--- Log opened Sun Dec 20 00:00:21 2009
05:04 < koeien> just got an e-mail from Niklas; HJScript is updated. woo! :)
05:09 < koeien> i'll purge the .ghc and .cabal of the chroots again
05:10 < koeien> h_buildbot: build
05:10 < h_buildbot> Build for all started. If one was running, no new one is started.
05:13 < h_buildbot> Build for ghc-6.8.3 failed. Check http://buildbot.happstack.com/ for details.
05:18 < h_buildbot> Build for ghc-6.10.2 OK. Test suite build failed. Check http://buildbot.happstack.com/ for details.
05:18 < h_buildbot> Build for ghc-6.12.1 failed. Check http://buildbot.happstack.com/ for details.
05:18 < koeien> ...
05:18 < koeien> ah
05:19 < koeien> h_buildbot: build ghc-6.12.1
05:19 < h_buildbot> Build for ghc-6.12.1 started. If one was running, no new one is started.
05:19 < koeien> for GHC 6.8.3 there seems to be a dependency error somewhere
05:24 < h_buildbot> Build for ghc-6.12.1 failed. Check http://buildbot.happstack.com/ for details.
05:37 < koeien> the test build also fails on my desktop w/GHC-6.10.4. Haven't pinpointed the issue yet
08:49 < stepcut> odd. the test build works on my desktop with 6.10.4 :-/
08:49 < koeien> platform?
08:55 < stepcut> ubuntu jaunty
08:55 < koeien> that is quite strange. I get the same error as on the buildbot here. Debian 5.0
08:56 < stepcut> very little changed between when the tests worked and when they broke -- I just changed the dependencies in .cabal
08:56 < koeien> yes, i saw that
08:57 < stepcut> does ghc-pkg describe happstack-state-0.4.1 show the tests as exposed ?
08:58 < stepcut> of course, there is also the weirdness that everything is fine on 6.12...
08:58 < koeien> no. Neither exposed, nor hidden
08:59 < stepcut> so it is as if happstack-state-0.4.1 was built with out the -ftests flag?
08:59 < koeien> yeah, i guess so
08:59 < koeien> let me try that more explicitly
09:00 < stepcut> actually, i think I see the problem
09:00 < koeien> it being?
09:00 < stepcut> nevermind, my eyes were playing tricks on me
09:01 < stepcut> in the build log it looked like happstack-state-tests got built...
09:02 < stepcut> gotta shower, bbiab
09:10 < koeien> hmm. i don't understand this. for some reason it works now on my desktop
09:10 < koeien> maybe i do. let me think
09:12 < koeien> oh. unregister-all.sh was removed from the repos...
09:12 < koeien> this could very well be a problem
09:16 < koeien> h_buildbot: build ghc-6.10.2
09:16 < h_buildbot> Build for ghc-6.10.2 started. If one was running, no new one is started.
09:18 < h_buildbot> Build for ghc-6.10.2 OK. Test suite build failed. Check http://buildbot.happstack.com/ for details.
09:22 < koeien> are we sure the ordering in bin/build-install-test-all.sh is correct?
09:23 < stepcut> seems fine to me..
09:24 < koeien> it has always worked[tm]
09:24 < stepcut> odd, I wonder how unregister-all got dropped
09:24 < koeien> AHA
09:25 < koeien> for some reason happstack-state is getting downloaded
09:25 < koeien> from Hackage
09:25 < stepcut> !
09:25 < koeien> let me verify
09:26 < koeien> cabal doesn't apply flags recursively to dependencies apparently
09:26 < koeien> (I don't know if I should expect this to happen.)
09:30 < stepcut> I am not sure that unregister-all.sh was ever *in* the repository -- what does your version look like?
09:30 < koeien> apparently I called it from the script
09:30 < stepcut> I wonder unregister-all.sh is something you wrote though ?
09:30 < koeien> yeah, could very well be
09:31 < koeien> but happstack-state is in my .cabal/packages folder now. something depends on it
09:31 < stepcut> I happen to have a bin/unregister-all.sh as well... but I think it is something that I wrote...
09:32 < stepcut> I should commit my version though.. it is obviously useful ;)
09:32 < koeien> yes, please do :)
09:33 < koeien> h_buildbot: pause
09:33 < h_buildbot> Already paused, or currently building. Try 'status'
09:33 < koeien> h_buildbot: status
09:33 < h_buildbot> Building ghc-6.8.3
09:33 < koeien> is that your patch?
09:33 < stepcut> ?
09:33 < koeien> hmm.
09:34 < koeien> never mind
09:35 < koeien> h_buildbot: status
09:35 < h_buildbot> Idle.
09:35 < koeien> h_buildbot: pause
09:35 < h_buildbot> Paused
09:35 < stepcut> ok, I pushed the patch
09:38 < koeien> darcs hangs for me
09:38 < koeien> while pulling
09:40 < stepcut> patch-tag is down :-/
09:40 < stepcut> will probably be up in a minute
09:40 < koeien> did you break it by pushing? ;)
09:40 < stepcut> no :)
09:40 < koeien> i like hg better than darcs.
09:41 < stepcut> I heard hg borrowed all the features I like of darcs :)
09:41 < stepcut> like, darcs send
09:41 < koeien> yes. darcs send rules
09:42 < koeien> but it's very useful to be able to say "I just tested repos version 243980234"
09:42 < koeien> afaik that's not possible w/darcs
09:42 < koeien> but w/e. darcs is also a good tool
10:01 < koeien> once 'happstack' is built, it wants happstack-state-0.4.1 from hackage for some reason
10:02 < stepcut> that is ... odd
10:02 < koeien> very odd, yes
10:02 < koeien> i do not have an explanation for this
10:02 < stepcut> but shouldn't the version for hackage work anyway?
10:02 < koeien> it works, but it doesn't get -ftests
10:02 < koeien> thereby making sure that the test build fails
10:07 < stepcut> right, I just realized that
10:10 < stepcut> perhaps it is just a bug in cabal? I am using Cabal 1.6.0.3 (with ghc 6.10.4)
10:10 < koeien> yeah i am beginning to think that, too
10:10 < koeien> on my desktop it works now. also 10.4
10:10 < stepcut> gotta, run, be back in a few hours
10:10 < koeien> ok ttyl
10:14 < HugoDaniel> hi
10:14 < HugoDaniel> back to happstack coding
10:15 < HugoDaniel> im going to update it to happs 4
10:16 < koeien> you mean 0.4?
10:16 < koeien> should be easy if you come from 0.3
10:19 < HugoDaniel> yes, 0.4
10:19 < HugoDaniel> sorry
10:19 < HugoDaniel> i dont tend to read the 0. :P
11:03 < vowyer> cabal upgrade happstack fails for me with multiple dependency problems related to base package. cabal install seems to not have that problem
11:03 < aavogt> cabal upgrade is broken independent of happstack
11:03 < vowyer> ah, ok
11:16 < jbe> Is happs-tutorial in a working state?  When I try to "cabal install happs-tutorial", I get an error building HStringTemplate-0.5 (Could not find module `Control.Monad.Writer': it is a member of the hidden package `mtl-1.1.0.2')
11:53 < HugoDaniel> why is the hjavascript needed for the happstack 0.4 ?
11:53 < koeien> hsx requires it afaicr
11:54 < HugoDaniel> hmm nice
11:55 < HugoDaniel> i really would like to see something like "server-side" javascript
11:55 < koeien> i really would like to see something like "client-side" haskell
11:55 < HugoDaniel> the future is probably going to pass through that... since javascript is starting to gain a huge momentum
11:55 < HugoDaniel> koeien, i also think a lot about that
11:55 < koeien> i don't think javascript is a particularly good or useful language
11:56 < HugoDaniel> its nice for simple scripts, i really enjoy the way it blends in flash
11:56 < HugoDaniel> adobe calls it "actionscript" but its the same
11:56 < HugoDaniel> anyway, about the client side haskell
11:57 < HugoDaniel> that would envolve gettin ghc (or a subset of it...) in a activeX controller
11:57 < HugoDaniel> an extension for firefox and another for opera
12:20 < aavogt> HugoDaniel: you can compile haskell to javascript
12:29 < HugoDaniel> and is it usable on the web ?
12:29 < HugoDaniel> wont it be like 5Mb of text in a .js file ? :/
12:33 < burp> happstack.com down?
12:33 < koeien> apparently
12:33 < koeien> patch-tag is down, too
12:37 < bnijk_> javascript is a joke
12:37 < HugoDaniel> why bnijk_ ?
12:38 < bnijk_> it has about three useful functions and then a thousand pain in the ass functions
12:39 < bnijk_> some of it has to do with how permissive the implementation is
12:50 < HugoDaniel> if i have a serverPartT like dir "test" (return . toResponse $ "TEST") , and go to http://localhost:8080/test or test/ or test/sdgfiosdfg it outputs "TEST"
12:51 < HugoDaniel> but if i have a serverPartT like dir "test/file" (return . toResponse $ "TEST2"), and go to http://localhost:8080/test/file it just fallsback to the next ServerPartT
12:51 < HugoDaniel> why ?
--- Log closed Sun Dec 20 13:02:41 2009
--- Log opened Sun Dec 20 13:02:41 2009
--- Log closed Sun Dec 20 13:02:41 2009
--- Log opened Sun Dec 20 13:02:45 2009
13:03 < HugoDaniel> ill just stop using dir
13:03 < HugoDaniel> its a pain
13:05 < bnijk_> HugoDaniel: http://cgi.4chan.org/gif/src/1261297594101.gif
13:05 < burp> *cabal haddock* :(
13:10 < HugoDaniel> bnijk_: thats exactly how i feel about dir
13:10 < HugoDaniel> and quite possibly fileServe :P
13:10 < bnijk_> oh really
13:14 < HugoDaniel> yeps
13:14 < HugoDaniel> just checked the source of dir
13:14 < HugoDaniel> its not that good
13:14 < bnijk_> is this the function for directory listings
13:14 < HugoDaniel> no
13:14 < HugoDaniel> for dir matching
13:14 < HugoDaniel> it only matches for the first string in a list of dirs
13:14 < HugoDaniel> thats why it doesn't match on the second example i provided
13:15 < HugoDaniel> line 913 of SimpleHTTP.hs
13:16 < HugoDaniel> anyway this should be easy to fix... let me give it a shot and send a patch
13:22 < jbe> For the record, I got happstack-tutorial installed.  Following the instructions at http://tutorial.happstack.com/tutorial/run-tutorial-locally did not work for me.  I had to pull wchogg's repo and "cabal install" from there: http://patch-tag.com/r/wchogg/happstack-tutorial/darcscommands
14:25 < stepcut> HugoDaniel: dir is not broken.. that is the way it supposed to work...
14:25 < stepcut> HugoDaniel: we split the path on the / and each dir looks at one part of the path
14:26 < stepcut> HugoDaniel: somewhere I have a 'dirs' function that can take something like "foo/bar/baz" and do what you expect
14:28 < stepcut>       dirs fp m =
14:28 < stepcut>           do let parts = splitDirectories (makeRelative "/" fp)
14:28 < stepcut>              foldr dir m parts
14:28 < stepcut>  
14:28 < stepcut> I thought I added that to SimpleHTTP already, but I guess not ;)
14:29 < HugoDaniel> ah ok
14:31 < HugoDaniel> dirs also doesn't work
14:32 < HugoDaniel> oh it does
14:32 < HugoDaniel> sorry
14:32 < HugoDaniel> my mistake again
14:38 < stepcut> :p
14:38 < stepcut> did your upgrade to 0.4 go smoothly/
14:39 < stepcut> compiling Haskell to actionscript bytecode will be much easier in the next version of flash, because they finally add tail-call optimization
14:40 < stepcut> maybe i'll actually finish my haskell->actionscript compiler then :p
14:41 < stepcut> unfortunately, adobe's ecmascript 4 spec got bonged by the javascript committee, so future versions of javascript will not be based on it :(
15:02 < mm_freak> hi there, sorry for asking in the wrong channel
15:02 < mm_freak> is there some way to integrate happstack into/with apache?
15:02 < mm_freak> (that is, without proxying)
15:05 < stepcut> mm_freak: not yet.. what would be the purpose?
15:06 < stepcut> mm_freak: there is fastcgi support, though it may need to be updated to work with 0.4
15:11 < mm_freak> stepcut: i need apache
15:12 < mm_freak> (until i want to reimplement the whole world)
15:12 < mm_freak> ((which is probably easy to do in haskell anyway))
15:15 < stepcut> currently we just use proxying
15:15 < burp> mod_happstack? :D
15:15 < burp> anyway there is probably nearly no performance penalty
15:16 < stepcut> burp: it's been done, though not with happstack.. http://www.st.ewi.tudelft.nl/~dolstra/mod_haskell/
15:20 < mm_freak> hmm
15:20 < mm_freak> i'll use proxying then
15:20 < stepcut> http://huygens.functor.nl/blog/?p=21
15:21 < stepcut> tutorial.happstack.com is actually proxied happstack server being proxied by the patch-tag happstack server :)
15:22 < mm_freak> hehe
15:31 < mm_freak> currently i'm a postgresql user
15:32 < mm_freak> do i have all those database features in happstack's builtin DBMS, too?
15:37 < burp> no
15:37 < burp> barely :>
15:39 < mm_freak> hehe ok
15:39 < burp> http://happstack.com/docs/0.4/happstack-ixset/Happstack-Data-IxSet.html
15:39 < burp> basically limited to this stuff
15:40 < stepcut> mm_freak: what do you mean by 'features'?
15:41 < mm_freak> stepcut: i use all that usual database stuff like referential integrity, indices, joins, partial result sets, etc.
15:42 < stepcut> happtsack provides a ACID data store, so you get the ACID propertiesp
15:43 < stepcut> possibly ... 'partial results' the results are already in memory, and when you do a query, you just get a reference to the existing data.. so all of Haskells laziness stuff applies
15:44 < stepcut> for relational table stuff, there is the IxSet data type -- but in happstack-state/macid, you can use (almost) any Haskell data type you want.. you are not limited to just tables with a few primative types
15:44 < stepcut> and instead of using SQL to query the database, you just write pure Haskell functions
15:47 < stepcut> so the question is less, 'does happstack have joins' and more 'how would I store this data in a convenient Haskell data structure'
15:47 < stepcut> maybe that means Data.Tree or Data.Map or some other type
15:51 < mm_freak> i understand
15:54 < stepcut> IxSet provides a Set with multiple indices -- which is suitable for many table type operations
15:55 < stepcut> if you really needed a full relational model, you could develop the types and functions and a library, unrelated to happstack, and just use it with happstack
15:55 < stepcut> I see people talk about implementing stuff like that in Haskell every now and then, but I don't think anyone has actually done it yet :-/
15:57 < stepcut> it's been my observation, though, that a lot of people use RDBMSs because they want ACID, and the relational mode is the price they have to pay :)
15:59 < stepcut> in fact, I think popularality of ORMs despite their problems, show that people don't really want to be working with a relational model -- they want to be working with a model that is native to their language
16:01 < HugoDaniel> i start my happstack ServerPartT's with a debug one, like this: http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5418#a5418
16:01 < HugoDaniel> is there a more complete debug ServerPartT's that i can use ?
16:02 < dons> stepcut: got most of happstack into Arch Linux, except for happstack package itself, which depends on hsx, which depends on an old version of haskell-src-exts
16:02 < dons> might be worth reviewing those dependencies
16:02 < stepcut> HugoDaniel: no? but if you write something we can put in in 0.5 :)
16:03 < dons> http://aur.archlinux.org/packages.php?ID=23713 but server is up to date
16:03 < stepcut> dons: unfortunately, HSX does not work with the new haskell-src-exts. Niklas says he will give a huge pile of love to whomever fixes that :)
16:03 < dons> you might want to remove it from the package then.
16:03 < dons> better not to have the feature, than to have it break all distros :)
16:04 < stepcut> dons: haskell supports multiple versions of haskell-src-exts, what's wrong with arch!
16:05 < dons> few distros support multiple packages
16:05 < stepcut> I would rather spend the effort updating hsx
16:05 < dons> so it will make distro packaging hard.
16:05 < stepcut> can't be that hard :p
16:05 < dons> well, depend on how important hsx is to the framework :)
16:06 < stepcut> to me, it is very important :)
16:06 < dons> working on xmonad, we have always tried to keep down deps to make packaging easier -- since that's key for adoption. but its different for happstack, since it isn't a user tool.
16:06 < stepcut> other people, not sure..
16:06 < dons> anyway, as it is, i can't package the happstack module due to this issue, as it would mean an old version of haskell-src-exts.
16:07 < stepcut> the happstack-{kitchensync} package is optional, and is the one that is supposed to depend on all the extra stuff you might want to use with happstack
16:07 < HugoDaniel> i can try to look @ hsx with haskell-src-exts
16:07 < dons> ok. so if we get as far as happstack-server, then that's still a usable system, right?
16:07 < stepcut> HugoDaniel: sweet! there is an HSP mailing list, you can ask Niklas for some guidance -- he usually responds quickly
16:07 < stepcut> dons: yes
16:07 < dons> good to know.
16:07 < dons> should i record this issue in a bug tracker?
16:08 < stepcut> dons: the happstack-{kitchensync} library is useful, but not essential. Unfontunately, I am not sure what to do about it. Either it depends on lots of libraries, or we split it into a dozen smaller packages. Both ways have their drawbacks
16:08 < dons> that's true.
16:08 < stepcut> feel free to file a bug.
16:09 < dons> its kind of like the 'xmonad-contrib'
16:09 < dons> bug url?
16:09 < dons> ?bug
16:09 < lambdabot> http://hackage.haskell.org/trac/ghc/newticket?type=bug
16:09 < dons> wrong project.
16:09 < stepcut> our requirement so far is that , cabal install happstack, must work cleanly
16:09 < HugoDaniel> :)
16:09 < stepcut> because most people aren't going to care about a few extra libraries anyway
16:09 < stepcut> but, interference with distro packaging is obviously an issue as well
16:10 < dons> http://code.google.com/p/happstack/
16:10 < HugoDaniel> i usually care about extra libs on my servers
16:10 < stepcut> http://code.google.com/p/happstack/issues/list
16:10 < stepcut> HugoDaniel: why?
16:10 < HugoDaniel> more points of failure
16:10 < stepcut> HugoDaniel: yeah.. tis an annoying issue
16:11 < stepcut> the related annoying problem is orphan instances
16:11 < stepcut> we want things like Data.Text to have Serialize instances, but that requires that either (a) Data.Text depends on happstack-data or (b) happstack-data depends on Data.Text
16:11 < stepcut> or we have some other optional library that contains a bunch of orphan instances
16:12 < stepcut> but then should that library depend on everything we want to have a Serialize instance for? Or do we have one extra package per data type that we want to add a Serialize instance for?
16:13 < stepcut> it's annoying to depend on everything in the world, and it is annoying to have dozens of tiny packages, and cabal does not and will not ever support having a single 'orphan instances' package where you optional compile only some of the modules
16:13 < HugoDaniel> :/
16:14 < HugoDaniel> brb, reboot
16:14 < stepcut> that type class issue is why we depend on hsx.. we want to provide a ToMessage instance for HSP stuff
16:14 < stepcut> HugoDaniel: regarding your servers -- you shouldn't be installing the libraries anyway, just the final executable
16:15 < HugoDaniel> happstack currrently runs in a jail
16:23 < stepcut> do you build the executable on the server outside the jail?
16:23 < stepcut> gotta run
16:23 < dons> what's the happstack logo... ?
16:24 < burp> and again down :|
16:25 < dons> http://archhaskell.wordpress.com/2009/12/20/happstack-0-4-a-scalable-framework-for-developing-web-applications-in-haskell/
16:25 < dons> in Arch.
16:54 < HugoDaniel> :)
16:55 < HugoDaniel> so when is the release date for happstack 0.5 ?
16:59 < dons> would be interesting to have a schedule release timetable.
17:01 < HugoDaniel> yes, with some sort of periodic release dates, like, for instance, 6 in 6 months or kind of..
17:18 < HugoDaniel> help please: http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5420#a5420
17:18 < HugoDaniel> i tried migrating into 0.4 and i get this erros
17:18 < HugoDaniel> otherwise the answer wont follow with html but with simple text
17:35 < HugoDaniel> setHeaderM "Content-Type" "text/html"
17:35 < HugoDaniel> now thats more like it :D
17:35 < HugoDaniel> i was wondering about a feature like this for quite some time, this is really great
--- Log opened Sun Dec 20 17:49:06 2009
--- Log opened Sun Dec 20 17:51:24 2009
--- Log closed Sun Dec 20 17:56:21 2009
--- Log opened Sun Dec 20 17:56:21 2009
--- Log closed Sun Dec 20 17:56:21 2009
--- Log opened Sun Dec 20 17:56:25 2009
--- Log closed Sun Dec 20 18:00:45 2009
--- Log opened Sun Dec 20 18:00:45 2009
--- Log closed Sun Dec 20 18:00:45 2009
--- Log opened Sun Dec 20 18:00:49 2009
--- Log closed Sun Dec 20 18:04:53 2009
--- Log opened Sun Dec 20 18:04:53 2009
--- Log closed Sun Dec 20 18:04:53 2009
--- Log opened Sun Dec 20 18:04:57 2009
18:08 < HugoDaniel> how do i return a html response ?
18:09 < HugoDaniel> im using "toHtml" but its not working, somehow its not calling the toResponse method of the Html data type
18:09 < mikste_> You mean as opposed to the default "text/plain"?
18:09 < HugoDaniel> yes
--- Log closed Sun Dec 20 18:10:26 2009
--- Log opened Sun Dec 20 18:10:26 2009
--- Log closed Sun Dec 20 18:10:26 2009
--- Log opened Sun Dec 20 18:10:29 2009
18:10 < mikste_> I'm using the HtmlString newtype from happstack-helpers.
--- Log closed Sun Dec 20 18:10:55 2009
--- Log opened Sun Dec 20 18:10:55 2009
--- Log closed Sun Dec 20 18:10:55 2009
--- Log opened Sun Dec 20 18:10:58 2009
18:11 < mikste_> It has an instance for ToMessage that changes the content type to "text/html"
18:12 < HugoDaniel> mikste_: on 0.4 ?
18:12 < mikste_> So I'll type something like (ok . HtmlString $ "Hello, World")
18:12 < mikste_> yes.
18:12 < HugoDaniel> ah
18:12 < HugoDaniel> thanks
18:12 < HugoDaniel> :)
18:12 < mikste_> I copied the the code from happstack-helpers locally into my codebase.
18:13 < HugoDaniel> i was doing toResponse.toHtml $ "blah"
18:14 < mikste_> The Text.Html module is used when you want to construct HTML programatically.
18:15 < HugoDaniel> its not the case
18:15 < mikste_> All you and I want to do is have strings interpreted as HTML by the web browser.
18:15 < HugoDaniel> indeed, but thats turning out not an easy task to do
18:16 < mikste_> So far I've only done basic things, so I'm not sure which is the better way to go.
18:17 < mikste_> I started with HStringTemplate since that's what the Happstack tutorial suggests.  After reading the original StringTemplate documentation it sounds like the way to go for me.
18:17 < HugoDaniel> well, im not being able to serve simple strings as html in 0.4
18:18 < HugoDaniel> im trying to read up on the code of simplehttp
--- Log closed Sun Dec 20 18:18:32 2009
--- Log opened Sun Dec 20 18:18:32 2009
--- Log closed Sun Dec 20 18:18:32 2009
--- Log opened Sun Dec 20 18:18:35 2009
18:19 < HugoDaniel> it doesn't call the "toResponse" of the type Html
18:19 < mikste_> http://hpaste.org/fastcgi/hpaste.fcgi/view?id=14627#a14627
18:19 < mikste_> That's what I use
--- Log closed Sun Dec 20 18:20:02 2009
--- Log opened Sun Dec 20 18:20:02 2009
--- Log closed Sun Dec 20 18:20:02 2009
--- Log opened Sun Dec 20 18:20:04 2009
18:20 < mikste_> I'm not using Text.Html, so what you need to do is probably different.
18:20 < HugoDaniel> well i just want simple strings to go as text/html
18:21 < HugoDaniel> ...and ill be fine
18:21 < HugoDaniel> after looking at the simplehttp code i found the instance of ToMessage for Html
18:21 < HugoDaniel> that was my approach
18:35 < HugoDaniel> :(
18:35 < HugoDaniel> still not working
18:39 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5424#a5424
18:39 < HugoDaniel> i always get hit on that error
18:39 < HugoDaniel> no matter what i try
18:41 < HugoDaniel> heck, ill just build the response myself and ditch this... its getting pretty frustrating
--- Log closed Sun Dec 20 18:43:37 2009
--- Log opened Sun Dec 20 18:43:37 2009
--- Log closed Sun Dec 20 18:43:37 2009
--- Log opened Sun Dec 20 18:43:40 2009
18:45 < mikste_> I don't think you should be calling toMessage directly like that.
18:45 < mikste_> toResponse does that for you.
18:45 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5426#a5426
18:46 < HugoDaniel> yeh
18:46 < HugoDaniel> it just feels like really broken :/
18:46 < HugoDaniel> im about to start using the http package directly :P
18:46 < HugoDaniel> or curl
18:52 < mikste_> Does anyone if it's possible to use a parameterized datatype with Happstack.State
18:53 < mikste_> http://hpaste.org/fastcgi/hpaste.fcgi/view?id=14642#a14642
18:53 < mikste_> This isn't working, but I don't know what the proper syntax should be if it's possible at all.
18:56 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5427#a5427
18:56 < HugoDaniel> its not working
18:56 < HugoDaniel> but its building out the content-type correctly
18:56 < HugoDaniel> ...which only makes it stranger
18:56 < mikste_> Oh, nevermind.  The compiler error I was getting was for something else in my code.  Adding in the necessary extensions seems to have triggered it.
18:57 < mikste_> HugoDaniel: There's a problem on line 1
18:57 < mikste_> You can't call (++) on a String and a Response
18:57 < HugoDaniel> im not
18:57 < mikste_> oh
18:57 < HugoDaniel> :P
18:57 < mikste_> right
18:57 < mm_freak> thanks people
18:57 < mm_freak> bye
18:59 < mikste_> HugoDaniel: I don't know why your output wouldn't start with "ANSWER" then.
18:59 < HugoDaniel> i see
18:59 < HugoDaniel> i was using the toHtml what just dumps the whole string in a nice <html></html> tags
18:59 < HugoDaniel> now im left in a dead end
19:00 < HugoDaniel> because in 0.4 HtmlString ain't working
19:00 < HugoDaniel> ...so its back to 0.3 for me i guess
19:01 < HugoDaniel> mikste_: i didnt paste "ANSWER", it was on a previous line, but it prints
19:01 < mikste_> oh
19:02 < mikste_> that makes sense then.
19:02 < HugoDaniel> bah
19:02 < mikste_> (show $ toResponse) turns your Response into a string.
19:02 < mikste_> I thought you were confused because the ANSWER bit wasn't showing.
19:03 < HugoDaniel> no
19:03 < HugoDaniel> im confused because its printing the html
19:03 < HugoDaniel> and not showing it correctly
19:03 < HugoDaniel> but thats because i was using the "toHtml" that wraps it all in <html> tags
19:03 < HugoDaniel> ...so im left with nothing
19:03 < HugoDaniel> because HtmlString seems to be borked in 0.4
19:05 < mikste_> Mind pasting a larger chunk of code?
19:05 < mikste_> I'm wrapping normal Strings into the HtmlString newtype wrapper, and it's working okay for me.  I might be able to spot something simple.
19:06 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5428#a5428
19:07 < HugoDaniel> it gives that strange "no instances" error
19:08 < HugoDaniel> mikste_: can you wrap a simple "<html><head><title>TEST</title></head><body>TESTTT</body></html>" and make it appear correctly ?
19:10 < mikste_> Okay, so it looks like you are using HStringTemplate too.
19:10 < mikste_> Take out that call to toHtml
19:10 < mikste_> toHtml :: (HTML a) => a -> Html
19:11 < HugoDaniel> done
19:12 < HugoDaniel> this seems like a strange dependency that was not correctly treated in 0.4
19:12 < HugoDaniel> ill try reinstalling syb
19:13 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5429#a5429 :(
19:13 < HugoDaniel> same error after reinstalling syb
19:13 < HugoDaniel> reinstalling happstack-helpers after reinstalling syb
19:14 < HugoDaniel> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5430#a5430
19:14 < HugoDaniel> still broken
19:14 < mikste_> I haven't tried using happstack-helpers since upgrading Happstack to 0.4.  I just copy a few bits as I need them.
19:15 < mikste_> let me dig up a short, working example for you to see...
19:15 < HugoDaniel> ok
19:16 < HugoDaniel> ill force installation of syb 0.1.0.1
19:16 < HugoDaniel> and try to reinstall happstack-helpers (again)
19:18 < mikste_> http://moonpatio.com/fastcgi/hpaste.fcgi/view?id=5431#a5431
19:18 < mikste_> This example doesn't need happstack-helpers
19:18 < HugoDaniel> hmmmf
19:18 < HugoDaniel> it works now
19:19 < HugoDaniel> it needed syb 0.1.0.1
19:19 < Raevel> these package version things are a killer :-(
19:19 < HugoDaniel> 1h30m spent on this :&
19:20 < mikste_> wow, no kiddin.g
19:20 < Raevel> hsp 0.4.5 failed to build for me now, it needed HJScript 0.4.4 and failed with 0.4.5
19:21 < mikste_> Well, I've got to go now.
--- Log closed Sun Dec 20 19:22:30 2009
--- Log opened Sun Dec 20 19:22:30 2009
--- Log closed Sun Dec 20 19:22:30 2009
--- Log opened Sun Dec 20 19:22:34 2009
19:51 < HugoDaniel> this dir, exactdir, and dirs are a bit awkward :/
19:51 < HugoDaniel> specially the way they handle the "/" char
19:51 < HugoDaniel> which one is? /dir, dir/, /dir/ how are browsers going to send them in ?
19:56 < HugoDaniel> exactdir requires the "/" char, while dir does not
--- Log closed Sun Dec 20 20:00:12 2009
--- Log opened Sun Dec 20 20:00:12 2009
--- Log closed Sun Dec 20 20:00:12 2009
--- Log opened Sun Dec 20 20:00:15 2009
--- Log closed Sun Dec 20 20:03:17 2009
--- Log opened Sun Dec 20 20:03:17 2009
--- Log closed Sun Dec 20 20:03:17 2009
--- Log opened Sun Dec 20 20:03:20 2009
--- Log closed Sun Dec 20 20:15:38 2009
--- Log opened Sun Dec 20 20:15:38 2009
--- Log closed Sun Dec 20 20:15:38 2009
--- Log opened Sun Dec 20 20:15:41 2009
--- Log closed Sun Dec 20 20:23:19 2009
--- Log opened Sun Dec 20 20:23:19 2009
--- Log closed Sun Dec 20 20:23:19 2009
--- Log opened Sun Dec 20 20:23:22 2009
--- Log closed Sun Dec 20 21:12:46 2009
--- Log opened Sun Dec 20 21:12:46 2009
--- Log closed Sun Dec 20 21:12:46 2009
--- Log opened Sun Dec 20 21:12:50 2009
21:31 < McManiaC> btw
21:31 < McManiaC> HTTP request failed with: <socket: 21>: hLookAhead: resource vanished (Connection reset by peer)
21:31 < McManiaC> what does this mean?
21:38 < dons> hmm.
--- Log closed Sun Dec 20 21:39:18 2009
--- Log opened Sun Dec 20 21:39:18 2009
--- Log closed Sun Dec 20 21:39:18 2009
--- Log opened Sun Dec 20 21:39:21 2009
22:03 < gwern> I see it sometimes with gitit. it doesn't seem to be important
22:22 < McManiaC> I see it all the time...
22:23 < McManiaC> everything seems to work so it really doesnt seem to be that important, but...
22:23 < McManiaC> ;)
--- Log closed Mon Dec 21 00:00:21 2009