EliasAmaral has quit [Read error: 104 (Connection reset by peer)]
thorat has quit [Remote closed the connection]
piggybox has joined #ocaml
just1n has joined #ocaml
scruffie has joined #ocaml
scruffie has quit []
zarvok has quit ["BitchX-1.1-final -- just do it."]
seafood has quit [Read error: 104 (Connection reset by peer)]
seafood has joined #ocaml
seafood has quit [Read error: 104 (Connection reset by peer)]
danly has quit [Remote closed the connection]
<flux>
I suppose you could hide the lazy list to a function that would simply take a stream and return two streams, which are copies of the original
<flux>
it's a shame there aren't functional streams in ocaml; they would look very much like haskell lazy lists
<flux>
("in ocaml" in the same sense the destructive streams are in.. as an extension.)
<flux>
there used to be such a beast with the distribution, but I never managed to make it work - and I think it was to be removed from future releases
beginner has quit [Read error: 110 (Connection timed out)]
<lde>
For a while i thought you were asking for something else and I thought this was #emacs ;->
<lde>
s/for/about/
malc_ has joined #ocaml
smimou has joined #ocaml
Mr_Awesome has quit ["time to impregnate a moth"]
ygrek has quit [Remote closed the connection]
ygrek has joined #ocaml
pango has quit [Remote closed the connection]
pango has joined #ocaml
_JusSx_ has joined #ocaml
tty56 has joined #ocaml
_blackdog has joined #ocaml
_blackdog has left #ocaml []
piggybox3 has joined #ocaml
_JusSx__ has joined #ocaml
tty56_ has quit [Read error: 110 (Connection timed out)]
piggybox has quit [Read error: 110 (Connection timed out)]
_JusSx_ has quit [Read error: 110 (Connection timed out)]
Submarine has quit [Read error: 113 (No route to host)]
slipstream has joined #ocaml
vorago has quit [Read error: 113 (No route to host)]
slipstream-- has quit [Read error: 110 (Connection timed out)]
visage has joined #ocaml
Lena has joined #ocaml
screwt8 has quit [Remote closed the connection]
<ygrek>
when using lablgtk2 with gtk 2.8 runtime I get errors (function blablabla not found in dll) - does that mean that I should fall back to the earlier version - 2.4?
<ygrek>
hm. no win32 packages on ftp.gtk.org for 2.4..
yodafinger_ has joined #ocaml
yodafinger has quit [Read error: 104 (Connection reset by peer)]
yodafinger_ is now known as yodafinger
cpfr has joined #ocaml
malc_ has quit ["leaving"]
jlouis_ has joined #ocaml
jlouis has quit [Read error: 110 (Connection timed out)]
slipstream-- has joined #ocaml
slipstream has quit [Read error: 110 (Connection timed out)]
visage has quit []
pango has quit [Remote closed the connection]
pango has joined #ocaml
<rwmjones>
ygrek, what's the error?
<rwmjones>
also are you linking to gtkInit.cmo or gtkInit.cmx?
<ygrek>
linking is ok
<rwmjones>
failing to do that usually causes the program to spew error messages shortly before crashing
<ygrek>
that was runtime error
<ygrek>
cmx
<ygrek>
but anyway - I switched to 2.6.9 and it is ok now
<rwmjones>
I'm using 2.10 (on Linux) and it works fine
<ygrek>
hm. g_intern_static_string it was
<ygrek>
in glib-2.0-0.dll
vorago has joined #ocaml
slipstream-- has quit [Read error: 110 (Connection timed out)]
rwmjones has quit ["Closed connection"]
visage has joined #ocaml
visage has quit [Client Quit]
_blackdog has joined #ocaml
_blackdog has left #ocaml []
<pango>
how would you progress in several maps at once? Map.{iter,map,fold} don't cut it, and I can't think of a way to get (key, value) tuples as a stream, or something like that...
<pango>
Map and Set module interfaces are kind of minimalistic... :/
<Smerdyakov>
Usually I do the single-map version, looking up matching values in another map manually inside the fold function.
<pango>
it's rather a parallel progress; in this case in keys order. But keys don't match exactly either
<pango>
each map stores different kind of events, and I want replay all events chronogically
<pango>
maybe that's not the right datastructure for the job...
<bluestorm_>
hm
<bluestorm_>
you could add them all in a proprity queue, then fold it
<bluestorm_>
rather easy to code, and not so innefficient
<bluestorm_>
s/proprity/priority/
<pango>
yes, using some variant type to tell apart the different kinds of events...
<pango>
problem is, some lists are in the 1e+7 events range
<pango>
I hope it'll be better than my first naive approach... 16GB processes aren't pretty ;)
noteventime has joined #ocaml
zarvok has joined #ocaml
Lena has quit ["Leaving."]
<flux>
pango, you can do that with set (it has partition or something, which is efficient), but not with map.
<flux>
pango, I've used an approach of having both a set and a map on data
<flux>
which I suppose is elegant.. but it could be more efficient
<flux>
in constant-factor-way
<pango>
not sure how Set.partition would help...
<bluestorm_>
hm
<bluestorm_>
i'm not sure a variant type is needed
<bluestorm_>
i mean
<bluestorm_>
you want to "replay" all events
<bluestorm_>
so you're using something a common interface for all events, aren't you ?
<bluestorm_>
maybe you could use the map -> pqueue transformation to extract the information you need for the replay at the same time
<bluestorm_>
and then you wouldn't have to carry details on the different kinds any more
<pango>
I'm trying to extract statistics from the whole data, but I'm not sure what statistics will be relevant in the end... Things are still fuzzy, so while I think "extracting/condensing" data as I read inputs will indeed be necessary to make the amount of data manageable, it gets in the way of flexibility and experimenting
<tsuyoshi>
if you have a one-to-one relationship between the values, why not just have pointers between them
<tsuyoshi>
e.g.
<tsuyoshi>
type foo = { data: string; bar: bar } and bar = { data: string; foo: foo }
<Smerdyakov>
You have to use lame recursive definitions to build a value of either type.
<tsuyoshi>
then just iterate over one of the maps
<pango>
I was thinking of using maps/sets/pqueues/whatever to store different condensed data over periods of time... But then comes the problem of crossing them
<tsuyoshi>
what kind of data is it?
<pango>
even plain array could do, if I knew keys range ahead of time
Jessehk has joined #ocaml
jaapweel has joined #ocaml
<Jessehk>
Hi all. :)
<Jessehk>
I'm trying to create a package for Arch linux, and I was wondering what else besides placing a package (in this case, lablgtk) in /usr/lib/ocal/site-lib do I have to do before ocamlfind recognises it?
<Jessehk>
do I have to "ocamlfind install" ?
<flux>
pango, with partition you can iterate n different sets efficiently?
<flux>
well, perhaps not optimally, though
<flux>
the complexity will stil be larger than O(n)..