--- Log opened Tue Mar 17 00:00:41 2009
00:33 < mae> wow
00:33 < mae> netsplit
00:33 < mae> ?
00:41 < mae> stepcut or saizan: question, do you think that libraries under the "happstack" umbrella which don't have any changes by 0.3 time should not get a version bump?, i.e. happstack will depend on an older version
00:41 < mae> to communicate that there are not any api changes
00:41 < mae> for that particular library
16:25 < solidsnack> ?localtime jsn
16:25 < solidsnack> ?localtime solidsnack
16:26 < solidsnack> mae: I've released the paper. Please let me know what you think. http://github.com/jsnx/members-only/raw/a5b71b24502bea051448e9b8ac0c6c9b941eee6a/notes/Sharding%20With%20Trees
16:31 < stepcut> solidsnack: cool
16:47 < solidsnack> stepcut: Do let me know what you think of it.
16:59 < stepcut> solidsnack: yeah, I am in a busy phase right now, but I do want to read it
17:10 < solidsnack> Aye.
17:52 < guenni> good morning!
17:52 < guenni> any cute guys present?
17:52 < guenni> oops, fc!
17:53 < guenni> serious now, I'd like to talk about Happs-ixset
17:53 < guenni> take me to your leader!
17:56 < guenni> mae: ping
18:00 < mightybyte> guenni: I'd help, but I have to go.  The channel is sparse at times, but when people are here we're generally pretty helpful.
18:01 < guenni> mightybyte: thx, but I was actually gonna offer my help, or rather participation
18:02 < guenni> if welcom
18:05 < solidsnack> guenni: This channel is logged. Leave a message.
18:06 < guenni> solidsnack: thx I will
20:08 < stepcut> ugh, there is a 'bug' in our cron function
20:11 < stepcut> threadDelay :: Int -> IO () where Int is in microseconds means that the longest time interval between runs is 35mins ? ((maxBound :: Int) `div` (10^6)) `div` 60
22:14 < corruptmemory> maybe this has been covered somewhere, but I've got a really strange situation
22:15 < corruptmemory> i have gitit and happstack 0.2.1 running behind an apache server
22:15 < corruptmemory> if I hit the apache server from the "inside" of the network everything is fine
22:16 < corruptmemory> hit it from the "outside" I get => Happstack 0.2.1 File not found
22:16 < corruptmemory> The URI's are the same.  In fact gitit doesn't even know what's "inside" or "outside" so I don't get it.
22:17 < corruptmemory> Anyone see this before?
22:17 < corruptmemory> Here's the url if anyone wants to try: http://www.corruptmemory.com/
22:26 < corruptmemory> is there possibly something about user agents and happstack?
22:27 < corruptmemory> Huh HA!  compression.
22:27 < corruptmemory> Probably something with gitit
22:27 < corruptmemory> sorry folks
22:27 < corruptmemory> if the client supports compressed responses everything is fine.
22:30 < gwern> hm, we thought we had covered the case where a client didn't support compressed
22:32 < corruptmemory> well, you can hit the site: wget -O test.html http://www.corruptmemory.com/
22:32 < corruptmemory> then curl --compressed http://www.corruptmemory.com/
22:33 < corruptmemory> first fails
22:33 < corruptmemory> second succeeds
22:33 < corruptmemory> looking at code now
22:33 < corruptmemory> but thoughts are welcome
22:34 < corruptmemory> This is a git-pull of gitit from yesterday.  No updates today
22:34 < corruptmemory> happstack installed from cabal
22:34 < corruptmemory> via cabal ;-)
22:34 < corruptmemory> from hackage
22:45 < mle> when I run cabal haddock on happstack-server, I get haddock: internal Haddock or GHC error: Maybe.fromJust: Nothing
22:46 < mle> or most of these really
22:46 < mle> is this normal, a known issue?
22:48 < mle> and is there a straightforward way to do https?
22:50 < stepcut> mle: haddock is broken, the docs only build with the latest version of GHC. When 6.10.2 is out it should work.
22:50 < stepcut> mle: we have prebuilt the docs for the latest stable release here, http://happstack.com/docs/0.2/index.html
22:51 < mle> ah, thanks!
22:52 < stepcut> for https you can use apache and setup a reverse proxy. Which is easier than it sounds. Someone wrote a tutorial on how to do it with happstack, but I can't find it off hand.
22:52 < gwern> the default assumption with haddock is that haddock is broken and not your markup
22:53 < stepcut> gwern: :)
22:53 < mle> gwern: a fromJust error generally indicates that, yeah
22:53 < stepcut> mle: I believe koeien wrote the tutorial on https + happstack, so if you see him, ask
22:55 < mle> okay
22:56 < stepcut> mle: or, if you are impatient, just read some random tutorial on mod_proxy and ReverseProxy, supposedly there is nothing special about using it with happstack
22:56 < mle> it sounds fairly standard... not going to worry about it until I need it.
22:57 < stepcut> grr, facebook is all screwy today
23:03 < gwern> hm, so hartman's blog post is kind of 'I don't like filestore being a record'...
23:04 < gwern> the irony is that filestore originally was a typeclass, not a record
23:04 < gwern> I forget why we changed it
23:07 < stepcut> :)
--- Log closed Wed Mar 18 00:00:43 2009