Available formats: content-negotiated html turtle (see SIOC for the vocabulary)
Back to channel and daily index: content-negotiated html turtle
These logs are provided as an experiment in indexing discussions using IRCHub.py, Irc2RDF.hs, and SIOC.
| 00:44:57 | <alexj> | hey there. just heard that people are having trouble with searchpath. can I get details? |
| 03:36:33 | <killfire> | following this page http://happs.org/HAppS/HAppS-Begin/public/ I get the following error when trying to run sp runhaskell -ihaskell haskell/Main.hs : <interactive>:1:130: attempting to use module `Main' (haskell/Main.hs) which is not loaded |
| 03:36:55 | <killfire> | this is after searchpath has tracked down and downloaded all the various HAppS-* |
| 03:37:07 | <killfire> | is there something I am doing wrong? |
| 03:37:56 | <alexj> | happs-begin is not very current. how did you get there? |
| 03:38:06 | <killfire> | google happs... second result.. |
| 03:38:21 | <alexj> | ah interesting. ok i'll fix that and redirect. |
| 03:38:22 | <killfire> | and it was the only docs I could find on happs in the .9 series... |
| 03:38:35 | <alexj> | just go to http://happs.org |
| 03:38:49 | <alexj> | but i'll fix the google results. |
| 03:39:18 | <killfire> | oh so start from the happs.org project? |
| 03:39:44 | <alexj> | yes. |
| 03:40:02 | <alexj> | not sure it is current either. we just fixed a bunch of thing for the release coming up shortly. |
| 03:41:01 | <killfire> | hmm... getting an error that Happs/Store/Util.hs cant find Control.Monad.Reader.... |
| 03:41:22 | <killfire> | src/HAppS/Store/Util.hs:21:7: Could not find module `Control.Monad.Reader': Use -v to see a list of the files searched for. |
| 03:41:45 | <alexj> | ok that I need to fix. |
| 03:41:57 | <alexj> | but that is werid. you should havve control.monad.reader |
| 03:42:01 | <alexj> | which ghc? |
| 03:42:14 | <killfire> | 6.8.2 |
| 03:42:23 | <killfire> | built on openbsd from 6.6 |
| 03:42:39 | <alexj> | ok |
| 03:42:47 | <alexj> | why can't you import control.monad.reader |
| 03:43:29 | <killfire> | hmm... let me try to figure that out |
| 03:43:46 | <killfire> | because that (should be) a happs thing... |
| 03:43:57 | <killfire> | s/should/shouldnt |
| 03:49:07 | <killfire> | hmm.. the only place it is showing up is /usr/local/lib/ghc/imports/Control/Monad/Reader, not in base/ etc ... which doesnt seem right... maybe I should head over to #ghc? or get up to openbsd current I guess. |
| 03:49:48 | <killfire> | a standard ./configure ; make; make install shouldnt build without that... |
| 03:55:32 | <killfire> | also, is the example State+HTTP on the bottom of the front page up to date (in the Examples box) |
| 17:30:30 | <non_socialist> | ok doing the runghc triple on many a package |
| 17:30:31 | <non_socialist> | whew |
| 17:30:36 | <non_socialist> | server compiling now |
| 17:31:50 | <non_socialist> | wait a second |
| 17:45:34 | <non_socialist> | HAppS-Data, HAppS-IxSet, HAppS-Server, HAppS-State, HAppS-Util, syb-with-class here is the list from yesterday, all installed yay!!! I also se on the happs.org site: HAppS- DNS SMTP Plugins HTTP ?? |
| 17:45:45 | <non_socialist> | blam |
| 17:45:51 | <non_socialist> | let em go get those too |
| 17:45:55 | <non_socialist> | me- |
| 17:54:09 | <non_socialist> | darcs get --partial --tag=0.9.2 http://happs.org/repos/HAppS-HTTP this fails, doesn't liek tag name |
| 17:54:15 | <non_socialist> | ... |
| 18:30:48 | <MarcWeber> | non_socialist: What will be your first app? |
| 19:35:17 | <non_socialist> | good question |
| 19:35:49 | <non_socialist> | maybe a search page for work procedures? |
| 19:37:17 | <MarcWeber> | non_socialist: Like 1) Never start ? 2) after having done thinking about how to solve the problem consider it beeing solved? :) |
| 19:38:15 | <MarcWeber> | I mean I'm starting using HAppS and you are too, right? MAybe we can do something together? |
| 19:39:22 | <non_socialist> | hm |
| 19:39:24 | <non_socialist> | ok |
| 19:39:26 | <non_socialist> | well |
| 19:39:33 | <non_socialist> | let me see |
| 19:40:00 | <non_socialist> | I dotn have 4 more modules listed on the page but have 7 installed |
| 19:40:27 | <non_socialist> | don't have dns http plugins or smtp |
| 19:40:34 | <non_socialist> | google doesn't seem to know where to get em |
| 19:40:35 | <non_socialist> | hm |
| 19:40:50 | <non_socialist> | well letme look at start guide and see if I can start the happs I have sofar.. |
| 19:41:10 | <MarcWeber> | non_socialist: cd to the happs-server directory. You can find some examples there. |
| 19:41:20 | <MarcWeber> | They are up to date and should compile. |
| 19:42:03 | <non_socialist> | ok |
| 19:43:05 | <non_socialist> | Ok I see the examples, how do I start happs? |
| 19:43:06 | <non_socialist> | :) |
| 19:46:21 | <MarcWeber> | Just ./<your compiled program> |
| 19:46:48 | <MarcWeber> | Default port ist 8000 So http://localhost:8000 should get you started. Not the comments "interested urls' .. |
| 19:47:35 | <non_socialist> | intrested url? |
| 19:49:16 | <MarcWeber> | typo " interesting urls: |
| 19:49:56 | <non_socialist> | ok chmod 700 AllIn.hs, then ./AllIn.hs and barfs |
| 19:50:05 | <non_socialist> | hm |
| 19:50:16 | <Saizan> | you've to compile it first |
| 19:50:34 | <Saizan> | e.g. ghc --make AllIn.hs |
| 19:50:48 | <Saizan> | that will produce an AllIn executable |
| 19:51:12 | <Saizan> | which you can run |
| 19:51:41 | <non_socialist> | oh |
| 19:52:02 | <non_socialist> | should I first do runghc configure, runghc build, runghc install? |
| 19:53:09 | <MarcWeber> | Do you see any .cabal file? |
| 19:53:10 | <Saizan> | no, the ones in Examples are not a cabal package |
| 19:53:24 | <non_socialist> | no |
| 19:53:24 | <non_socialist> | ok |
| 19:53:34 | <MarcWeber> | Just use ghc --make file.hs |
| 19:53:42 | <MarcWeber> | maybe with --package XX --package YY |
| 19:54:32 | <non_socialist> | cmopiled using ghc --make AllIn.hs, then ./AllIn.......No suitable handler found |
| 19:54:56 | <non_socialist> | hm |
| 19:57:30 | <MarcWeber> | non_socialist: Try the ones in set. (HAven't understood all of AllIn yet myself) |
| 20:03:20 | <non_socialist> | do I have to compile all of the b4 they work? |
| 20:03:22 | <non_socialist> | or.. |
| 20:03:27 | <non_socialist> | I am starting with basics |
| 20:03:28 | <non_socialist> | heh |
| 20:05:16 | <non_socialist> | ok hello world works loll |
| 20:10:55 | <non_socialist> | ok I grabbed some pacakges using darcs |
| 20:11:02 | <non_socialist> | DNS SMTP Plugins HTTP |
| 20:11:14 | <non_socialist> | now where is the Setup.hs for these ones? |
| 20:11:22 | <non_socialist> | sem different format |
| 20:14:34 | <non_socialist> | hmmmmmm |
| 20:55:20 | <mightybyte> | I have the following data structure "data State = State {users :: Map String User} deriving (Show,Read,Typeable,Data)" |
| 20:55:40 | <mightybyte> | And I declared State to be an instance of Component. |
| 20:56:27 | <mightybyte> | But I'm getting an error "No instances for (Default (Map String User)..." |
| 20:56:52 | <mightybyte> | How do I make Map String User an instance of Default? |
| 20:58:20 | <mightybyte> | @seen Lemmih |
| 21:02:12 | <Saizan> | mightybyte: have you made a StartState instance for State? |
| 21:02:21 | <mightybyte> | Yes |
| 21:02:38 | <mightybyte> | Want me to hpaste it? |
| 21:06:26 | <mightybyte> | Pasted: http://hpaste.org/5643 |
| 21:11:32 | <Saizan> | oh, you shouldn't derive Default in deriveAll |
| 21:11:54 | <mightybyte> | Ok, I have tried it both ways. I also tried the commented out block instead of the deriveAll. |
| 21:13:00 | <mightybyte> | And when I do that, it gives me even more "No instances" messages. |
| 21:13:18 | <Saizan> | using deriveAll omitting ''Default? |
| 21:13:31 | <mightybyte> | AllIn uses deriveAll to derive Default |
| 21:13:53 | <Saizan> | does it use a Map? |
| 21:14:01 | <mightybyte> | No |
| 21:14:31 | <Saizan> | that's the point |
| 21:14:55 | <Saizan> | since Map has some non-exported consturctors the deriving mechanisms can't work well |
| 21:15:22 | <mightybyte> | Ok, I figured it had something to do with Map. |
| 21:15:45 | <mightybyte> | Ok, I'll try removing Default from the deriveAll call. |
| 21:17:13 | <Saizan> | deriveAll adds the "NewData" instance |
| 21:17:30 | <mightybyte> | When I do that, I get a no instances error for Default State and Default (Map String User) amongst others. |
| 21:18:16 | <Saizan> | arising from? |
| 21:19:38 | <mightybyte> | I think it's saying Component State |
| 21:20:20 | <mightybyte> | @hoogle Default |
| 21:23:37 | <mightybyte> | Maybe I just need to manually make a defaultValue method for Map |
| 21:25:50 | <Saizan> | that would work if you can |
| 21:27:36 | <mightybyte> | Why wouldn't I be able to? |
| 21:29:25 | <Saizan> | i tried for Seq once and couldn't because of class contexts but i don't remember the details |
| 21:30:45 | <mightybyte> | Hmmm, I may be experiencing the same thing. |
| 21:32:33 | <mightybyte> | Well this seems to be a problem. It seems like using Maps in a Component would be common. |
| 21:33:06 | <Saizan> | yeah |
| 21:33:29 | <mightybyte> | ...unless I'm supposed to use something like IxSet instead of Map. |
| 21:34:04 | <Saizan> | instance (Indexable a b, Data a, Ord a, Default a) => Default (IxSet a) where defaultValue = ISet [] |
| 21:34:56 | <Saizan> | that's the Default instance for IxSet |
| 21:36:43 | <Saizan> | but it also has a SYBWC.Data hand-written instance |
| 21:39:39 | <mightybyte> | Ahhh |
| 21:40:17 | <mightybyte> | Here's an interesting comment from the code |
| 21:40:35 | <mightybyte> | "If you want an instance for all types then import HAppS.Data.Default.Generic" |
| 21:40:45 | <mightybyte> | Talking about an instance of Default |
| 21:41:21 | <mightybyte> | And indeed, that seems to solve the problem. |
| 21:41:26 | <Saizan> | uh |
| 21:48:36 | <mightybyte> | Well, partially. |
| 21:49:15 | <mightybyte> | I still have instance errors for NormalizeD State, DefaultD State, and XmlD State |
| 21:49:41 | <mightybyte> | As well as those same 3 for the component User |
| 22:33:36 | <mightybyte> | Lemmih: Hey, I've been having trouble getting my component to instance Default. |
| 22:34:21 | <mightybyte> | http://hpaste.org/5643 |
| 22:38:30 | <Lemmih> | I see no error message. |
| 22:38:46 | <Lemmih> | Ah, yes. Data.Map |
| 22:39:31 | <mightybyte> | Yeah |
| 22:39:38 | <mightybyte> | Any suggestions? |
| 22:40:04 | <Lemmih> | Don't use Xml+Default. It is ridiculously fragile. |
| 22:40:44 | <Lemmih> | Why is User a component? |
| 22:40:55 | <Lemmih> | You don't want that. |
| 22:41:17 | <Lemmih> | Or well, I'm pretty sure you don't want that. |
| 22:41:31 | <Lemmih> | A component is kinda like a global IORef. |
| 22:45:03 | <mightybyte> | Oh, I thought User had to be a component because State was a Component. |
| 22:45:30 | <Lemmih> | User is just like Map or String. |
| 22:45:56 | <mightybyte> | Ok, I thought the components had to cascade down the tree. |
| 22:46:25 | <Lemmih> | They do. But User is not a component. |
| 22:46:42 | <mightybyte> | How do I not use Xml+Default? I didn't know I was using it. |
| 22:47:38 | <Lemmih> | Don't use 'deriveAll' and write your own instances for Serialize. |
| 22:51:46 | <mightybyte> | Ok, but that doesn't fix the Data.Map problem |
| 22:52:26 | <Lemmih> | Right, it just circumvents it. |
| 22:53:13 | <Lemmih> | Oh and 'deriveNewData' has to go as well. |
| 22:54:33 | <mightybyte> | Ahhhh |
| 22:54:53 | <mightybyte> | But keep mkMethods? |
| 22:55:06 | <Lemmih> | Yes, mkMethods is good. |
| 22:55:29 | <mightybyte> | Now it's telling me that I need a Component User instance |
| 22:55:39 | <mightybyte> | Do I need to take User out of the dependency list? |
| 22:55:41 | <Lemmih> | Yes. |
| 22:55:45 | <mightybyte> | Ok |
| 22:58:02 | <mightybyte> | I just posted an update. http://hpaste.org/5643#a3 |
| 22:58:08 | <mightybyte> | Still won't compile. |
| 22:58:37 | <mightybyte> | It's complaining about the same 3 instances. NormalizeD, DefaultD, and XmlD. |
| 22:59:25 | <Lemmih> | You need an instance for Serialize. |
| 22:59:59 | <mightybyte> | Ahh. Why isn't the compiler telling me that? |
| 23:00:02 | <Lemmih> | It tries to use Xml by default which results in less than optimal error messages. |
| 23:02:56 | <mightybyte> | Ok. So is "instance (Serialize State) => Component State" enough or do I have to create it manually? |
| 23:03:23 | <Lemmih> | instance Serialize State where ... |
| 23:03:29 | <Lemmih> | Yes, the latter. |
| 23:04:26 | <Lemmih> | instance Serialize State where putCopy (State users) = contain $ safePut users; getCopy = contain $ liftM State safeGet |
| 23:04:32 | <Lemmih> | Likewise for User. |
| 23:05:22 | <Lemmih> | DrIFT or TH can be used to ease the burden. |
| 23:05:44 | <mightybyte> | Ok |
| 23:08:15 | <mightybyte> | Is Serialize defined in HAppS or an external library? |
| 23:08:25 | <Lemmih> | It's defined in HAppS. |
| 23:09:25 | <alexj> | it would be nice if it was defined somewhere else. we really should not have to be thinking about serialiatiion ourselves. |
| 23:09:26 | <Lemmih> | Importing HAppS.State should give you everything you need. |
| 23:10:46 | <mightybyte> | Ok, I see it. |
| 23:11:24 | <Lemmih> | alexj: Searchpath shouldn't try to use tag 0.9.2 anymore. |
| 23:12:05 | <alexj> | ok. i'll take all the tags out until we do a release or should I just use a different set of tags? |
| 23:12:15 | <mightybyte> | How do I modify that instance line for User? Not sure how safePut should be used for two arguments. |
| 23:12:16 | <alexj> | do you want just to send me a patch of default.map? |
| 23:13:49 | <Lemmih> | mightybyte: putCopy (User name passwd) = contain $ put name >> put passwd |
| 23:14:13 | <Lemmih> | getCopy = contain $ liftM2 User get get |
| 23:14:43 | <mightybyte> | Ok, get and put being from the State monad? |
| 23:15:07 | <Lemmih> | No, from Data.Binary. You can use safeGet and safePut instead. |
| 23:16:18 | <mightybyte> | Ok, I'll play around with this and see where I get. Thanks for the help. |
| 23:19:23 | <Lemmih> | alexj: I'm not sure what repo default.map is in. |
| 23:19:53 | <alexj> | searchpath |
Back to channel and daily index: content-negotiated html turtle