shagster]

Since somebody was nice enough to make a yearly donation for uptime, I have a few bucks to put towards the bar bill ^H^H^H^H^H^H expenses

02:30:46 [rafa]

rafa (rafa@ebt.ee.usyd.edu.au) has joined #openacs

02:30:57 [rafa]

hi

02:31:11 [shagster]

hi

02:31:51 [rafa]

any nes about the 4.5b release?.

02:32:16 [shagster]

Don is working on it, 'bout the only think I know

02:33:40 [rafa]

ok, thanks

02:34:05 [shagster]

You might check the bboards for more info though

02:34:12 [shagster]

I'm about a day behind right now

02:44:18 [rbm]

moo

02:45:12 [shagster]

oom

02:48:49 [rbm]

(__)

02:48:49 [rbm]

(oo)

02:48:49 [rbm]

/------\/

02:48:49 [rbm]

/ | ||

02:48:49 [rbm]

* /\---/\

02:48:49 [rbm]

~~ ~~

02:48:51 [rbm]

...."Have you mooed today?"...

02:49:41 [shagster]

well, that is a personal question! :)

02:53:27 [rbm]

paje: apt?

02:53:27 [paje]

rbm: wish i knew

02:54:05 [rbm]

paje: apt is the Advanced Package Tool, an ultra-cool and very nice package management front-end that works with .deb's and .rpm's

02:54:05 [paje]

OK, rbm.

03:28:36 [davb]

hmmmmm

03:28:46 [davb]

ns_returnnotfound is crashing AOLserver without an error in the log

03:34:03 [davb]

we need to fix the "you have to reboot every time you change something" MS-like behavior of OpenACS

03:34:49 [davb]

ok ns_returnnotfound is definitely crashing aolserver without any sort of nice message.

03:35:56 [davb]

hmmmm. I wonder what ns_returnnotfound is doing.

03:40:46 [markd2]

markd2 (~Snak@r-41.81.alltel.net) has joined #openacs

03:44:38 [davb]

hi markd2

03:45:02 [markd2]

yo

03:45:32 [davb]

can you think of any reason why calling ns_returnnotfound from a tcl file would crash aolserver with no errors or anything in the log?

03:46:02 [markd2]

bad mojo crashing aolserver from tcl

03:46:13 [markd2]

I'd run aolserver in gdb and look at the stack trace

03:46:14 [davb]

yeah.

03:46:15 [talli]

someone say mojo?

03:46:17 [markd2]

there may be something Obvious there

03:46:21 [davb]

ah cool. thanks.

03:46:24 [markd2]

tailli has mega mojo

03:46:25 [talli]

* talli hides his markd2 voodoo doll

03:46:31 [davb]

* davb looks to see if gdb is installed

03:46:38 [markd2]

mojo argito, meester roboto

03:47:02 [talli]

markd2, how's life outside of nebraska?

03:47:44 [markd2]

cold!

03:47:52 [markd2]

it was like 11 on the way over

03:47:56 [markd2]

gtg. catch folks tomorrow

03:47:58 [markd2]

markd2 has quit ("brrrrr")

03:48:00 [talli]

late mar

03:48:45 [davb]

ok.

03:48:46 [til]

til has quit (Read error: 113 (No route to host))

03:55:50 [davb]

argh

04:01:00 [davb]

hmmm gdb seems to crash aolserver too!

04:07:46 [davb]

anyone know how to use gdb? it seems to not like me.

04:10:22 [davb]

nm

04:10:31 [davb]

I got it to work, gdb that is

04:11:14 [davb]

hmmm: Program received signal SIGSEGV, Segmentation fault.

04:11:14 [davb]

[Switching to Thread 4101 (LWP 3347)]

04:11:14 [davb]

0x08116ce9 in TclReExec (re=0x876358c, string=0x8729e8c, len=62,

04:11:14 [davb]

details=0x87635b4, nmatch=0, pmatch=0x876cd88, flags=0)

04:11:14 [davb]

at ./../generic/regexec.c:171

04:11:15 [davb]

171 int flags;

04:12:38 [davb]

hmmm. maybe my aolserver is broken...

04:13:39 [davb]

* davb rebuilds aolserver

04:19:47 [davb]

ok still crashes. weird.

04:42:31 [dlk]

dlk (dlk@as2-1-4.va.g.bonet.se) has joined #openacs

04:43:28 [davb]

davb has quit ("Client Exiting")

04:50:39 [andyn]

andyn (~andy@12-254-190-230.client.attbi.com) has joined #openacs

06:15:27 [dlk]

dlk has quit (Read error: 113 (No route to host))

06:37:57 [andyn]

andyn has quit (Remote closed the connection)

06:37:57 [talli]

talli has quit (Read error: 104 (Connection reset by peer))

06:56:59 [andyn]

andyn (~andy@12-254-190-230.client.attbi.com) has joined #openacs

10:37:49 [talli]

talli (~talli@talli.dialup.access.net) has joined #openacs

11:38:49 [k2pts]

k2pts (~nkd@adsl-168-174.cytanet.com.cy) has joined #openacs

11:39:19 [k2pts]

paje: seen talli?

11:39:19 [paje]

talli was last seen on #openacs 7 hours, 51 minutes and 19 seconds ago, saying: late mar [Mon Mar 4 20:50:00 2002]

11:46:10 [talli]

talli has quit (Read error: 104 (Connection reset by peer))

11:49:54 [k2pts]

paje: where is nowhere?

11:49:54 [paje]

k2pts: wish i knew

11:50:08 [k2pts]

paje: what's your favorite song?

11:50:08 [paje]

bugger all, i dunno, k2pts

11:50:26 [k2pts]

paje: do you like chicks?

11:50:26 [paje]

no idea, k2pts

11:50:50 [k2pts]

paje: insult paje :)

11:50:51 [paje]

paje :) is nothing but a festering petrification of lumpish snake assholes.

11:51:18 [k2pts]

paje: paje: paje: insult paje

11:51:19 [paje]

k2pts: i'm not following you...

11:51:43 [k2pts]

paje: insult paje: insult paje: insult paje

11:51:43 [paje]

paje: insult paje: insult paje is nothing but a wretched half-mouthful of coughed-up chicken piss.

11:52:06 [k2pts]

paje: insult I am

11:52:06 [paje]

I am is nothing but a fen-sucked ass-full of fermented rat retch.

11:52:31 [k2pts]

paje: insult chump

11:52:31 [paje]

chump is nothing but a spongy pile of festering gunk.

11:52:38 [k2pts]

paje: insult chump:

11:52:39 [paje]

chump: is nothing but a spam-sucking quart of tickle-brained Sun IPC manuals.

11:52:39 [chump]

Not understood: is nothing but a spam-sucking quart of tickle-brained Sun IPC manuals.

11:52:50 [k2pts]

paje: insult chump: paje

11:52:50 [paje]

chump: paje is nothing but a fulminating puddle of unintelligent poopy.

11:52:51 [chump]

Not understood: paje is nothing but a fulminating puddle of unintelligent poopy.

11:53:39 [k2pts]

chump: say hi to paje:

11:53:39 [chump]

Not understood: say hi to paje:

11:53:47 [k2pts]

loggy: hi

11:54:15 [k2pts]

paje: seen loggy?

11:54:15 [paje]

loggy was last seen on #openacs 4 days, 12 hours, 42 minutes and 17 seconds ago, saying: 0) 2002-02-28 23:10:31 <denshi> loggy, what's new in the blog today? [Thu Feb 28 16:13:58 2002]

11:54:29 [k2pts]

loggy, what's new?

11:55:28 [loggy]

I'm logging. I found 12 answers for 'what's new' (showing 0...4)

11:55:29 [loggy]

0) 2002-03-05 11:54:29 <k2pts> loggy, what's new?

11:55:30 [loggy]

1) 2002-03-05 11:54:15 <paje> loggy was last seen on #openacs 4 days, 12 hours, 42 minutes and 17 seconds ago, saying: 0) 2002-02-28 23:10:31 <denshi> loggy, what's new in the blog today? [Thu Feb 28 16:13:58 2002]

11:55:31 [loggy]

2) 2002-02-28 23:10:31 <denshi> loggy, what's new in the blog today?

11:55:32 [loggy]

3) 2002-02-27 21:13:32 <denshi> what's new, talli?

11:55:33 [loggy]

4) 2002-02-13 01:09:18 <denshi> what's new with you?

11:55:45 [k2pts]

paje: seen davb?

11:55:45 [paje]

davb was last seen on #openacs 7 hours, 35 minutes and 2 seconds ago, saying: ok still crashes. weird. [Mon Mar 4 21:21:47 2002]

11:55:45 [k2pts]

paje: bye

11:55:54 [paje]

bye k2pts

11:56:13 [k2pts]

paje: insult: loggy, paje: insult: loggy

11:56:47 [paje]

k2pts: huh?

11:57:07 [k2pts]

paje: insult loggy, paje: insult loggy, ..

11:57:28 [paje]

loggy, paje: insult loggy, .. is nothing but a milk-livered gob of imp-bladdereddle-headed dog vomit.

11:57:28 [paje]

I'm logging. I don't understand 'paje: insult loggy, .. is nothing but a milk-livered gob of imp-bladdereddle-headed dog vomit.', paje. Try /msg loggy help

11:58:26 [k2pts]

k2pts has left #openacs

12:18:19 [jim]

jim has quit ("[x]chat")

13:14:04 [dlk]

dlk (dlk@walter.ita.chalmers.se) has joined #openacs

13:35:09 [davb]

davb (~dave@rrcs-nys-24-97-22-203.biz.rr.com) has joined #openacs

13:41:27 [davb]

I created a general comment on March 2, and it is still in the database. I will test some more.

13:41:48 [davb]

but the chump is still not udating nightly. weird.

13:46:29 [k2pts]

k2pts (~nkd@adsl-168-174.cytanet.com.cy) has joined #openacs

13:47:21 [davb]

hi k2pts

13:47:26 [k2pts]

hey davb

13:47:27 [k2pts]

the problem with general_comments is in the bboard_garbage_collect_postgresql function

13:47:49 [k2pts]

both bboard messages and general comments are stored as object_type='acs_message'

13:47:51 [davb]

ah, I don;'t have bboard installed :)

13:48:22 [k2pts]

they are deleted whenever bboard_garbage_collect is executed (once a day by default)

13:48:27 [davb]

so the objects types are not used correctly.

13:48:34 [k2pts]

exactly

13:48:42 [k2pts]

have you made any progress with RSS

13:48:46 [davb]

excellent.

13:48:58 [davb]

no, we are doing a big data conversion over here at work.

13:49:13 [k2pts]

ok, have you seen my comments about the favorites issue

13:49:29 [davb]

I think so, but I seem not to remember :)

13:49:48 [k2pts]

basically you don't want to use the categorization package for favorites, IMO

13:49:54 [k2pts]

you need a separate package

13:50:02 [davb]

ok.

13:50:24 [k2pts]

which could use categories but for categorization -- not for maintaining each user's favorites (you'll need a separate table for that)

13:50:35 [k2pts]

like general-comments

13:50:39 [davb]

right. that is what I meant :)

13:50:43 [k2pts]

ok

13:50:44 [k2pts]

:)

13:51:04 [k2pts]

how are things?

13:51:55 [davb]

not bad. my site is mostly working under OpenACS 4. I am still way behind on the secret project though because every time that guy changes his mind, I don't feel like working on it and I have to get remotivated all over again.

13:52:07 [k2pts]

:)

13:52:08 [k2pts]

heh

13:52:37 [davb]

for some reason the same copy of ns_xml I was using with the openacs 3 site stopped working for xslt.

13:52:56 [k2pts]

it works for me

13:53:11 [davb]

for example: [05/Mar/2002:08:53:37][10232.5126][-sched:7-] Error: unknown command

13:53:11 [davb]

unknown command

13:53:11 [davb]

while executing

13:53:12 [davb]

"ns_xml parse_xslt $xslt_file"

13:53:17 [k2pts]

have you compiled it to work with xslt...it's not set by default...you have to edit the Makefile

13:53:32 [davb]

yes. I was using the XSLT under the old site :).

13:53:34 [k2pts]

I can send you mine if you want to

13:53:35 [davb]

for chump.

13:53:37 [k2pts]

ah ok

13:53:46 [k2pts]

that's strange

13:54:01 [davb]

yes.

13:54:43 [davb]

k2pts: maybe not a bad idea.

13:54:47 [k2pts]

heh

13:55:26 [davb]

this is just weird. I can't imagine why the xslt commands are missing.

13:55:41 [k2pts]

moment

13:56:45 [k2pts]

are you sure you compiled nsxml with xslt on?

13:56:53 [davb]

yes.

13:56:57 [k2pts]

I mean are you using the same installation as before?

13:57:20 [davb]

yes. i did not change the aolserver stuff at all. I double checked all the paths etc...

13:57:29 [k2pts]

strange

13:57:40 [k2pts]

reboot, it works with windows :)

14:00:24 [k2pts]

davb: I have to head out...I'll log in later this evening...

14:00:32 [davb]

ok bye.

14:00:52 [k2pts]

we can talk about the categorization and RSS package when you are done with the data migration

14:01:02 [davb]

cool.

14:01:09 [k2pts]

great

14:01:10 [k2pts]

l8r

14:01:14 [k2pts]

k2pts has left #openacs

14:13:48 [talli]

talli (~talli@lti-4.dialup.access.net) has joined #openacs

14:29:59 [k2pts]

k2pts (~nkd@adsl-168-174.cytanet.com.cy) has joined #openacs

14:44:54 [k2pts]

k2pts has left #openacs

14:55:53 [markd2]

markd2 (~markd2@r-41.16.alltel.net) has joined #openacs

15:23:44 [davb]

ok. I am seriously annoyed.

15:23:52 [davb]

ns_xml is acting up.

15:24:34 [davb]

I wonder if my libxslt is broken or something

15:30:49 [davb]

ok, I don't have time, I'll have to go insane later...

15:40:33 [talli]

is it good when you ask someone for a reference for an accountant, and they tell you, "this guy's good. he's very liberal, if you know what i mean" ?

15:40:50 [markd2]

uhhh.....

15:41:04 [talli]

that's my feeling too

15:43:56 [markd2]

I want my accountants to be as conservative as possible

16:06:42 [dlk]

dlk has quit (Read error: 110 (Connection timed out))

16:38:15 [andyn]

andyn has quit (Remote closed the connection)

16:38:37 [andyn]

andyn (~andy@12-254-190-230.client.attbi.com) has joined #openacs

17:48:49 [talli]

IT'S SO QUIET IN HERE!!!

17:48:55 [rbm]

* rbm makes some noise

17:49:04 [talli]

that's better

17:50:45 [markd2]

* markd2 squeals like a pig

17:51:59 [talli]

* talli runs in circles, skipping like a donkey and paddling his own behind

17:52:42 [markd2]

* markd2 sighs

17:52:45 [markd2]

not again!

17:52:51 [markd2]

* markd2 reaches for his banjo

17:59:02 [hazmat]

.seen k2pts

18:02:29 [markd2]

seen k2pts

18:02:41 [markd2]

paje, seen kp2ts

18:02:41 [paje]

I haven't seen 'kp2ts', markd2

18:03:52 [talli]

he was here earlier

18:04:02 [talli]

but his pops came and he had to go "offline"

18:07:57 [markd2]

pops?

18:11:17 [talli]

soda pops

18:11:24 [markd2]

ah

18:11:37 [markd2]

didn't know if that was some weird Cyprus thing involving parents or food or something

18:11:59 [talli]

it was his fahdah, soda pops

18:12:55 [hazmat]

mirror, mirror, on the wall when will oacs4 be released? :-)

18:12:55 [rbm]

* rbm makes a Debian package

18:13:46 [rbm]

<with heavy accent> Master, patient you must be. But using the CVS version, no problems there will be.

18:14:09 [talli]

</with heavy accent>

18:14:22 [rbm]

* rbm notes talli is not in IE mode :)

18:14:40 [talli]

i didn't want to sounds like austrian weight lifter for the rest of the day

18:15:41 [rbm]

I love that halls commercial "Brrreath my frrriend, brrreath!"

18:15:57 [rbm]

* rbm wonders if he spelled that right

18:16:01 [rbm]

paje: spell breath

18:16:02 [paje]

rbm: 'breath' may be spelled correctly

18:16:24 [markd2]

breathe is the variant you're looking for

18:16:32 [rbm]

markd2: Aha! Thanks :)

18:17:00 [markd2]

my speling and grammer was so bad in skoo, I be surprized I can do it so goodly now

18:17:34 [rbm]

Someday you guys should stop by #linux on fslc.usu.edu. There's a guy there that mispells 13 out of 15 words. It's amusing.

18:18:02 [talli]

you mean, "eetz umuzing"

18:18:12 [markd2]

heh

18:18:48 [talli]

haha

18:19:24 [talli]

i can't spell in hackerese yet. this bothers me

18:19:38 [rbm]

talli: ph33r my sp33l1ng!

18:20:02 [talli]

hmm... maybe it shouldn't

18:20:24 [rbm]

talli: I don't really know how to spell in hackerese, and personally I don't like it ;)

18:20:57 [talli]

but i want to be 133t!!!

18:22:10 [markd2]

t4LL1 w1L n3V3r b3 L3eT H4><()R!!!

18:22:12 [hazmat]

* hazmat sprinkles pixie dust on talli

18:22:58 [markd2]

u 5huL c|-|4T w1F Rz0Lph m()r3. h3 b3 l33t

18:23:49 [rbm]

SEGFAULT: core dumped

18:23:49 [chump]

Label SEGFAULT not found.

18:27:54 [markd2]

heh

18:28:01 [markd2]

I have learned at the feet of masters

18:31:56 [k2pts]

k2pts (~nkd@adsl-168-174.cytanet.com.cy) has joined #openacs

18:31:57 [shagster]

shagster has quit (Read error: 104 (Connection reset by peer))

18:34:44 [hazmat]

hi k2pts

18:35:01 [k2pts]

hey hazmat

18:37:27 [rbm]

paje: seen k2pts?

18:37:28 [paje]

k2pts was last seen on #openacs 2 minutes and 27 seconds ago, saying: hey hazmat [Tue Mar 5 11:37:00 2002]

18:37:34 [rbm]

:)

18:37:36 [andyn]

andyn has quit (Remote closed the connection)

18:37:48 [k2pts]

hey rbm

18:37:57 [k2pts]

I was playing with paje earlier this afternoon

18:38:24 [rbm]

k2pts: Oh really? What did you think? I'm going to install more factoid packs some time so he can become a little smarter.

18:38:33 [rbm]

s/some time/sometime/

18:38:46 [k2pts]

I was trying to let the bots fight, e.g.

18:39:00 [k2pts]

paje: insult loggy, paje insult

18:39:01 [paje]

loggy, paje insult is nothing but a spongy plate of thick seagull puke.

18:39:01 [paje]

I'm logging. I don't understand 'paje insult is nothing but a spongy plate of thick seagull puke.', paje. Try /msg loggy help

18:39:21 [k2pts]

paje: insult chump: insult paje

18:39:21 [paje]

chump: insult paje is nothing but a spongy plate of off-color anal warts.

18:39:21 [chump]

Not understood: insult paje is nothing but a spongy plate of off-color anal warts.

18:39:58 [davb]

k2pts: can you please email me your nsxml.so that works with xslt? I seem to have broken mine. I think it might be my version of libxslt so I want to test to see if the one you compiled will work.

18:40:16 [k2pts]

davb: sure give me a moment

18:42:46 [davb]

thanks alot!

18:43:18 [k2pts]

davb: just sent nsxml.so, ns_xml.c, Makefile, Readme

18:43:23 [k2pts]

aolserver-3.3.1ad13

18:43:25 [davb]

excellent.

18:43:27 [davb]

me too.

18:43:47 [k2pts]

compile under [aolserver-src-]root/

18:44:01 [k2pts]

/usr/local/src/root/nsxml/....

18:44:10 [davb]

ok cool.

18:44:53 [davb]

bbl

18:45:23 [k2pts]

ok, dave, bye

19:17:09 [hazmat]

offtopic lars has a nice write up about his experience with ad at pinds.com

19:18:58 [k2pts]

k2pts has left #openacs

19:38:35 [hazmat]

* hazmat is about to make himself unpopular again.

19:38:43 [talli]

no more zope

19:38:47 [talli]

go away zope.

19:38:52 [hazmat]

don brought it up,

19:38:55 [talli]

haha

19:38:57 [talli]

where?

19:38:58 [hazmat]

and with fud

19:39:50 [talli]

but...

19:39:54 [talli]

zope doesn't scale well

19:40:03 [talli]

and zeos is not a very good answer

19:40:25 [hazmat]

bullshit

19:40:36 [hazmat]

zope has more public high profile sites than openacs.

19:40:42 [talli]

like?

19:40:55 [hazmat]

and in what way does standard zope scale well.

19:41:03 [talli]

i don't know one zope site that handles heavy traffic

19:41:09 [hazmat]

openacs/acs clustering has been done by how many people?

19:41:11 [talli]

cbsnewyork is the only one

19:41:18 [hazmat]

www.activestate.com

19:41:20 [hazmat]

for one

19:41:21 [talli]

well, no one HAS to cluster oacs

19:41:38 [hazmat]

really have you load tested openacs4?

19:42:00 [talli]

no. but if it didn't work, then you just use AOLserver

19:42:01 [hazmat]

zope is faster in my testing.

19:42:11 [hazmat]

aolserver is different than openacs.

19:42:40 [talli]

but it's a component

19:43:09 [talli]

and i'm not sure that activestate is such a heavy site

19:43:28 [rbm]

Yeah! My first real Debian package lives!!! <evil laugh>AHAHAHAHAHAHAH</evil laugh>

19:43:29 [talli]

nor do i know of many other sites that are particularly complex running zope

19:44:17 [talli]

and i'm not buying anything about zope as long as mailing lists and yahoo groups are used for collaborative development

19:44:25 [talli]

especially if its a collaborative server

19:44:53 [hazmat]

bah.. mailing lists are how 99% of the open source world develops... so take your beef to every other project regarding that.

19:45:11 [hazmat]

s/develops/developers

19:45:24 [hazmat]

activestate is a farily high traffic site.

19:45:25 [talli]

every other project isn't a collaborative server for web application environments

19:45:27 [hazmat]

uniontrib.com

19:46:18 [hazmat]

yes, but most developers prefer mailing lists.

19:47:03 [hazmat]

the only reason the openacs uses them is toot their own horn imo, esp. considering the *delete* level of collaborative development because most dev takes place by *email* anyways.

19:47:04 [talli]

and wikis, too, right?

19:47:19 [talli]

what's your point?

19:47:41 [hazmat]

wikis are a preference thing. they are easy collaborative documentation.

19:47:48 [hazmat]

personally i'm not a big fan.

19:48:05 [talli]

no one here has any problem with you in OACS, nor with zope (beyond the fact that i give you shit for it)

19:48:09 [hazmat]

my point is that don's post was FUD>

19:48:11 [talli]

but that post you made was totally unnecessary

19:48:13 [talli]

bullshit

19:48:29 [talli]

no one posts anything liek that to the mailing lists, do they?

19:48:44 [talli]

you just spouted the typical zope response

19:48:50 [talli]

we have zclasses

19:48:58 [hazmat]

no, but they don't make unsubstantiated claims about other platform's scalability unless they themselves have used it.

19:49:10 [talli]

bullshit

19:49:19 [talli]

listen, if you have a problem, don't listen to the boards

19:50:09 [talli]

you didn't respond to the issue of object-relational mapping and how it's pretty much fact that they don't scale well or work as well as straight procedural-relational systems

19:50:18 [talli]

you went in with zope propoganda

19:50:37 [hazmat]

no i went in to clear up statements that zope was unscalable that where vague and general.

19:50:38 [markd2]

zopaganda

19:50:49 [talli]

aD built some pretty good sites with ACS4.2, like the world bank, which is huge and quite complex

19:51:07 [talli]

OACS isn't even released yet so only people with smaller sites have been using it

19:51:09 [hazmat]

yourpoint?

19:51:34 [andyn]

andyn (~andy@12-254-190-230.client.attbi.com) has joined #openacs

19:51:41 [talli]

your post is as much FUD as don's may have been, but don has the right because he runs the community

19:51:47 [talli]

he knows what he is doing.

19:51:59 [talli]

that wasn't an invitation for a zope flame war, although you set it up for one

19:52:31 [talli]

that's not good community participation, that's just flamebait

19:52:58 [talli]

if you post your tests of the scalability of zope vs OACS, then that would be something worth reading

19:53:01 [hazmat]

your right. the last line was uncalled for... i'm writing a qualification for it right now.

19:53:21 [talli]

but going into another community and lighting up one of the leaders for his FUD with more FUD is not good

19:54:38 [hazmat]

so here's my additional response, comments welcome.

19:54:41 [hazmat]

to add a qualification to my response. i realize that the openacs is unreleased and that the last line of my response was uncalled for. but i disliked the unsubstantiated claims about zope's scalability that were both vague and general.

19:55:07 [hazmat]

+ my apolgies to anyone who might have been offended.

19:55:23 [talli]

it's not like you called anyone a nazi

19:55:41 [talli]

if you feel that strongly, why not use the test you made between zope and OACS4?

19:56:03 [talli]

don said in his post he's willing to be proven wrong

19:56:20 [hazmat]

its not a very self contained test, it was between a development version of the python respository in the openacs and zope.

19:56:43 [hazmat]

there were lots of additional factors that don't make it good for a general test case imo.

19:56:56 [hazmat]

but for me it was a real world test.

19:57:08 [talli]

but you were using python with OACS?

19:58:29 [hazmat]

no (hadn't gotten to that part), the majority of the site was all in tcl.

19:58:44 [talli]

how about thi

19:58:45 [talli]

this

19:58:47 [hazmat]

python was only going to be used for webservices.

19:59:05 [talli]

i'm willing to set up a test of a tuned OACS4 system vs a tuned Zope system to put all of this shit to rest

19:59:08 [hazmat]

and doing dependency graphs, and some file manip.

19:59:55 [hazmat]

hmm... now thats interesting.

20:01:00 [hazmat]

but the thing about zope is that its an app development platform, so every site is going to be different under the hood. which depending on the skill of the implementer can make a marked difference to the performance.

20:01:25 [hazmat]

openacs is more about taking a community framework and building on top of it.

20:01:28 [talli]

there ya go

20:02:02 [talli]

an application development system is much different than a database system

20:02:03 [rbm]

hazmat: I think OpenACS is a web development framework, with community features built-in for you.

20:02:13 [rbm]

not the other way around.

20:02:19 [hazmat]

true.

20:03:15 [hazmat]

but can you separate out the community features from the web development framework... its pretty grounded throughout the toolkit., of course that brings up the question of what is a web dev platform.

20:03:56 [hazmat]

to be honest, as a web development platform i think the openacs is a bit inflexible, as a community systems its fantastic.

20:04:46 [talli]

that's the rub, and the point that needs to be made

20:04:58 [talli]

zope has more flexibility, but one needs to know it through and through, as you do

20:05:15 [talli]

but that flexibility does come at a price, which is scalability. hence zeos

20:05:20 [hazmat]

to me its all about integration capabilities.

20:05:26 [talli]

if you know zeos, maybe it's worth it

20:05:37 [hazmat]

i'm never had to use zeo for scalability.

20:05:39 [talli]

most people in the OACS community don't want to come near software clusters

20:05:53 [hazmat]

i disagree wrt to the scaling though.

20:06:10 [hazmat]

don gave me numbers which put the openacs at around 12-15hits/s

20:06:20 [hazmat]

which zope is easily capable of .

20:06:24 [hazmat]

think about it.

20:06:34 [hazmat]

in openacs you have to make round trips to the db for every query

20:07:00 [hazmat]

incurring i/o overhead and rdbms overhead.

20:07:33 [hazmat]

in zope all your data is local, a good deal of it will be in object cache if the object has been accessed recently so all your iteration is local.

20:07:50 [hazmat]

plus add in the native indexing/search capabilities with zcatalog.

20:08:12 [talli]

ok

20:08:22 [talli]

what if, using zope, you need to do a complex query?

20:08:34 [hazmat]

such as

20:08:59 [talli]

you're stuck either with zclasses or hacking some of your own code

20:09:09 [hazmat]

you can use the catalog to index properties (and methods) of a zope object.

20:09:27 [talli]

and zclasses do not guarantee nearly as good performance as straight sql

20:09:30 [k2pts]

k2pts (~nkd@adsl-168-174.cytanet.com.cy) has joined #openacs

20:09:31 [hazmat]

you can make the query in a python script, and do any filtering and post/pre processing you want.

20:09:55 [rbm]

My only problem with Zope is that it looks like the implementors went crazy to worship the OO gods. I love OO but it's not an answer to all problems, in fact it has its own problems when used incorrectly.

20:09:56 [hazmat]

which can be more natural than sql.

20:10:15 [k2pts]

Check the openfts-general mailing list for some discussion between zope's indexing/search and openfts...I'm sorry I can't stay and continue this discussion, later Kapil

20:10:22 [k2pts]

k2pts has left #openacs

20:10:29 [hazmat]

will do

20:13:34 [rbm]

hazmat: RDBMS or non-RDBMS have its advantages and disadvantages. I personally think RDBMS approach is much more scalable.

20:14:14 [hazmat]

rdbms approaches are more portable and openaccess definitely

20:14:23 [hazmat]

more proven as well.

20:15:31 [hazmat]

awesome, someone added support for snowball stemmers to openfts.

20:15:47 [hazmat]

that gives like 11 language stemmers to openfts.

20:20:32 [hazmat]

but wrt to rdbms vs. non-rdbms, its a bit of a non issue for me, as long as i can convert data between them as nesc. to me its *all* about two things integration with external systems, and speed of development.

20:23:27 [davb]

does openacs have anything yet to integrate external systems?

20:23:29 [rbm]

right, you can do that with both systems (I presume). Although there are _a lot_ of tools to deal with integration to RDBMSs, and a general much larger know-how. Homever, scalability and reliability are important.

20:23:46 [rbm]

davb: define external systems.

20:24:08 [davb]

* davb defers to hazmat :) I am not sure what he meant.

20:24:34 [hazmat]

* hazmat is wondering why donb referenced zope at all in that bboard thread...

20:25:57 [rbm]

paje: seen jim?

20:25:57 [paje]

jim was last seen on #openacs 20 hours, 18 minutes and 7 seconds ago, saying: oop, it made an error when I tried to look at your publ bookmarks :) [Mon Mar 4 17:09:50 2002]

20:26:05 [davb]

looks like he was just comparing systems that have an abstract persistance layer kinda thingy...

20:26:17 [davb]

but without any actual numbers.

20:26:40 [markd2]

having nursed many systems back to health that have collapsed under load, I don't care how fast the development was

20:26:41 [hazmat]

wrt to integrataion to external systems you can integrate most things with an rdbms, but again your developing your own custom infrastructure to do it, as opposed to natively talking to the external system. for instance doing auth AND permission mapping from an ldap directory. or talking corba. or talking xml documents, or talking to matlab, etc...

20:26:42 [rbm]

I think he just tossed something around, not with the intention of dissing zope.

20:26:52 [markd2]

it doesn't matter if it melts down in the real world

20:26:59 [rbm]

There was a HUGE thread about zope about 1 or 1.5 years ago on the bboards.

20:27:04 [hazmat]

scalability is something needs to be understood at designed time.

20:27:15 [hazmat]

er.. at least taken into consideration.

20:27:24 [hazmat]

which is factored into dev. time

20:27:52 [hazmat]

trying to add in scalabiity after the fact is a hack.

20:27:58 [hazmat]

imo

20:28:35 [shagster]

shagster (~mkovach@63.90.248.161) has joined #openacs

20:29:05 [hazmat]

rbm: i've read that thread, and talked to some of the particpants (chrisMc and albert langer)...

20:29:59 [hazmat]

about the thread.

20:30:55 [rbm]

I don't have any experience with Zope internals to comment on it. The most information I have about it is that thread and a couple articles I've read.

20:31:51 [hazmat]

ic

20:32:29 [rbm]

The article on last month's LJ showed Zope's DTML and I didn't like it (which says nothing about Zope's internals or how it scales or not).

20:33:04 [hazmat]

those articles are pretty useless imo. ZPT would be a more interesting tech. to cover.

20:33:15 [hazmat]

actually there was an article in DrDobbs about it.

20:33:23 [rbm]

what's that?

20:33:30 [hazmat]

dtml isn't anything special..

20:33:35 [rbm]

ZPT I mean

20:34:17 [hazmat]

ZPT is zope presentation templates... its a fairly clean separation of design and content that allows designers to use their normal tools golive/dreamweaver/frontpage.

20:34:26 [hazmat]

to develop templates.

20:34:52 [hazmat]

the transformation is based on a xml namespace based attribute language.

20:35:55 [hazmat]

which because it doesn't introduce special tags, means that every page is viewable in a browser in its raw format.

20:38:16 [hazmat]

wrt to that original thread, lots of things have changed on the zope side of the equation.. for one the documentation has gotten loads better, there are about a half-dozen books in print, with several online books that take people through learning zope.

20:40:40 [rbm]

"xml namespace based attribute language" wow, introduce "Java" or ".NET" and your phrase would be fully buzzword-compliant :)

20:41:08 [davb]

its simpler than that. it is a really nice design.

20:41:10 [hazmat]

perhaps ;) , it is what it is.

20:42:09 [hazmat]

* hazmat decides thats enough zopaganda for at least a month :-)

20:42:32 [rbm]

hazmat: So what brought you to OpenACS land?

20:43:34 [hazmat]

originally... wow, way back when i was trying to decide on a platform for a school portal for students that was to replace a bunch of aging perl scripts backed by dbm files.

20:43:59 [hazmat]

i evaluated zope, openacs, enhydra, midguard (and raw php)...

20:44:27 [hazmat]

i eventually choose zope, but i was fascinated by the (open)acs and read through the panda books

20:44:35 [hazmat]

i came out west after school.

20:44:44 [hazmat]

and there was a local ad office nearby where i live.

20:45:07 [hazmat]

so i went to a bootcamp, and was really impressed by the vision and the people (aure made a big impression on me).

20:46:06 [hazmat]

i got a contract job for doing oracle/aolserver work at dreamworks about the same time acs4 was coming out and championed its usage at the (actually a subsidary by the name of pop)

20:46:35 [hazmat]

s/at the/for the site

20:46:41 [rbm]

dreamworks uses AOLserver?

20:47:02 [hazmat]

they used to for some of their sites. no longer.

20:47:06 [rbm]

hazmat: I, like you, was also brought to ACS by the vision.

20:47:25 [hazmat]

yup, the vision was what caught me.

20:47:52 [hazmat]

they stopped development with acs and aolserver because hiring people was a bitch, very hard to find qualified people at the time.

20:48:26 [hazmat]

i was contracting at the time, i didn't want to go full time because i'm not a big fan of the entertainment industry... lots of perks though..

20:49:39 [hazmat]

eventually they shifted the site and development over to perl, with template toolkit and mysql as a caching layer in front of oracle, it was pretty slick.. same arch. they use at citysearch.

20:50:53 [rbm]

now there's an interesting thought... MySQL as a caching layer ahead of Oracle...

20:50:57 [rbm]

How did they do that?

20:52:19 [hazmat]

mysql can run in memory. moved all the frequently used queries and data to mysql. i didn't stay around much longer after the switch to perl, so the exact nature of it is unknown to me.

20:54:16 [davb]

http://www.joelonsoftware.com/articles/NothingIsSimple.html

20:54:17 [chump]

G: http://www.joelonsoftware.com/articles/NothingIsSimple.html from davb

20:54:22 [davb]

G:|Nothing is Simple

20:54:22 [chump]

titled item G

20:54:39 [davb]

G:"If you've spent more than 20 minutes of your life writing code, you've probably discovered a good rule of thumb by now: nothing is as simple as it seems."

20:54:40 [chump]

commented item G

20:54:49 [rbm]

So they had one "pool" for SELECTs and another "pool" for DML?

20:55:19 [rbm]

PostgreSQL can cache a lot of stuff in memory too.

20:57:24 [talli]

talli has left #openacs

20:57:48 [hazmat]

rbm: this was with apache

20:58:05 [hazmat]

rbm: i believe the use of the cache was more transparent to the end programmer.

21:00:34 [hazmat]

but this wasn't a simple cache, they needed a fast sql interface to commonly accessed data, since mysql can run entirely in mem, and is pretty fast, it was an appropriate choice, imo.

21:03:38 [davb]

http://www.adaptiveobjectmodel.com/

21:03:39 [chump]

H: http://www.adaptiveobjectmodel.com/ from davb

21:03:45 [davb]

H:|MetaData and Adaptive Object-Model Pages

21:03:46 [chump]

titled item H

21:03:49 [rbm]

Yeah, as long as all the persistent data and transactions were done in Oracle

21:10:50 [davb]

http://www.oreillynet.com/pub/a/javascript/synd/2002/03/01/css_layout.html

21:10:50 [chump]

I: http://www.oreillynet.com/pub/a/javascript/synd/2002/03/01/css_layout.html from davb

21:11:01 [davb]

I:|Introduction to Javascript

21:11:02 [chump]

titled item I

21:11:04 [davb]

ack

21:11:13 [davb]

I:|Introduction to CSS Layour

21:11:14 [chump]

titled item I

21:11:15 [davb]

argh

21:11:17 [davb]

I:|Introduction to CSS Layout

21:11:18 [chump]

titled item I

21:14:59 [davb]

yeah!

21:15:10 [davb]

ns_xml appears to be fixed.

21:21:52 [davb]

ok ns_xml works, but it still hates my stylesheet.

21:21:53 [davb]

hmmmm.

21:23:09 [hazmat]

davb: you could try a command line tool that might give better diagnostics while developing (saxon comes to mind).

21:23:23 [hazmat]

does libxslt have a command line front end?

21:23:25 [davb]

cool, i'll have to try that.

21:23:28 [davb]

yes. xsltproc.

21:23:35 [davb]

It gave me a different error :)

21:23:46 [davb]

I am annoyed because it worked last week,

21:24:16 [hazmat]

* hazmat was about to recommend komodo but realized its no longer free.

21:24:38 [hazmat]

komodo has a cool xslt debugger... i think its based on some apache stuff

21:26:17 [davb]

thanks.

21:26:21 [hazmat]

yes.. its appears to be based on xerces (syntax checks) xalan (iterative debugging)

21:27:12 [davb]

oops: sltproc: Symbol `xmlDefaultSAXHandler' has different size in shared object, consider re-linking

21:28:06 [davb]

oops, time to go home. thanks for the help.

21:28:23 [davb]

davb has quit ()

22:21:33 [davb]

davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs

22:23:16 [davb]

it looks like the encoding error is coming from libxslt, but I don't see it with xsltproc.

22:23:42 [markd2]

markd2 has quit ("Bork")

22:31:59 [hazmat]

davb: do you have an encoding set on the stylesheet?

22:41:39 [davb]

* davb checks

22:42:35 [davb]

in the <?xml ?> part? there is this: <xsl:output indent="yes" encoding="ISO-8859-1" method="html" />

22:42:59 [davb]

in the xml tag there is no encoding. but there is in an XML file it is trying to include.

22:43:18 [hazmat]

try adding an encoding to the xml part

22:43:26 [davb]

ok cool. thanks for the advice.

22:49:18 [davb]

http://www.zveno.com/courses/samples/XML-For-Managers/

22:49:18 [chump]

J: http://www.zveno.com/courses/samples/XML-For-Managers/ from davb

22:49:31 [davb]

http://www.zveno.com/courses/samples/XML-App-Dev-Tcl/

22:49:31 [chump]

K: http://www.zveno.com/courses/samples/XML-App-Dev-Tcl/ from davb

22:49:43 [davb]

J:|XML for Managers, Free Online Course

22:49:43 [chump]

titled item J

22:50:08 [davb]

K:| XML Application Development Using Tcl, Free Online Course

22:50:08 [chump]

titled item K

22:56:42 [andyn]

andyn has quit (Read error: 104 (Connection reset by peer))

23:00:28 [andyn]

andyn (~andy@12-254-190-230.client.attbi.com) has joined #openacs

23:26:30 [davb]

ok I cheated. I took the encoding part out of the nav.xml

23:27:29 [davb]

hmm. still didn't build the missing pages.

23:31:58 [talli]

talli (~talli@lti-4.dialup.access.net) has joined #openacs

23:45:16 [paje]

paje has quit ("regrouping; bbiab")

23:47:24 [xemacs5]

xemacs5 (~chatzilla@user39.net262.oh.sprint-hsd.net) has joined #openacs

23:51:08 [xemacs5]

xemacs5 has quit (Remote closed the connection)

23:53:50 [davb]

ok, i am reindexing the chump files.

23:54:59 [talli]

davb, is there a way of suggesting tighter RDF integration with the CR, or is that kind of a silly idea

23:55:41 [hazmat]

RDF integration as in a source of , or as in a store for?

23:55:57 [hazmat]

it might be nice to export dublin core metadata from the CR

23:56:07 [talli]

hazmat, exactly

23:56:24 [talli]

how can we encourage the dublic core metadata throughout the OACS?

23:58:15 [talli]

* talli is also discussing this with aaronsw in #swhack

23:58:16 [hazmat]

i haven't looked at the cr in a while, but someof it is in the cr by default, creation, author, content type, title, keywords.. what other stuff did you have in mind?

23:58:44 [talli]

that stuff