--- Log opened Mon Dec 21 00:00:21 2009
11:56 < xkb> hi
11:57 < xkb> I'm having problems installing happs-tutorial. cabal install happs-tutorial fails on a type match error
11:57 < xkb> Couldn't match expected type `ServerPartT IO a'
11:57 < xkb>            against inferred type `[Char]'
11:57 < xkb>     In the second argument of `smartserver', namely `"happs-tutorial"'
11:57 < xkb> Any ideas on how to fix this?
11:57 < xkb> I tried it on both my mac and my windows xp machine
12:03 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5476#a5476
12:03 < HugoDaniel> AAAAAAAAARRRRRRRRRRRGGGGGGGGGGGHHHHHHHHHHHH
12:03 < HugoDaniel> HELPPPPPPPP!!!!!!!!!!!!!
12:08 < HugoDaniel> (the drama)
15:03 < HugoDaniel> i keep getting segfaults with happstack :(
15:03 < burp> segfault? o0
15:03 < HugoDaniel> ill paste the whole code once im finished... its reallllllyyyyyy messy
15:03 < HugoDaniel> yes :(
15:03 < burp> how are you able to segfault haskell? :D
15:03 < HugoDaniel> the server is runing in ghci
15:04 < HugoDaniel> and all the sudden ghci exists and a nice "Segmentation fault" message is printed on the terminal
15:04 < HugoDaniel> burp: my code is messy :/
15:04 < HugoDaniel> i guess
15:04 < HugoDaniel> and i use the function arguments stack at the maximum most of the time
15:04 < HugoDaniel> or whatever that was outputing the other error that i pasted
15:05 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5476#a5476  <- this is at 20 :P
15:06 < burp> wtf :D
15:08 < sm> evening all
15:08 < HugoDaniel> hi s&m
15:08 < sm> I'm getting "HTTP request failed with: stack overflow" on the console.. does this sound like a happstack error message ?
15:09 < HugoDaniel> :D
15:09 < HugoDaniel> dont be sad... im getting segmentation faults
15:09 < HugoDaniel> :D
15:09 < sm> dear me
15:09 < sm> that's not so good
15:10 < sm> well I wonder how to track down this kind of thing. Lots of trace statements I suppose
15:10 < sm> I am wondering if it's related to the size of the request's output, about 400k
15:11 < HugoDaniel> :/
15:11 < HugoDaniel> quite probably
15:11 < HugoDaniel> if i have the time, after this project, im going to try some stress tests on happstack
15:12 < HugoDaniel> because i keep feeling it is quite unstable
15:12 < sm> that would be great
15:12 < sm> I heard 0.4 fixes at least one such issue
15:15 < sm> after adding the trace statements it works. And still when I remove them again
15:17 < sm> side question, how do you get happstack to not write "request" on the console ?
15:20 < sm> HugoDaniel: what's your project ?
15:33 < vowyer> sm: using the logging facilities you can divert all messages to log files
15:33 < vowyer> sm: the guestbook example included does that
15:36 < sm> vowyer: thanks
16:14 < stepcut> sm: the stack overflow is probably due to something in your code which is lazily folding over a large number of elements creating lots of thunks which later get forced
16:15 < stepcut> HugoDaniel: did you add -fcontext-stack=N where N > 20 ?
16:18 < stepcut> HugoDaniel: happstack has been quite stable in my experience. Does your code depend on other libraries or using the FFI ?
16:32 < HugoDaniel> sorry, im busy, ill answer as soon as i can
16:32 < HugoDaniel> but i didn't add it
16:32 < HugoDaniel> i just changed the number of args in my functions
16:33 < HugoDaniel> brb
18:08 < HugoDaniel> hi
18:08 < HugoDaniel> i have a server that listens on a socket
18:08 < HugoDaniel> how do i make it a daemon ?
18:08 < HugoDaniel> so that i can logout of my shell and it stills listen on ?
18:17 < stepcut> HugoDaniel: on FreeBSD?/
18:18 < stepcut> HugoDaniel: i would make an init script so that the server restarts automatically when the server does?
18:20 < HugoDaniel> no
18:22 < sm> stepcut: thanks
18:26 < MAN_> HugoDaniel: you could use cron.
18:27 < MAN_> btw, people, is the happstack-helpers package part of the suite? There are some funny notes of in hackage.
18:28 < MAN_> and, at least one function (simpleserver (')) differs from documentation .
18:29 < MAN_> brb
18:32 < stepcut> HugoDaniel: disown ?
18:33 < stepcut> HugoDaniel: what distro?
18:33 < stepcut> MAN_: happstack-helpers is a 3rd party thing
18:43 < ac> just downloaded the new release of happstack with cabal and I'm trying out the "new project" command
18:44 < ac> that works fine, but when I try to run bin/run.sh I get: trhsx: Error at SrcLoc {srcFilename = "src/GuestBook/View.hs", srcLine = 34, srcColumn = 1}:
18:44 < ac> Illegal instance declaration
18:47 < MAN_> ac: do you have the trhsx program (it's an extra library)
18:48 < stepcut> MAN_: that is an error message from trhsx..
18:48 < MAN_> stepcut: Oh...
18:48 < ac> yeah, I definitely have trhsx. I'm guessing I just need to upgrade one of the happstack packages
18:48 < ac> I had the older version and I just ran "cabal install happstack"
18:49 < MAN_> I solve that same problem in my box, yesterday, installing libghc6-harp-dev
18:49 < MAN_> *solved
18:49 < stepcut> ac: do you perhaps have an older version of trhsx installed somewhere ?
18:50 < stepcut> ac: that error message is due to use a version of hsx that is too old
18:51 < ac> stepcut: I'm not sure. `ghc-pkg list trhsx` shows nothing, so I'm figuring it's part of one of my happstack packages
18:51 < stepcut> ac: trhsx gets install as part of hsx
18:51 < stepcut> you should be using hsx-0.5.5
18:52 < ac> I have hsx-0.5.6
18:52 < stepcut> what does, which trhsx, say?
18:53 < stepcut> is hsx-0.5.6 installed in the global package config or your user package config?
18:55 < ac> stepcut: user
18:55 < bnijk_> i still think my haskell logo was the best one
18:55 < ac> stepcut: most everything is in user
18:55 < bnijk_> http://community.haskell.org/~hexpuem/logo_contest/haskell_5.png
18:55 < ac> bnijk_: really? hah
18:56 < bnijk_> no art in the other ones ;)
18:56 < stepcut> ac: I am wondering if hsx-0.5.6 installed trhsx in ~/.cabal/bin, but there is also an older version of trhsx in /usr/local/bin or something?
18:56 < stepcut> bnijk_: I liked that one too, maybe we can adapt it to be a happstack logo somehow?
18:56 < bnijk_> by all means
18:56 < bnijk_> are you in charge of that?
18:57 < stepcut> bnijk_: sure :)
18:57 < bnijk_> lol
18:57 < bnijk_> so long as i get credit ;)
18:57 < ac> stepcut: maybe I should downgrade to 0.5.5?
18:57 < stepcut> bnijk_: do you wish to make any changes to make it more happstacky?
18:57 < bnijk_> i'll think about it
18:57 < bnijk_> i don't know if i have it in svg
18:57 < stepcut> ac: no, 0.5.6 is better -- I am just behind
18:58 < stepcut> ac: is ~/.cabal/bin in your PATH?
18:59 < ac> stepcut: yeah, and it's using the right one
18:59 < stepcut> ac: what version of haskell-src-exts?
18:59 < ac> stepcut: seems to be two: haskell-src-exts-0.4.8, haskell-src-exts-1.0.0
18:59 < stepcut> ac: http://groups.google.com/group/HAppS/browse_thread/thread/850ec6efb68cd61b
19:01 < stepcut> ac: I would update haskell-src-exts to the latest version that hsx works with (check hsx.cabal) and then rebuild hsx against it.
19:01 < stepcut> gotta run, bbl
19:02 < ac> ahah
19:02 < ac> stepcut: thanks for the help :)
19:02 < ac> someone should update the version dependency
19:03 < stepcut> yeah, I'll mention that on the hsp list
19:51 < ac> I had to explicitly install haskell-src-exts-1.0.1 because there's a much newer version, and hsx depends on <1.1
19:54 < Raevel> seems to me it would be a good idea to always explicitly set the version numbers instead of using ranges
20:22 < stepcut> ac: yes
20:23 < stepcut> ac: but if you can fix hsx to work with haskell-src-exts >1.1 we will love you forever :)
20:26 < stepcut> our bug mostly obviously appears at the simplifier stage.. but I need to look at the desugared code and see if the bug exists there or not
20:26 < stepcut> oops
20:26 < stepcut> wrong window :)
20:27 < stepcut> Raevel: it is good to set the lower bound explicitly, but setting an exact dependency would mean that people would not be able to use newer, better versions unless we up'd the dependency
20:33 < Raevel> sure, except in my experience new versions often break things as well :-/
--- Log closed Tue Dec 22 00:00:24 2009