IRC log of fenfire on 2005-04-08

Timestamps are in UTC.

02:31:11 [mudyc_]
mudyc_ has joined #fenfire
02:48:05 [rubberpaw]
rubberpaw has quit
02:48:13 [mudyc]
mudyc has quit
04:03:22 [rubberpaw]
rubberpaw has joined #fenfire
06:11:33 [vegai]
06:11:48 [tuukkah]
yes, scary
06:14:33 [majukati]
06:21:06 [tuukkah]
huomenta majukati
06:22:00 [vegai]
have to start thinking about programs that write essays too, just for balance
06:25:46 [majukati]
06:27:01 [tuukkah]
vegai, yeah
06:33:41 [benja_]
06:33:47 [majukati]
06:33:53 [tuukkah]
huomenta benja_
06:34:01 [benja_]
* benja_ woke up too late to be at uni at 10
06:34:06 [tuukkah]
did you get home well yesterday?
06:34:06 [benja_]
06:34:22 [majukati]
benja_: ok, not a problem. i need to walk there
06:34:23 [benja_]
tuukkah: yeah, I took the bus to the center and walked from there
06:34:30 [benja_]
majukati, ok
06:34:43 [benja_]
BUT I forgot I had my washing machine turn yesterday
06:34:53 [benja_]
and I really needed to wash. ARGH ARGH :-(
06:35:22 [tuukkah]
perhaps you wash now while eating breakfast and before coming to university ?-)
06:35:36 [majukati]
it's good that there aren't smellies yet ;)
06:36:01 [benja_]
I should probably go there at 10 and see whether the person takes their turn
06:36:15 [ibid]
it turns out i didn't lose the elections after all
06:36:21 [benja_]
hm? :)
06:36:23 [ibid]
i'll be the fifth(!) deputy member
06:36:41 [ibid]
(of the faculty council in the second staff group)
06:38:10 [tuukkah]
hmm, moin 1.3.4-1 has come to Sarge
06:38:50 [ibid]
* ibid upgrades flowerpot
06:39:21 [tuukkah]
* tuukkah wonders how it handles data migration
06:39:37 [majukati]
ibid: is the function of that machine of being a flowerpot?
06:39:54 [ibid]
06:42:01 [majukati]
benja_: so, when we see then?
06:43:04 [benja_]
depends on whether the machine is free or not :-)
06:43:52 [benja_]
if it's not, 10:40 perhaps
06:44:47 [tuukkah]
how come we always have these freedom issues
06:45:10 [majukati]
tuukkah: ?
06:45:52 [tuukkah]
now free software isn't enough, we need a free machine too ;-)
06:47:22 [tuukkah]
ibid, the upgrade breaks the wikis until they're manually migrated
06:48:36 [ibid]
tuukkah: can you migrate them?
06:48:41 [tuukkah]
06:51:02 [tuukkah]
an impressive version number: libtheora0 0.0.0.alpha3-1
06:59:05 [ibid]
Preparing to replace moin 1.2.4-1 (using .../archives/moin_1.3.4-1_all.deb) ...
06:59:05 [ibid]
Unpacking replacement moin ...
07:03:29 [benja_]
07:03:39 [benja_]
someone reserved the machines from 9-11
07:03:54 [tuukkah]
07:03:56 [benja_]
looks like they'll probably be finished in 30min, though
07:04:16 [ibid]
Setting up moin (1.3.4-1) ...
07:04:19 [ibid]
07:04:22 [benja_]
so I could probably use the machines for the last half of the hour
07:04:46 [benja_]
majukati: how about meeting at 12 for lunch?
07:04:55 [benja_]
and doing some work on IRC before that
07:05:27 [vegai]
have you ever seen this when trying rsync over ssh
07:05:32 [vegai]
[vegai@kontu haskell-skeleton]$
07:05:32 [vegai]
filter rules are too modern for remote rsync.
07:05:32 [vegai]
rsync error: syntax or usage error (code 1) at exclude.c(1119)
07:05:42 [majukati]
benja_: ok
07:07:02 [tuukkah]
vegai, perhaps you should set some simpler rules then :-)
07:07:32 [vegai]
of course, I wouldn't be wondering about this if it weren't for the fact that this used to work
07:07:46 [vegai]
I have no idea what filtering rules it's even talking about
07:07:50 [ibid]
tuukkah: upgrade complete
07:08:27 [vegai]
agh, server rsync: "rsync version 2.5.6cvs protocol version 26
07:08:28 [vegai]
07:09:00 [vegai]
debian :-(
07:09:40 [vegai]
is that really the version in debian stable?
07:09:43 [tuukkah]
ibid, noticed
07:10:03 [tuukkah]
rsync 2.5.5-0.6 stable
07:10:33 [vegai]
yep, that's the one
07:12:59 [tuukkah]
what's wrong with that ?-)
07:13:37 [vegai]
the "cvs" in the version line makes me fear
07:14:27 [vegai]
I think I'll just compile my own rsync until the admin updates
07:19:07 [vegai]
is the next major version of debian stable coming soon, btw?
07:19:46 [benja_]
07:19:57 [benja_]
07:20:13 [tuukkah]
vegai, it always is ;-)
07:20:22 [vegai]
07:21:33 [tuukkah]
and like this moin upgrade, it took them ages to prepare it, and I don
07:21:46 [tuukkah]
I don't see anything better than in upstream
07:22:14 [antont]
moin? that's not moinmoin i suppose?
07:22:23 [tuukkah]
that exactly
07:22:57 [benja_]
do you guys know what 'moin' / 'moin, moin' mean, actually? :)
07:23:00 [vegai]
ah, bugger.
07:23:19 [vegai]
for some reason, the "--cvs-exclude" command doesn't work for rsync over ssh anymore
07:23:40 [tuukkah]
benja_, yes. there's this nice bot on #moin that you need to say "Moin!" to :-)
07:23:47 [benja_]
07:24:16 [benja_]
it's northern German (and perhaps used in other languages/countries too, dunno 'bout that)
07:24:28 [tuukkah]
07:24:36 [benja_]
with an 'n'?
07:24:43 [benja_]
heh :)
07:24:44 [tuukkah]
well, without :-)
07:24:49 [benja_]
oh, ok :)
07:24:55 [majukati]
07:26:03 [antont]
is upstream some wiki too, then?
07:26:27 [tuukkah]
at least in debian, "upstream" means the original developers
07:27:41 [benja_]
somehow I'm wondering whether all the moin-saying in northern Germany is partly a game for tourists ;-)
07:27:58 [benja_]
as in, you can feel like you belong if you say 'moin' like everybody =)
07:28:08 [benja_]
but perhaps it's really just sort of local pride =)
07:29:21 [vegai]
yep, this is a feature of rsync-2.6.4
07:29:50 [antont]
ah yes i know that 'upstream' but read you somehow wrong that thouth it couldn't mean that
07:30:12 [antont]
"better than in upstream" has an extra 'than' i guess
07:31:54 [tuukkah]
antont, no, at least I mean that while it took Debian a long time to package the new moin, it doesn't have any enhancements if compared to the upstream release
07:39:03 [benja_]
ok, they did let me use the machine for the last part of their turn :)
07:39:10 [antont]
ah now i finally got you :)
07:39:25 [antont]
(busy here talking with students)
07:39:36 [antont]
yeah i've never looked at moin really
07:40:09 [tuukkah]
well, it's wiki and it's in python, so you'd suppose there couldn't be anything better %-)
07:40:10 [antont]
we've been using zwikis 'cause happened to have zope for some other reason, but are not too happy with them
07:40:33 [antont]
though zwiki has been developing .. quite a lot actually, it has seemed on the surface
07:40:58 [antont]
yeah .. but there are more php written things and they are used more
07:41:46 [tuukkah]
well, perhaps I'm a language elitist
07:42:10 [antont]
there seem to be (a lot of?) people moving from php to python, tho
07:42:33 [antont]
and (at least some) php ppl tend to agree that python is better for complex systems
07:42:43 [antont]
of course wikis can be simple too :o
07:43:02 [antont]
but dunno how it goes
07:44:37 [antont]
of to eat now, bbl.
08:13:15 [benja_]
laundry's in the dryer =)
08:13:22 [majukati]
08:13:38 [tuukkah]
benja_, va du e dukti :-)
08:14:40 [benja_]
du va så dukti du er ser kul
08:15:04 [benja_]
whatever that means
08:15:06 [majukati]
08:15:14 [benja_]
(what you said -- what I said probably means nothing =))
08:15:16 [benja_]
majukati, ?
08:15:37 [majukati]
benja_: I don't want to work with awt anymore
08:15:47 [tuukkah]
benja_, well, drop the last three words and it makes sense :-)
08:16:17 [benja_]
majukati, what's the problem?
08:16:37 [benja_]
tuukkah: so what did we say?
08:16:53 [tuukkah]
the same thing, basicly, just formulated differently =)
08:17:10 [benja_]
tuukkah: can you translate?
08:17:21 [tuukkah]
but I don't think I know how to say it in any other language
08:17:27 [benja_]
blah! :)
08:18:58 [majukati]
benja_: i use awt.image.MemoryImageSource but it's not working at all
08:19:31 [benja_]
ah, images...
08:19:55 [ibid]
actually, stable sarge is now closer than ever
08:20:00 [ibid]
freeze is imminent
08:20:04 [tuukkah]
benja_, literally, I said "what you are hard-working" and you replied "you were so hard-working, you"
08:20:09 [ibid]
if one believes the rms, naturally
08:20:49 [ibid]
tuukkah: in which language? it looks like swedish but isn't...
08:21:20 [tuukkah]
spoken swedish, not written
08:21:33 [ibid]
08:21:38 [benja_]
08:22:06 [tuukkah]
benja_, the cabal ;-)
08:22:09 [ibid]
release managers
08:22:15 [tuukkah]
08:22:34 [ibid]
08:22:42 [benja_]
and I was wondering what Stallman had to do with Sarge releases :)
08:22:46 [ibid]
08:22:46 [tuukkah]
nothing :-)
08:22:53 [vegai]
speaking of the cabal, I started a haskell skeleton for Cabalized programs:
08:23:23 [vegai]
"mule's bridge" of the millennium, eh?
08:23:28 [ibid]
i didn't know he uses the definitive article before his initials
08:24:06 [ibid]
i see my one suggestion to cabal actually was used :)
08:24:20 [vegai]
08:24:36 [benja_]
is anybody opposed to me making the iterators used by swamp into javolution Realtime objects?
08:25:03 [benja_]
i.e., if you use them inside a Javolution PoolContext, they'll be re-used when the pool context is exited
08:25:21 [benja_]
08:25:30 [ibid]
i suggested the use of the dctrl format :)
08:25:59 [tuukkah]
benja_, I think that's good. I suppose it'll make the code a lot faster? you did profile, didn't you?
08:26:41 [benja_]
tuukkah: yes, it looks like currently most objects created are swamp iterators
08:26:53 [majukati]
benja_: it's fine
08:26:53 [vegai]
ah, right
08:26:54 [benja_]
it *is* possible to work around that in a couple of ways, but this seems more elegant
08:27:39 [tuukkah]
08:29:39 [tuukkah]
vegai, something like that can be useful in cabal, yes
08:30:16 [rubberpaw]
rubberpaw has quit
08:30:30 [vegai]
I'll just need to come up with the test-runners and I can actually start doing something
08:31:55 [tuukkah]
wouldn't it be funny to try to beat javolution perfomance with Haskell ?-)
08:32:14 [vegai]
would that be a challenge? :P
08:32:17 [vegai]
what's javolution?
08:32:41 [tuukkah]
was it java framework for realtime apps
08:33:49 [antont]
how is haskell performance, then? i saw something somewhere that scheme people or someone was concerned about it
08:34:03 [vegai]
it's getting better
08:34:19 [tuukkah]
and probably depends on the application
08:34:24 [vegai]
ghc moves in a steady pace, afaik
08:34:55 [vegai]
on the usenet, somebody tested the naive fibonacci function with both ghc-6.4 and ocaml, and it was faster in ghc
08:34:56 [tuukkah]
in some microbenchmark it did really well, didn't it?
08:35:17 [vegai]
yes, some
08:35:21 [vegai]
and awful in most
08:35:39 [tuukkah]
hmm, I didn't read them as awful :-)
08:36:04 [vegai]
"worse than Java" is equal to "awful" in my vocabulary :P
08:37:04 [benja_]
it *is* possible to work around that in a couple of ways, but this seems more elegant
08:37:07 [benja_]
08:37:12 [benja_]
ww (wrong window)
08:39:15 [tuukkah]
vegai, in some microbenchmarks, java does better than gcc. so...?
08:39:43 [vegai]
riight :)
08:40:13 [vegai]
08:40:55 [tuukkah]
what about it?
08:41:17 [vegai]
I guess you meant Sun's java, right?
08:41:44 [tuukkah]
probably, that's what people usually mean with java
08:42:22 [vegai]
antont: anyway, the implementations obviously all have their strong and weak sides
08:42:53 [vegai]
Gambit, for instance, seems to be a very snappy scheme implementation
08:46:30 [vegai]
but I believe that in a larger program (especially when you're past prototyping), features like strong static typing can be very helpful
08:48:21 [vegai]
(too bad this isn't a religion)
08:55:13 [tuukkah]
you would have to think a bit before porting org.fenfire.swamp into Haskell...
08:55:52 [antont]
yeah, i've been reading a bit of some of the proposed typing systems for Python
08:56:08 [antont]
like zope3 has this Interface system
08:56:18 [vegai]
08:56:29 [vegai]
python's getting weirder and weirder
08:56:39 [antont]
well zope3 is not python
08:56:53 [antont]
and the typing things that people do with/for python are not python either
08:56:56 [vegai]
well no, but
08:57:20 [antont]
i don't know if python will ever adopt any of those mechanisms
08:57:26 [antont]
but several have been proposed
08:59:11 [antont]
the "Python Enterprise Application Kit" has one, there when you get an object, and you need it to conform to some interface / be of some type, you call somehow call adapt(object), and it rises an exception if there's a problem (wrong type / does not implement the interface)
08:59:47 [antont]
ah that seems to be the PyProtocols thing,
09:00:16 [antont]
the zope3 thing was more like in java
09:02:09 [antont]
or well perhaps that is similar too, 'cause it says it can use the zope (and twisted) interfaces
09:04:39 [antont]
vegai: how do you think it has gotten weird(er), btw?-)
09:04:57 [vegai]
well, they were just talking about removing map/filter/lambda
09:05:12 [antont]
for python 3000, yes
09:05:25 [antont]
i guess map&filter because of list comprehensions
09:06:04 [antont]
lambda i don't know why really, nor why should .. i guess it has it's uses
09:06:20 [vegai]
not as many if you remove map and filter :)
09:06:36 [antont]
well the list comprehensions are there
09:06:54 [antont]
i guess the only reason to remove map&filter is if list comprehensions are a better syntax for them
09:07:06 [vegai]
that's a big if
09:07:26 [antont]
yes, and there's still a lot of time for everyone to think & argue about it
09:07:43 [tuukkah]
list comprehension syntax is [variable for variable in ... if variable ...] :-)
09:08:02 [benja_]
tuukkah: if <any condition> actually
09:08:04 [antont]
i've found myself always using list comprehensions, 'cause i never remember the order or arguments for map directly, and even less for filter
09:08:22 [tuukkah]
benja_, is that of any use?
09:09:00 [benja_]
tuukkah, well, for one thing, the expression can involve the variable at an arbitrary point
09:09:04 [antont]
[show(vob) for vob in vobs if vob.visible]
09:09:06 [tuukkah]
oh, I know: [variable for variable in ... if random() % 2 == 0]
09:09:15 [benja_]
expressions not involving the variable are probably useful in some very edge cases :)
09:09:22 [benja_]
tuukkah: hey, yeah, that's nice :)
09:09:30 [antont]
yeah tuukkah's was a good one :)
09:10:50 [vegai]
interesting that % is defined for floats
09:11:00 [vegai]
or is it?
09:11:04 [antont]
09:11:11 [vegai]
is it interesting, I meant :)
09:11:15 [antont]
>>> 1.1 % 1
09:11:15 [antont]
09:12:04 [vegai]
there we have a nastiness in python, kinda
09:12:13 [vegai]
tuukkah's expression would give empty list every time
09:12:41 [vegai]
I guess some language might give a type error
09:12:42 [tuukkah]
oops, [variable for variable in ... if random() < 0.5]
09:13:00 [antont]
this i found nasty too:
09:13:09 [antont]
>>> 0.01 == float('nan')
09:13:09 [antont]
09:13:17 [antont]
>>> float('nan') is float('nan')
09:13:17 [antont]
09:14:21 [vegai]
huh... what's nan as float?
09:14:23 [antont]
.. so i ended up doing > 0 to check it is not 'nan'
09:14:45 [antont]
it's a float that's not a number
09:15:04 [antont]
at least this 3d engine i'm using (soya) can return it sometimes when getting the angle between two vectors etc
09:15:59 [tuukkah]
if one of the vectors is 0
09:17:31 [antont]
Vector is a class in Soya
09:17:52 [tuukkah]
a vector of length 0, that is
09:18:07 [antont]
ah right
09:20:08 [antont]
hm i'm still not able to reproduce that
09:21:01 [antont]
anyhow, one stand people take regarding the weaknesses & flaws in Python is that it's better to try to fix them, instead of making a new language
09:21:19 [antont]
but surely at least in research there's always room for more :)
09:24:32 [vegai]
I wonder what Jonne's mystery language
09:24:34 [vegai]
09:24:49 [tuukkah]
09:25:04 [antont] seems to be about that float('nan') comparison
09:33:02 [benja_]
argl! 12:32
09:33:21 [benja_]
majukati: I've taken the laundry out of the dryer, will take the next bus to uni
09:33:31 [benja_]
should be there at ca 13:05
09:33:38 [benja_]
where should I look for you?
09:39:04 [tuukkah]
ibid, the migration should be complete now
09:39:28 [majukati]
benja_: we may start at piato
09:40:39 [benja_]
09:42:02 [ibid]
right :)
09:45:02 [tuukkah]
right to what?
09:49:30 [ibid]
12:38 <tuukkah> ibid, the migration should be complete now
09:50:54 [rubberpaw]
rubberpaw has joined #fenfire
10:20:33 [tuukkah]
could it have been wrong, then?
10:22:50 [ibid]
10:29:05 [rubberpaw]
rubberpaw has quit
11:29:58 [ffdarcsbot]
libvob: Benja Fallenstein <>, make mouse events in ViewportLob work
11:51:02 [rubberpaw]
rubberpaw has joined #fenfire
12:41:55 [ffdarcsbot]
fenfire: Benja Fallenstein <>, optimizations (create less objects)
12:54:32 [ffdarcsbot]
fenfire: Benja Fallenstein <>, ARGH
12:59:03 [rubberpaw]
rubberpaw has quit
13:13:21 [ffdarcsbot]
fenfire: Benja Fallenstein <>, start a non-patent-related public version of Fenfire ('FenHaystack' :-))
14:02:26 [rubberpaw]
rubberpaw has joined #fenfire
14:22:59 [ibid]
majukati: this may be relevant to your thesis ->
14:39:28 [rubberpaw]
rubberpaw has quit
16:02:33 [rubberpaw]
rubberpaw has joined #fenfire
16:49:10 [ffdarcsbot]
libvob: Benja Fallenstein <>, hack ClipLob to not change the matching parent, but to place nested clipping cses with the same matching parent into a vobmatcher hierarchy
16:49:20 [ffdarcsbot]
libvob: Tuukka Hastrup <>, fix awt line drawing by substracting one pixel
16:59:29 [ffdarcsbot]
fenfire: Benja Fallenstein <>, add a list of thumbnails of other spatial views for thhis node to the 'public Fenfire'
17:05:39 [rubberpaw]
rubberpaw has quit
17:18:44 [ffdarcsbot]
fenfire: Benja Fallenstein <>, fix bug that prevented nodes on the canvas from being shown in a canvas buoy
17:43:31 [ibid]
ibid has quit
18:02:43 [rubberpaw]
rubberpaw has joined #fenfire
18:06:08 [ibid]
ibid has joined #fenfire
18:19:17 [ibid]
wow, shfs is so cool :)
18:28:00 [rubberpaw]
rubberpaw has quit
19:37:32 [majukati]
plaah, installing a debian is not fun
19:40:39 [majukati]
why to have four broken cds?-/
19:51:29 [ibid]
don't get broken cds, get good cds :)
19:53:49 [majukati]
no, this just stops where i press 'install base'
19:54:01 [majukati]
when eventually
19:54:41 [ibid]
well, i can't help you there
19:54:46 [ibid]
it works for me :)
19:55:51 [majukati]
install the base system and it starts the installing program again
19:56:14 [majukati]
i have no idea, i see no errors
19:58:48 [ibid]
is this on a laptop?
19:58:58 [majukati]
no, an old machine
19:59:39 [ibid]
which version of debian?
20:00:00 [majukati]
20:00:06 [ibid]
20:00:17 [ibid]
look for any errors in the second virtual console
20:00:25 [majukati]
there are none
20:00:40 [ibid]
i'd have to see it to help further :(
20:02:49 [majukati]
var/log/messages says that /sbin/udbootstrap exited
20:02:54 [rubberpaw]
rubberpaw has joined #fenfire
20:03:08 [majukati]
but it has the command line that i can run myself :)
20:04:05 [majukati]
oh, in virtual console 4 i have the message: Segmentation fault
20:04:16 [majukati]
20:04:28 [ibid]
how much memory?
20:04:36 [majukati]
20:05:01 [ibid]
not very old, then? :)
20:10:45 [majukati]
i used to had older stable but lost those cds in a move
20:21:06 [majukati]
or actually i just didn't find those. will use *floppies* tomorrow!
20:24:18 [ibid]
you only need two if you have net access from the comp