--- Log opened Tue Apr 21 00:00:34 2009
00:00 < mightybyte> Do you know if anyone runs live sites from within ghci?
00:01 < mightybyte> Seems like having the repl there might allow some nice on-the-fly mods.
00:09 < mightybyte> mae: Do you know a way to prevent a type from getting indexed by inferIxSet?
00:10 < mightybyte> data Item = Item { a :: Foo, b :: Bar, c :: [Foo] }
00:11 < mightybyte> $(inferIxSet "ItemDB" ''Item 'noCalcs [''Foo])
00:12 < mightybyte> Items in c interefere in some way with the intended index by item a.
00:13 < mightybyte> It still does this if you make "c :: [Funk]" where "newtype Funk = Funk { unFunk :: Foo } deriving..."
03:08 < mae> hallo
03:09 < TheRealMarko> Hi
03:10 < mae> ok, fixed a bug with the lockfile patch
03:10 < mae> works good for me
03:25 < h_buildbot> Build for ghc-6.8.3 OK. Test suite ran. 14/77 test cases failed. Check http://buildbot.happstack.com/ for details.
03:32 < h_buildbot> Build for ghc-6.10.2 OK. Test suite ran. 14/77 test cases failed. Check http://buildbot.happstack.com/ for details.
04:03 < mae> just pushed patch for auto-build
04:03 < mae> we need to test 3 things now to get release out
04:04 < mae> 1. locking mechanism, 2. auto builder (it needs to kill more gracefully -- or we need to handle harsh killings better) 3. haddock needs to build on everything with 6.10.2 (this way haddocks work on hackage)
04:05 < mae> wait actually, scratch 2, the killing mechanism is find, it was my ctrl-c that didn't allow the lockfile to be deleted
04:05 < mae> do we need to capture exit signals to make the lockfile not as annoying?
04:05 < mae> find == fine
04:05 < mae> man i'm tired
04:06 < mae> well anyways, be back tommorrow
04:19 < h_buildbot> Build for ghc-6.8.3 OK. Test suite ran. 14/77 test cases failed. Check http://buildbot.happstack.com/ for details.
04:26 < h_buildbot> Build for ghc-6.10.2 OK. Test suite ran. 14/77 test cases failed. Check http://buildbot.happstack.com/ for details.
09:21 < stepcut> mae: doesn't waitForTerminate already catch killing signals for graceful shutdown purposes?
10:13 < mae> stepcut: yeah, i think i need to add that to Happstack.Util.AutoGen, right now the loop gets killed abnormally if you ctrl-c it
10:14 < mae> gotta run to work
10:14 < mae> bbl
10:14 < mae> err AutoBuild
10:14 < mae> not autogen
10:14 < mae> also, if you could tweak that error message, that would be helpful
10:14 < mae> right now it is vague about where the lock file is located
10:44 < mae> i am heading to work, please try this command on the head dev branch if anyone can:
10:46 < mae> (inside the example app): happstack build auto "sh bin/build.sh" ./app
11:39 < neoswish> how to compile Happstack.Helpers
11:39 < neoswish> on windows
11:44 < Zao> I didn't have much luck at all last I tried building HappStack on Windows.
11:45 < neoswish> Happstack.Server is compiled ok, but i need some functionality from Helpers, especially exactdir to match root path
12:03 < wchogg> Sorry, but I think happstack-helpers is still relying on some functionality from the FileManip package.  I'm going to try & replicate what I did to make Happstack build on Windows and just port the few functions I need over to use unix-compat instead of unix.
12:03 < wchogg> It'll probably be fixed sometime this week.
12:03 < wchogg> Then happstack-tutorial can finally build on windows again...oof.
12:04 < TomMD> Is Happstack hoping to have TLS support some day?
12:05 < stepcut> TomMD: we have TLS today! via apache reverse proxy...
12:06 < stepcut> TomMD: or, the new FastCGI support I suspect
12:07 < wchogg> stepcut : you played with the fastcgi package yet?
12:08 < stepcut> nope
12:10 < stepcut> TomMD: native TLS support would be nice though, since one of the foundational ideas behind happstack is to reduce IT overhead by getting rid of all the extra servers (apache, mysql, etc) and having a single, easy to deploy binary
12:11 < TomMD> stepcut: thats exactly what I ment.  So TLS patches would be accepted if they were in a reasonable state, I take it.
12:12 < Zao> What's TLS in this context, as I assume it's not Thread-Local Storage.
12:14 < stepcut> TomMD: I would think so
12:14 < stepcut> transport layer security
12:14 < stepcut> used for https
12:18 < TomMD> The RFC is just 100 pages too - do you know off-hand if the TLS 1.2 RFC is more or less self contained?
12:19 < TomMD> Nevermind - it looks pretty much stand alone.
12:19  * stepcut has no idea
--- Log closed Wed Apr 22 00:00:35 2009