--- Log opened Thu Apr 02 00:00:07 2009
01:20 < altmattr> hi happs'ers, can you sell me on happstack?  I know, big question, I just want to know the features to look into first
01:20 < altmattr> also, how does the data-persistence store my data?  Just a little technical detail?
01:27 < sm> hi altmattr, did you see happstack.com ?
01:27 < altmattr> sm: yep, poked around a little
01:27 < altmattr> sm: always better to talk to users I think :)
01:28  * sm does too.. indeed there isn't much high-level overview there
01:28 < sm> I won't do a big sales job just now, but I can throw out featuroids or answer qs
01:28 < altmattr> :)
01:29 < sm> the data persistence writes your haskell types to disk in a custom transactional format.. not all happs users use that
01:29 < sm> contrary to what some folks might think!
01:35 < sm> alrighty then
04:34 < altmattr> sm: sorry about vanishing before
04:34 < altmattr> sm: thanks for the comments
04:34 < altmattr> :)
04:34 < sm> no problem
07:03 < dcoutts> mae: sounds like the stdcall feature would be quite easy actually
07:04 < dcoutts> Language.C already records the attributes
12:02 < h_buildbot> Build for ghc-6.8.3 OK. Test suite ran. 14/77 test cases failed. Check http://buildbot.happstack.com/ for details.
12:15 < h_buildbot> Build for ghc-6.10.1 OK. Test suite ran. 14/77 test cases failed. Check http://buildbot.happstack.com/ for details.
12:32 < h_buildbot> Build for ghc-6.8.3 OK. Test suite ran. 15/77 test cases failed. Check http://buildbot.happstack.com/ for details.
13:44 < h_buildbot> Build for ghc-6.8.3 OK. Test suite ran. 15/77 test cases failed. Check http://buildbot.happstack.com/ for details.
13:51 < h_buildbot> Build for ghc-6.10.1 OK. Test suite ran. 14/77 test cases failed. Check http://buildbot.happstack.com/ for details.
16:40 < TomMD> I earlier commented on a persistence issue and have now made a basic (fifty line) program to show what I'm running into
16:41 < TomMD> I'm sure this is me mis-reading or perhaps completely missing some part of Happstack tutorial, but if someone could look at the code (I'll paste) and let me know then I'd appreciate it.
16:41 < TomMD> @paste
16:42 < TomMD> The code is at: http://hpaste.org/fastcgi/hpaste.fcgi/view?id=3227#a3227
16:47 < mightybyte> Is there any documentation describing the TH convenience functions in Happstack?
16:48 < wchogg> mightybyte : there might be more in head.  I've added a good bit of docs.
16:48 < wchogg> I'm not sure in particular which functions you mean
16:48 < wchogg> TomMD : So nothing is getting saved?
16:48 < mightybyte> Well, inferIxSet, deriveSerialize, and deriveAll for starters.
16:49 < wchogg> mightybyte : I think there's doc strings for all those now.  If more explanation is needed, then I'll add to the docs for 0.3
16:50 < wchogg> TomMD : but if you use startSystemState instead I presume it works, yes?
16:51 < mightybyte> wchogg: And I just recently ran across inferRecordUpdaters which sounds useful.
16:52 < wchogg> mightybyte : yeah, that's also documented a bit more in the darcs head
16:52 < mightybyte> wchogg: I don't see a doc string for inferIxSet.
16:52 < wchogg> mightbyte : well damn, I'll fix that
16:52 < mightybyte> wchogg: Or is it only in the development repo?
16:54 < wchogg> mightybyte : No, you're right, somehow I missed that one.
16:54 < mightybyte> Ok
16:54 < wchogg> thanks for noticing
16:54 < mightybyte> Yeah, TH is still voodoo to me otherwise I'd contribute.
16:55 < mightybyte> I was just trying to figure out how to use calculated indexes.
16:59 < mightybyte> I see how to do it if you're not using inferIxSet, but I wasn't sure how with it.
16:59 < mightybyte> What's the difference between the '' and ' prefixes in the TH calls?
17:01 < wchogg> I'm a little shaky on that distinction myself, honestly.
17:03 < mightybyte> I see in the inferIxSet call that 'noCalcs is the only one with one quote.  The rest have two.  So I wasn't sure what exactly to put there.
17:04 < wchogg> Ah, I think you want to give the calcs method with just a single tick
17:05 < wchogg> That might be the distinction, now that I think about it, that types are '' and values are '
17:05 < mightybyte> Ahhh, ok.  That makes sense.
17:06 < mightybyte> So instead of noCalcs I just put my own function there?
17:06 < wchogg> right
17:06 < wchogg> preceded by '
17:06 < mightybyte> f :: a -> (,) --a tuple of some size.
17:07 < mightybyte> So in IxSet Foo, I'd do $(inferIxSet "FooDB" ''Foo 'f [''Key])
17:08 < mightybyte> With "f :: Foo -> (Key1, Key2, ...)" rather
17:09 < wchogg> I _think_ that's right.
17:09 < mightybyte> Ok, I'll give that a try.
17:16 < mightybyte> Well...it compiles.
20:39 < sm> let me ask here too - do you think folks installing my project with cabal are going to be annoyed by the additional time/hassle required to download and build happstack and deps, if it's not providing core features ?
20:40 < sm> I'd like to enable the web interface out of the box, but not having happstack in the dependencies makes it an easier install. Or does it
20:49  * sm thinks: off by default until it's a core feature
21:50 < sm> folks, what is the way to match on / ? I saw it mentioned recently
21:51 < MarcWeber> Are you talking about foo/dir/ ? I'm not sure. In the past there was one patch adding the full url. You could just match the last char..
21:51 < MarcWeber> But I don't know exactly
21:55 < gwern> @seen mae
21:56 < gwern> oh. no lb
22:05 < dancor> why isn't lb in here
22:31 < gwern> guess whoever is running lb hasn't picked up the darcs repo changes adding #happs to default
23:34 < dibblego> do any of you guys do consulting writing applications with happstack?
23:40 < mightybyte> Possibly :)
23:46 < dibblego> possibly?
23:47 < mightybyte> Well, I haven't been aware of a market for happstack developers before now, but I might be interested.
23:49 < dibblego> I'm more interested in a market where the consulting client doesn't dictate for or against HAppS
23:50 < mightybyte> Ahhh.  Well, most people would probably consider that a dicey choice of frameworks.
23:50 < mightybyte> But recently, there's been a lot of improvement.
23:56 < mightybyte> Contract development isn't what I do full-time, but I might be interested in a project.
--- Log closed Fri Apr 03 00:00:06 2009