Nabble - Haskell - Haskell-Cafe

Re: Re: How does cabal determine version ranges in the dependencies?

2009-11-23T10:38:49Z

On Mon, 2009-11-23 at 09:50 -0200, Maurício CA wrote:

> > I recently started porting cabal-install to Freebsd. When
> > I looked at its dependencies on hackage, I noticed HTTP
> > (>=4000.0.2 && <4001). However the latest HTTP version on
> > hackage is 4000.0.8. That struck me as kinda odd. How can cabal
> > tell that it won't be compatible with HTTP version 4001?
>
> If it doesn't break dependencies, it won't be called http 4001,
> it will be called 4000.0.9 :)
>
> Check:
>
> http://www.haskell.org/haskellwiki/Package_versioning_policy

I see. Thanks for the link. I wasn't aware of the versioning policy.
Just to clarify though, wouldn't the next higher major version by 4000.1
rather than 4001?

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Re: Wiki software?

2009-11-23T10:29:18Z

On Sun, Nov 22, 2009 at 11:56 AM, Antoine Latter <aslatter@...> wrote:

>
> Running 'pandoc --strict' over the Markdown readme.text takes:
>
> ~0.09s with pandoc built against parsec-2
> ~0.19s with pandoc built against parsec-3
>
> on my machine.
>
> I have a branch of parsec-3 which seems to brings us back to parsec-2
> numbers, but also fails the rst-reader test-case in the pandoc testing
> suite:
>
> http://community.haskell.org/~aslatter/code/parsec/cps

In reply to my own post, the branch of parsec posted now passes all of
the pandoc test cases.

If there are any other consumers of the parsec library that have tests
I can run let me know.

The 'many' combinator is one of those things that can look right, be
wrong, yet work for almost everything.

Antoine
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

RE: Re: Idea for a very simple GUI llibrary

2009-11-23T10:04:02Z

RE: [Haskell-cafe] Re: Idea for a very simple GUI llibrary

Thinking of a parallel with Java for a second, is there a GUI library out there that's structured like Java Swing? Meaning, there is a GUI library that has a small platform-specific GUI foundation (e.g. a per platform implementation of the core AWT functionality) and the rest of the functionality is pure haskell?

Supporting cross platform guis is often a bit ... complicated. Java attempted to resolve their debug-everywhere nightmare with AWT by making the per-platform bit as small as possible, and building everything else in Java.

I guess in theory gtk and wxWidgets take on this support burden, but you do get some fairly hefty imperative apis as a result. Perhaps it would make sense to focus efforts on stabilising a small 'core gui' library that can act as the foundation stone for all manner of pure haskell gui libraries?*

Or perhaps this already exists?

Just a thought.

Cheers,
Sam

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Re[2]: Re: Idea for a very simple GUI llibrary

2009-11-23T08:19:02Z

Yeah, I found many pages about it (some making fun of a interpreter that "follows all H98 spec"). But i'm still crawling to understand who is who and what are the pages to get help in Haskell community :D

And well, tha's bad, i've really enjoyed Hugs, but i'll have to use GHC instead :D

Thanks.

On Mon, Nov 23, 2009 at 2:14 PM, Bulat Ziganshin <bulat.ziganshin@...> wrote:

Hello Juan,

Monday, November 23, 2009, 7:01:39 PM, you wrote:

> But in HUGS i can't. It says:
> ERROR "conjunction.hs":1 - Unrecognised character `\8743'

hugs doesn't accept unicode source files

there are lots of unicode support problems in both haskell
implementations. probably we have some wiki page what describes current
state

--
Best regards,
Bulat mailto:Bulat.Ziganshin@...

--
Juan Maiz Lulkin Flores da Cunha
---------------------------------------------------------------------------
Softa Consultoria para Desenvolvimento
http://www.softa.com.br
http://www.mailee.me - Finalmente e-mail marketing 2.0
http://www.linkedin.com/in/juanmaiz
http://workingwithrails.com/recommendation/new/person/9354-juan-maiz
“The most exciting breakthroughs of the 21st century will not occur because of technology but because of an expanding concept of what it means to be human”
John Naisbitt

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re[2]: Re: Idea for a very simple GUI llibrary

2009-11-23T08:14:24Z

Hello Juan,

Monday, November 23, 2009, 7:01:39 PM, you wrote:

> But in HUGS i can't. It says:
> ERROR "conjunction.hs":1 - Unrecognised character `\8743'

hugs doesn't accept unicode source files

there are lots of unicode support problems in both haskell
implementations. probably we have some wiki page what describes current
state

--
Best regards,
Bulat mailto:Bulat.Ziganshin@...

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Re: Idea for a very simple GUI llibrary

2009-11-23T08:02:24Z

Forgot the URL: http://github.com/softa/rl

On Mon, Nov 23, 2009 at 2:01 PM, Juan Maiz <juanmaiz@...> wrote:

Hi folks, my name is Juan Maiz and i'm starting to study Haskell (again). Is anyone from Brazil in the list?i

I'm currently reading The Haskell Road to Logic, Maths and Programming and having a lot of (geek) fun. By the way, i found this: http://www.mail-archive.com/haskell-cafe@.../msg45406.html . It was sent by Mattias Bengtsson. I made something similar using Ruby's Treetop, with the difference that my concern was to validate and generate truth tables for formulas using Unicode.

So... my first noob question to the list: What is the difference of Unicode support in HUGS and GHCI ? In GHCI I can do something like:

(∧) = (&&)
Prelude> True ∧ False
False

(The carachter is the unicode for conjunction U+2227)

But in HUGS i can't. It says:

ERROR "conjunction.hs":1 - Unrecognised character `\8743'

Somebody?

On Mon, Nov 23, 2009 at 1:26 PM, Eric Kow <eric.kow@...> wrote:

Minor aside.

> -> Could be the first GUI to build on hackage :)

If you have wxWidgets installed, the new fully Cabalised wxHaskell
builds just fine. It's quite handy/refreshing for 'cabal install wx' to
finally "just work" :-)

Unless you mean "build on the Hackage server" which should also be
possible in principle, although the wxHaskell folks may want to have
a quick look at
http://hackage.haskell.org/packages/archive/wxcore/0.12.1.2/logs/failure/ghc-6.10

--
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAksKqZgACgkQBUrOwgisBPmGBQCfRnW1dcLV3JaLi2p5PppsO+XK
7uwAoMVhCV00sdNctgAjw2TyQGs6TyNv
=tnXa
-----END PGP SIGNATURE-----

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

--
Juan Maiz Lulkin Flores da Cunha
---------------------------------------------------------------------------
Softa Consultoria para Desenvolvimento
http://www.softa.com.br
http://www.mailee.me - Finalmente e-mail marketing 2.0
http://www.linkedin.com/in/juanmaiz
http://workingwithrails.com/recommendation/new/person/9354-juan-maiz
“The most exciting breakthroughs of the 21st century will not occur because of technology but because of an expanding concept of what it means to be human”
John Naisbitt

Re: Re: Idea for a very simple GUI llibrary

2009-11-23T08:01:39Z

Hi folks, my name is Juan Maiz and i'm starting to study Haskell (again). Is anyone from Brazil in the list?i

I'm currently reading The Haskell Road to Logic, Maths and Programming and having a lot of (geek) fun. By the way, i found this: http://www.mail-archive.com/haskell-cafe@.../msg45406.html . It was sent by Mattias Bengtsson. I made something similar using Ruby's Treetop, with the difference that my concern was to validate and generate truth tables for formulas using Unicode.

So... my first noob question to the list: What is the difference of Unicode support in HUGS and GHCI ? In GHCI I can do something like:

(∧) = (&&)
Prelude> True ∧ False
False

(The carachter is the unicode for conjunction U+2227)

But in HUGS i can't. It says:

ERROR "conjunction.hs":1 - Unrecognised character `\8743'

Somebody?

On Mon, Nov 23, 2009 at 1:26 PM, Eric Kow <eric.kow@...> wrote:

Minor aside.

> -> Could be the first GUI to build on hackage :)

If you have wxWidgets installed, the new fully Cabalised wxHaskell
builds just fine. It's quite handy/refreshing for 'cabal install wx' to
finally "just work" :-)

Unless you mean "build on the Hackage server" which should also be
possible in principle, although the wxHaskell folks may want to have
a quick look at
http://hackage.haskell.org/packages/archive/wxcore/0.12.1.2/logs/failure/ghc-6.10

--
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAksKqZgACgkQBUrOwgisBPmGBQCfRnW1dcLV3JaLi2p5PppsO+XK
7uwAoMVhCV00sdNctgAjw2TyQGs6TyNv
=tnXa
-----END PGP SIGNATURE-----

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: ANNOUNCE: deepseq-1.0.0.0

2009-11-23T07:47:58Z

On 19/11/09 12:17, Simon Marlow wrote:

> Ok, unless there are any further objections, I'll change the names back to
>
> class NFData a where
> rnf :: a -> ()
>
> and also add
>
> deepseq :: a -> b -> b
>
> but I'll leave the module name as Control.DeepSeq.

I made this change and uploaded deepseq-1.1.0.0 on Friday.

There's also an updated parallel-2.1.0.0, but after discussions with
Phil Trinder and other Parallel Haskell gurus, I think there may be
further changes forthcoming. Upshot: the new version of parallel is
still changing, you might want to wait until things settle down
(hopefully not long) before switching.

Cheers,
Simon
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Idea for a very simple GUI llibrary

2009-11-23T07:26:18Z

Minor aside.

> -> Could be the first GUI to build on hackage :)

If you have wxWidgets installed, the new fully Cabalised wxHaskell
builds just fine. It's quite handy/refreshing for 'cabal install wx' to
finally "just work" :-)

Unless you mean "build on the Hackage server" which should also be
possible in principle, although the wxHaskell folks may want to have
a quick look at
http://hackage.haskell.org/packages/archive/wxcore/0.12.1.2/logs/failure/ghc-6.10

--
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

signature.asc (204 bytes) Download Attachment

Kind polymorphism

2009-11-23T07:09:09Z

Hello,

Are there currently any known problems that would hinder the
implementation of kind polymorphism [1], e.g. unresolved inelegancies or
technical limitations, or is it only a matter of finding the time to
implement it?

Thanks,

Martijn.

[1] http://hackage.haskell.org/trac/haskell-prime/wiki/KindInference
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: How does cabal determine version ranges in the dependencies?

2009-11-23T03:50:07Z

> I recently started porting cabal-install to Freebsd. When
> I looked at its dependencies on hackage, I noticed HTTP
> (>=4000.0.2 && <4001). However the latest HTTP version on
> hackage is 4000.0.8. That struck me as kinda odd. How can cabal
> tell that it won't be compatible with HTTP version 4001?

If it doesn't break dependencies, it won't be called http 4001,
it will be called 4000.0.9 :)

Check:

http://www.haskell.org/haskellwiki/Package_versioning_policy

Best,
Maurício

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

How does cabal determine version ranges in the dependencies?

2009-11-23T03:36:55Z

Hi everyone

I recently started porting cabal-install to Freebsd. When I looked at its dependencies on hackage, I noticed HTTP (>=4000.0.2 && <4001). However the latest HTTP version on hackage is 4000.0.8. That struck me as kinda odd. How can cabal tell that it won't be compatible with HTTP version 4001?

regards
david
--
Preisknaller: GMX DSL Flatrate für nur 16,99 Euro/mtl.!
http://portal.gmx.net/de/go/dsl02
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Idea for a very simple GUI llibrary

2009-11-23T02:37:30Z

> Nice idea. I will try it if you write runGUI :-)

Sure, just let me know :)

If this is to be done, I think it's better that the person writing
the Haskell code do not write runGUI, so the implementation
details wouln't discourage ideas that make life easier for users.

> This is an imperative style library. For more Haskellian GUI
> library ideas, see Fruit (http://www.haskell.org/fruit/) and TVs
> (http://www.haskell.org/haskellwiki/TV). They may not pass the
> "builds" constraint :-P

I do remember looking at TVs and also Fudgets as sugested by
Keith.

But there's a unfilled hole for a library that's conceptually
simple. I believe that with the library I described users
(begginers in Haskell?) could even use QuickCheck and HUnit with
their GUI code.

Thanks for your comments,
Maurício

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Some useful TH templates

2009-11-23T01:50:52Z

Hi Yair,

> I wrote some Template Haskell templates that I think may be of use to others.
>
> The first generates "in" and "with" functions for newtypes.

This looks very nice. Have you thought about putting this code in to
the Derive package? (http://community.haskell.org/~ndm/derive, and
also on Hackage).

It provides a set of derivations, which can be called from Template
Haskell, and a command line program for applying them. By putting them
inside Derive you'll get lots of nice things for free, and it will be
easier for people to use your code.

Thanks, Neil
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Idea for a very simple GUI llibrary

2009-11-23T00:50:28Z

Has there been "real world" adoption of any of these, in the shape of
a moderately complex end-user application that is not just a library
demo?

martin

On Mon, Nov 23, 2009 at 8:48 AM, Keith Holman <holmak@...> wrote:

> You should also check out Fudgets and "Tangible Functional
> Programming." Fudgets is a really old Haskell UI library concept;
> Tangible FP is a recent Google talk about a UI library inspired by
> Haskell types.
>
> 2009/11/22 Luke Palmer <lrpalmer@...>:
>> Nice idea. I will try it if you write runGUI :-)
>>
>> This is an imperative style library. For more Haskellian GUI library
>> ideas, see Fruit (http://www.haskell.org/fruit/) and TVs
>> (http://www.haskell.org/haskellwiki/TV). They may not pass the
>> "builds" constraint :-P
>>
>> Luke
>>
>> 2009/11/22 Maurício CA <mauricio.antunes@...>:
>>> Hi,
>>>
>>> Here is a sketch for a library with these properties:
>>>
>>> -> Easy to test. All Haskell code can be tested in a text
>>> terminal. Also, testing code that uses the library can also be
>>> done without using a GUI.
>>>
>>> -> Extremely easy to document and use.
>>>
>>> -> Not even close to Gtk2hs power, but enough for small
>>> applications.
>>>
>>> -> Could be the first GUI to build on hackage :)
>>>
>>> What we need is:
>>>
>>> -> MyState. A user suplied type for application state.
>>>
>>> -> WidId. A user suplied type for widget identifiers.
>>>
>>> -> Gui wi. A type capable of describing an interface with all of
>>> its state. It's an instance of Eq.
>>>
>>> -> Event wi. A type for events.
>>>
>>> -> Prop. A type for properties than can related to a WidId.
>>>
>>> Running an application would be like this:
>>>
>>> main = runGUI
>>> initState -- An initial MyState.
>>> event -- :: MyState -> DiffTime -> Event WidId -> MyState
>>> props -- :: WidId -> [Prop]
>>> action -- :: MyState -> DiffTime -> IO (Maybe (MyState,Gui
>>> WidId))
>>> timeout -- :: DiffTime
>>>
>>> DiffTime parameters for callbacks are always the time elapsed
>>> since application started.
>>>
>>> From initState and event, the implementation of runGUI can save a
>>> state that optionally changes with time.
>>>
>>> From props, it can get details on what to present in widgets
>>> associated with a WidId (selected state, picture to draw etc.).
>>>
>>> action presents a chance for using IO, and optionally change state
>>> and GUI description.
>>>
>>> timeout is the maximum time runGUI implementation is allowed to
>>> wait between calls to action.
>>>
>>> Examples for those types:
>>>
>>> newtype MyState = {
>>> lastUpdate :: DiffTime,
>>> builtGui :: Bool,
>>> earthCoordinates :: (Double,Double),
>>> map :: SVG,
>>> ...
>>> }
>>>
>>> data WidId = XCoord | YCoord | MapWindow | ReloadButton ...
>>>
>>> data Gui widid = TitleWindow (Gui widid)
>>> | Tabs [(String,Gui widid)]
>>> | PressButton String widid
>>> | Selection [String] widid
>>> | ...
>>> deriving Eq
>>> {-
>>> Eq is needed by runGUI to detect if GUI has
>>> changed after the last call to action.
>>> -}
>>>
>>> data Event widid = ButtonPressed widid
>>> | FileSelected String widid
>>> | OptionSelected String widid
>>> | ...
>>>
>>> data Prop widid = Active Bool
>>> | Text String
>>> | Draw SVG
>>> | ...
>>>
>>> I believe this can represent most kinds of simple applications,
>>> and be efficient enough for practical use.
>>>
>>> It's interesting that all of this can be designed, implemented and
>>> tested independent of runGUI implementation. Actually, if you want
>>> a pet project and want to write and design the Haskell part, I may
>>> probably be able to write runGUI for you :)
>>>
>>> Best,
>>> Maurício
>>>
>>> _______________________________________________
>>> Haskell-Cafe mailing list
>>> Haskell-Cafe@...
>>> http://www.haskell.org/mailman/listinfo/haskell-cafe
>>>
>> _______________________________________________
>> Haskell-Cafe mailing list
>> Haskell-Cafe@...
>> http://www.haskell.org/mailman/listinfo/haskell-cafe
>>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe@...
> http://www.haskell.org/mailman/listinfo/haskell-cafe
>

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Idea for a very simple GUI llibrary

2009-11-22T19:18:12Z

You should also check out Fudgets and "Tangible Functional
Programming." Fudgets is a really old Haskell UI library concept;
Tangible FP is a recent Google talk about a UI library inspired by
Haskell types.

2009/11/22 Luke Palmer <lrpalmer@...>:

> Nice idea. I will try it if you write runGUI :-)
>
> This is an imperative style library. For more Haskellian GUI library
> ideas, see Fruit (http://www.haskell.org/fruit/) and TVs
> (http://www.haskell.org/haskellwiki/TV). They may not pass the
> "builds" constraint :-P
>
> Luke
>
> 2009/11/22 Maurício CA <mauricio.antunes@...>:
>> Hi,
>>
>> Here is a sketch for a library with these properties:
>>
>> -> Easy to test. All Haskell code can be tested in a text
>> terminal. Also, testing code that uses the library can also be
>> done without using a GUI.
>>
>> -> Extremely easy to document and use.
>>
>> -> Not even close to Gtk2hs power, but enough for small
>> applications.
>>
>> -> Could be the first GUI to build on hackage :)
>>
>> What we need is:
>>
>> -> MyState. A user suplied type for application state.
>>
>> -> WidId. A user suplied type for widget identifiers.
>>
>> -> Gui wi. A type capable of describing an interface with all of
>> its state. It's an instance of Eq.
>>
>> -> Event wi. A type for events.
>>
>> -> Prop. A type for properties than can related to a WidId.
>>
>> Running an application would be like this:
>>
>> main = runGUI
>> initState -- An initial MyState.
>> event -- :: MyState -> DiffTime -> Event WidId -> MyState
>> props -- :: WidId -> [Prop]
>> action -- :: MyState -> DiffTime -> IO (Maybe (MyState,Gui
>> WidId))
>> timeout -- :: DiffTime
>>
>> DiffTime parameters for callbacks are always the time elapsed
>> since application started.
>>
>> From initState and event, the implementation of runGUI can save a
>> state that optionally changes with time.
>>
>> From props, it can get details on what to present in widgets
>> associated with a WidId (selected state, picture to draw etc.).
>>
>> action presents a chance for using IO, and optionally change state
>> and GUI description.
>>
>> timeout is the maximum time runGUI implementation is allowed to
>> wait between calls to action.
>>
>> Examples for those types:
>>
>> newtype MyState = {
>> lastUpdate :: DiffTime,
>> builtGui :: Bool,
>> earthCoordinates :: (Double,Double),
>> map :: SVG,
>> ...
>> }
>>
>> data WidId = XCoord | YCoord | MapWindow | ReloadButton ...
>>
>> data Gui widid = TitleWindow (Gui widid)
>> | Tabs [(String,Gui widid)]
>> | PressButton String widid
>> | Selection [String] widid
>> | ...
>> deriving Eq
>> {-
>> Eq is needed by runGUI to detect if GUI has
>> changed after the last call to action.
>> -}
>>
>> data Event widid = ButtonPressed widid
>> | FileSelected String widid
>> | OptionSelected String widid
>> | ...
>>
>> data Prop widid = Active Bool
>> | Text String
>> | Draw SVG
>> | ...
>>
>> I believe this can represent most kinds of simple applications,
>> and be efficient enough for practical use.
>>
>> It's interesting that all of this can be designed, implemented and
>> tested independent of runGUI implementation. Actually, if you want
>> a pet project and want to write and design the Haskell part, I may
>> probably be able to write runGUI for you :)
>>
>> Best,
>> Maurício
>>
>> _______________________________________________
>> Haskell-Cafe mailing list
>> Haskell-Cafe@...
>> http://www.haskell.org/mailman/listinfo/haskell-cafe
>>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe@...
> http://www.haskell.org/mailman/listinfo/haskell-cafe
>

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: What's wrong with code.haskell.org ?

2009-11-22T17:12:49Z

>> code.h.o and community.h.o have rather flaky hosting, and have been
>> going down often recently.

Yes, we know that code/community.h.o can be somewhat flaky. However,
it does claim to have been up continuously for the last 79 days. I
suspect it is that (a) the service daemons occasionally die for no
obvious reason, or (b) some users are running processes that sometimes
hog the CPU, memory, or network bandwidth.

>> They have hardware problems - dying HDs and such.

No, that is a different machine: hackage/darcs.h.o. A new server is
on order, but we do not yet have a delivery date.

Regards,
Malcolm

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Idea for a very simple GUI llibrary

2009-11-22T16:59:10Z

Nice idea. I will try it if you write runGUI :-)

This is an imperative style library. For more Haskellian GUI library
ideas, see Fruit (http://www.haskell.org/fruit/) and TVs
(http://www.haskell.org/haskellwiki/TV). They may not pass the
"builds" constraint :-P

Luke

2009/11/22 Maurício CA <mauricio.antunes@...>:

> Hi,
>
> Here is a sketch for a library with these properties:
>
> -> Easy to test. All Haskell code can be tested in a text
> terminal. Also, testing code that uses the library can also be
> done without using a GUI.
>
> -> Extremely easy to document and use.
>
> -> Not even close to Gtk2hs power, but enough for small
> applications.
>
> -> Could be the first GUI to build on hackage :)
>
> What we need is:
>
> -> MyState. A user suplied type for application state.
>
> -> WidId. A user suplied type for widget identifiers.
>
> -> Gui wi. A type capable of describing an interface with all of
> its state. It's an instance of Eq.
>
> -> Event wi. A type for events.
>
> -> Prop. A type for properties than can related to a WidId.
>
> Running an application would be like this:
>
> main = runGUI
> initState -- An initial MyState.
> event -- :: MyState -> DiffTime -> Event WidId -> MyState
> props -- :: WidId -> [Prop]
> action -- :: MyState -> DiffTime -> IO (Maybe (MyState,Gui
> WidId))
> timeout -- :: DiffTime
>
> DiffTime parameters for callbacks are always the time elapsed
> since application started.
>
> From initState and event, the implementation of runGUI can save a
> state that optionally changes with time.
>
> From props, it can get details on what to present in widgets
> associated with a WidId (selected state, picture to draw etc.).
>
> action presents a chance for using IO, and optionally change state
> and GUI description.
>
> timeout is the maximum time runGUI implementation is allowed to
> wait between calls to action.
>
> Examples for those types:
>
> newtype MyState = {
> lastUpdate :: DiffTime,
> builtGui :: Bool,
> earthCoordinates :: (Double,Double),
> map :: SVG,
> ...
> }
>
> data WidId = XCoord | YCoord | MapWindow | ReloadButton ...
>
> data Gui widid = TitleWindow (Gui widid)
> | Tabs [(String,Gui widid)]
> | PressButton String widid
> | Selection [String] widid
> | ...
> deriving Eq
> {-
> Eq is needed by runGUI to detect if GUI has
> changed after the last call to action.
> -}
>
> data Event widid = ButtonPressed widid
> | FileSelected String widid
> | OptionSelected String widid
> | ...
>
> data Prop widid = Active Bool
> | Text String
> | Draw SVG
> | ...
>
> I believe this can represent most kinds of simple applications,
> and be efficient enough for practical use.
>
> It's interesting that all of this can be designed, implemented and
> tested independent of runGUI implementation. Actually, if you want
> a pet project and want to write and design the Haskell part, I may
> probably be able to write runGUI for you :)
>
> Best,
> Maurício
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe@...
> http://www.haskell.org/mailman/listinfo/haskell-cafe
>

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

some help debugging ghci build on powerpc

2009-11-22T16:22:43Z

I'm trying to build ghc so that ghci will be included under linux power-pc.

The build dies here:

~/ghc6-6.10.4/rts$ /home/briand/ghc6-6.10.4/ghc/stage1-inplace/ghc -optc-O -optc-Wall -optc-W -optc-Wstrict-prototypes -optc-Wmissing-prototypes -optc-Wmissing-declarations -optc-Winline -optc-Waggregate-return -optc-I../includes -optc-I. -optc-Iparallel -optc-Ism -optc-DCOMPILING_RTS -optc-fomit-frame-pointer -optc-DNOSMP -optc-I../gmp/gmpbuild -optc-I../libffi/build/include -optc-fno-strict-aliasing -optc-w -H32m -O -optc-O2 -I../includes -I. -Iparallel -Ism -DCOMPILING_RTS -package-name rts -optc-DNOSMP -static -I../gmp/gmpbuild -I../libffi/build/include -I. -dcmm-lint -c StgCRun.c -o StgCRun.o
/tmp/ghc21791_0/ghc21791_0.s: Assembler messages:

/tmp/ghc21791_0/ghc21791_0.s:15:0:
Error: junk at end of line, first unrecognized character is `@'

/tmp/ghc21791_0/ghc21791_0.s:19:0:
Error: junk at end of line, first unrecognized character is `@'

That relates to this line(s):

StgRun:
.quad .StgRun,.TOC.@tocbase,0
.size StgRun,24
.globl StgReturn

It _looks_ like it's grabbing the right stuff from StgCRun.c,
i.e. it's really trying to compile for power pc.

I was hoping that this was a bit of cruft and someone could direct me
to a fix for the assembly portion so I can get a little farther along.

Thanks,

Brian

P.S. Or somebody could also tell me if trying to have the latest ghc
working on power pc is even worth working on.

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Idea for a very simple GUI llibrary

2009-11-22T16:19:00Z

Hi,

Here is a sketch for a library with these properties:

-> Easy to test. All Haskell code can be tested in a text
terminal. Also, testing code that uses the library can also be
done without using a GUI.

-> Extremely easy to document and use.

-> Not even close to Gtk2hs power, but enough for small
applications.

-> Could be the first GUI to build on hackage :)

What we need is:

-> MyState. A user suplied type for application state.

-> WidId. A user suplied type for widget identifiers.

-> Gui wi. A type capable of describing an interface with all of
its state. It's an instance of Eq.

-> Event wi. A type for events.

-> Prop. A type for properties than can related to a WidId.

Running an application would be like this:

main = runGUI
initState -- An initial MyState.
event -- :: MyState -> DiffTime -> Event WidId -> MyState
props -- :: WidId -> [Prop]
action -- :: MyState -> DiffTime -> IO (Maybe (MyState,Gui WidId))
timeout -- :: DiffTime

DiffTime parameters for callbacks are always the time elapsed
since application started.

From initState and event, the implementation of runGUI can save a
state that optionally changes with time.

From props, it can get details on what to present in widgets
associated with a WidId (selected state, picture to draw etc.).

action presents a chance for using IO, and optionally change state
and GUI description.

timeout is the maximum time runGUI implementation is allowed to
wait between calls to action.

Examples for those types:

newtype MyState = {
lastUpdate :: DiffTime,
builtGui :: Bool,
earthCoordinates :: (Double,Double),
map :: SVG,
...
}

data WidId = XCoord | YCoord | MapWindow | ReloadButton ...

data Gui widid = TitleWindow (Gui widid)
| Tabs [(String,Gui widid)]
| PressButton String widid
| Selection [String] widid
| ...
deriving Eq
{-
Eq is needed by runGUI to detect if GUI has
changed after the last call to action.
-}

data Event widid = ButtonPressed widid
| FileSelected String widid
| OptionSelected String widid
| ...

data Prop widid = Active Bool
| Text String
| Draw SVG
| ...

I believe this can represent most kinds of simple applications,
and be efficient enough for practical use.

It's interesting that all of this can be designed, implemented and
tested independent of runGUI implementation. Actually, if you want
a pet project and want to write and design the Haskell part, I may
probably be able to write runGUI for you :)

Best,
Maurício

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Graph drawing library for Haskell

2009-11-22T12:21:01Z

On Fri, Nov 20, 2009 at 9:36 PM, Victor Mateus Oliveira
<rhapsodyv@...> wrote:
> I'm looking for something more integrated with a gui library. The
> jgraph integrates with swing, so you can move, create, delete, have
> popup menus, select nodes, and so on.
>
> I haven't found yet.. If there isn't, I thinking in create one lib
> with wxHaskell using wxDC... But by now, I really prefer to use one
> existing library.

Blobs [http://www.cs.york.ac.uk/fp/darcs/Blobs/] might be a better
starting point than implementing from scratch on wxdc

martin
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Existential type question

2009-11-22T10:59:58Z

Am Sonntag 22 November 2009 19:24:48 schrieb Michael Snoyman:

> Hi all,
>
> I've come across some code I just can't figure out how to write
> appropriately. Below is a silly example that demonstrates what I'm trying
> to do. I don't really have the appropriate vocabulary to describe the
> issue, so I'll let the code speak for itself. In particular, I'm trying to
> understand what the correct type signatures for unwrapMyData and bin should
> be.
>
> Thanks,
> Michael
>
> ---
>
> {-# LANGUAGE MultiParamTypeClasses #-}
> {-# LANGUAGE ExistentialQuantification #-}
> {-# LANGUAGE FlexibleInstances #-}
> {-# LANGUAGE FlexibleContexts #-}
> class Monad m => MonadFoo x m where
> foo :: x -> m a
>
> data MyData a = forall i. Integral i => MyLeft i
>
> | MyRight a
>
> instance Monad MyData where
> return = MyRight
> (MyLeft i) >>= _ = MyLeft i
> (MyRight x) >>= f = f x
> instance Integral i => MonadFoo i MyData where
> foo = MyLeft
>
> bar :: MonadFoo Int m => Int -> m String
> bar 0 = return "zero"
> bar i = foo i
>
> baz :: String -> MyData String
> baz "zero" = MyRight "Zero"
> baz _ = MyLeft (-1 :: Integer)
>
> --This works: unwrapMyData (MyLeft i) = foo (fromIntegral i :: Integer)
> unwrapMyData (MyLeft i) = foo i -- This is what I'd like to work

I don't see how it could (except perhaps...). The type would be

unwrapMyData ::(forall i. (Integral i => MonadFoo i m)) => MyData a -> m a

- the MyData a -> m a part is obviously necessary, but MyLeft can wrap *any* integral
type, so you indeed need instances MonadFoo i m for *every* integral type. You can't
specify that condition in Haskell.

You could try

instance MonadFoo a [] where
foo _ = []

instance MonadFoo a Maybe where
foo _ = Nothing

(more general,
instance (MonadPlus m) => MonadFoo a m where
foo _ = mzero
)

unwrapMyData :: (MonadPlus m) => MyData a -> m a

that might be got to work, but it's very likely not even remotely what you want
(unwrapMyData :: MyData a -> [a] would work with the above instance, I think, but it would
be rather pointless).

> unwrapMyData (MyRight a) = return a
>
> bin i = do
> a <- bar i

You must specify which instance of MonadFoo to use here, otherwise it is impossible to
determine which instance of bar to use, and the type of bin is irredeemably ambiguous
(show . read).

> b <- unwrapMyData $ baz a
> return $ b ++ "!!!"
>
> instance Show a => MonadFoo a IO where
> foo = fail . show -- I know, it's horrible...
>
> main = do
> res <- bin 0
> putStrLn res

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Existential type question

2009-11-22T10:24:48Z

Hi all,

I've come across some code I just can't figure out how to write appropriately. Below is a silly example that demonstrates what I'm trying to do. I don't really have the appropriate vocabulary to describe the issue, so I'll let the code speak for itself. In particular, I'm trying to understand what the correct type signatures for unwrapMyData and bin should be.

Thanks,
Michael

---

{-# LANGUAGE MultiParamTypeClasses #-}
{-# LANGUAGE ExistentialQuantification #-}
{-# LANGUAGE FlexibleInstances #-}
{-# LANGUAGE FlexibleContexts #-}
class Monad m => MonadFoo x m where
    foo :: x -> m a

data MyData a = forall i. Integral i => MyLeft i
              | MyRight a

instance Monad MyData where
    return = MyRight
    (MyLeft i) >>= _ = MyLeft i
    (MyRight x) >>= f = f x
instance Integral i => MonadFoo i MyData where
    foo = MyLeft

bar :: MonadFoo Int m => Int -> m String
bar 0 = return "zero"
bar i = foo i

baz :: String -> MyData String
baz "zero" = MyRight "Zero"
baz _ = MyLeft (-1 :: Integer)

--This works: unwrapMyData (MyLeft i) = foo (fromIntegral i :: Integer)
unwrapMyData (MyLeft i) = foo i -- This is what I'd like to work
unwrapMyData (MyRight a) = return a

bin i = do
    a <- bar i
    b <- unwrapMyData $ baz a
    return $ b ++ "!!!"

instance Show a => MonadFoo a IO where
    foo = fail . show -- I know, it's horrible...

main = do
    res <- bin 0
    putStrLn res

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: Re: Wiki software?

2009-11-22T09:56:13Z

On Sat, Nov 21, 2009 at 3:09 PM, Gwern Branwen <gwern0@...> wrote:
> This doesn't surprise me; but how much slower are we talking?
>
> If it's not at the point that a browser of a Gitit wiki could notice
> the difference, then it seems to me that the dep ought to be loosened:
> the parsec/quickcheck/base diamond dependency problem is one of the
> worst ones a user can run into, the hardest to resolve, and one that
> can arise in the course of ordinary safe use of Haskell.

Running 'pandoc --strict' over the Markdown readme.text takes:

~0.09s with pandoc built against parsec-2
~0.19s with pandoc built against parsec-3

on my machine.

I have a branch of parsec-3 which seems to brings us back to parsec-2
numbers, but also fails the rst-reader test-case in the pandoc testing
suite:

http://community.haskell.org/~aslatter/code/parsec/cps

Antoine
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: What's wrong with code.haskell.org ?

2009-11-22T09:11:42Z

Svein Ove Aas <svein.ove@...> writes:

> On Sun, Nov 22, 2009 at 4:39 AM, John Millikin <jmillikin@...> wrote:
>> code.h.o and community.h.o have rather flaky hosting, and have been
>> going down often recently. The only solution seems to be waiting until
>> the admins notice.
>>
> They have hardware problems - dying HDs and such.
>
> IIRC there were plans to replace the problematic hardware, but I don't
> know how far that has gotten. Maybe we should have a fundraiser for
> the purpose?

Hope they can fix this problem completely, then running stable...

-- Andy

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: haskell code from hi

2009-11-22T08:41:36Z

I really get the point, and what I should have been doing. Anyways shit happens!

Thanks for the answers though.

2009/11/22 Svein Ove Aas <svein.ove@...>

On Sun, Nov 22, 2009 at 1:04 AM, Uwe Hollerbach <uhollerbach@...> wrote:
> Ouch... my condolences, but I think you're screwed. I think the .hi
> files are purely interface info, and the .o files have all the info on
> what to actually do (and getting to .hs files from .hi+.o is gonna be
> like going from sausage to pig, in any case). If you haven't messed
> with the disk, I think your best bet might be to try and undelete
> files. That might be as messy as looking at the raw disk image and
> trying to recover disk sectors, or possibly there are still entire
> files there that are just not referenced by directory entries. Either
> (or any) way, it's a bit chancy...
>

And this why you always, *always* use a revision-control system or at
least good backups.

.hi files actually do contain some code, specifically what ghc decides
can be inlined (see ghc --show-iface), but it's not the sort of code
you'd get any use from. Have a look, you'll see.

--
Svein Ove Aas

--
Ozgur Akgun

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: haskell code from hi

2009-11-22T08:14:26Z

On Sun, Nov 22, 2009 at 1:04 AM, Uwe Hollerbach <uhollerbach@...> wrote:

> Ouch... my condolences, but I think you're screwed. I think the .hi
> files are purely interface info, and the .o files have all the info on
> what to actually do (and getting to .hs files from .hi+.o is gonna be
> like going from sausage to pig, in any case). If you haven't messed
> with the disk, I think your best bet might be to try and undelete
> files. That might be as messy as looking at the raw disk image and
> trying to recover disk sectors, or possibly there are still entire
> files there that are just not referenced by directory entries. Either
> (or any) way, it's a bit chancy...
>

And this why you always, *always* use a revision-control system or at
least good backups.

.hi files actually do contain some code, specifically what ghc decides
can be inlined (see ghc --show-iface), but it's not the sort of code
you'd get any use from. Have a look, you'll see.

--
Svein Ove Aas
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: What's wrong with code.haskell.org ?

2009-11-22T08:05:22Z

On Sun, Nov 22, 2009 at 4:39 AM, John Millikin <jmillikin@...> wrote:
> code.h.o and community.h.o have rather flaky hosting, and have been
> going down often recently. The only solution seems to be waiting until
> the admins notice.
>
They have hardware problems - dying HDs and such.

IIRC there were plans to replace the problematic hardware, but I don't
know how far that has gotten. Maybe we should have a fundraiser for
the purpose?

--
Svein Ove Aas
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

ANN: GPipe-1.1.0 with greatly improved performance

2009-11-22T04:22:59Z

Hi,

GPipe 1.1.0 is now available on Hackage. This update includes the following:

* BIG performance increase:

The previous shader cache keys grew exponential in size in relation to the GPipe program. This is now fixed with a completely new shader generator so the cache key sizes are linear relative to the program. In addition, the cache now uses patricia trees instead of an ordinary map, which will be a big performance win in programs with many shaders. In the simple example on http://www.haskell.org/haskellwiki/GPipe I got an performance increase of factor 2.7(!), and expect it to be even larger for more complex programs.

* Other minor fixes:

Fixed so min, max, abs and sign no longer works on Vertex Int and Fragment Int types, since its not supported in GLSL 1.20.

Reexport of Vec-Boolean instances so you dont need to import that package on its own any more.

Added IfB instances for the associated type Color.

Cheers
Tobias

Hitta kärleken! Klicka här Hitta en dator som passar dig!
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: How to understand such a newtype ?

2009-11-22T01:50:26Z

On Sun, Nov 22, 2009 at 3:47 AM, Felipe Lessa <felipe.lessa@...> wrote:
> ...
> Well, “ReaderT XConf (StateT XState IO) a” is *the* type :).
> It's a monad that is a Reader of XConf and has a State of XState.

... and also wraps a monad to allow IO access inside the X monad.

Cristiano
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: How to understand such a newtype ?

2009-11-21T20:03:46Z

thanks your quick answer first !

> :m + Control.Monad.Reader
> :t ReaderT
ReaderT :: (r -> m a) -> ReaderT r m a

> :m + Control.Monad.State
> :t StateT
StateT :: (s -> m (a, s)) -> StateT s m a

data XConf = XConf {...}
data XState = XState {...}

Both ReaderT and StateT needs a function as its 1st parameter, then in `ReaderT XConf (StateT XState IO) a`, which parts is the function parameter for ReaderT and StateT respectively ?

Felipe Lessa wrote:

On Sat, Nov 21, 2009 at 06:20:40PM -0800, zaxis wrote:
> newtype X a = X (ReaderT XConf (StateT XState IO) a)
> #ifndef __HADDOCK__
> deriving (Functor, Monad, MonadIO, MonadState XState, MonadReader XConf,
> Typeable)
> #endif
>
> In `X (ReaderT XConf (StateT XState IO) a)`, X is a type constructor, how to
> understand `(ReaderT XConf (StateT XState IO) a)` ?

Well, “ReaderT XConf (StateT XState IO) a” is *the* type :).
It's a monad that is a Reader of XConf and has a State of XState.
This means you can use, for example,

ask :: X XConf

and

get :: X XState

> And why use `#ifndef __HADDOCK__` ?

Because Haddock used to have difficulties in processing some
directives, like that “deriving (..., MonadState XState, ...)”
which is part of the GeneralizedNewtypeDeriving extension.

HTH,

--
Felipe.
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

fac n = foldr (*) 1 [1..n]

Re: What's wrong with code.haskell.org ?

2009-11-21T19:39:08Z

code.h.o and community.h.o have rather flaky hosting, and have been
going down often recently. The only solution seems to be waiting until
the admins notice.

On Sat, Nov 21, 2009 at 18:31, Andy Stewart <lazycat.manatee@...> wrote:

> Hi all,
>
> I can't use `darcs get --partial http://code.haskell.org/gtk2hs/` to get
> source code, and can't push patch.
>
> What's wrong with code.haskell.org?
> Rest for weekend? :)
>
> -- Andy
>
> _______________________________________________
> Haskell-Cafe mailing list
> Haskell-Cafe@...
> http://www.haskell.org/mailman/listinfo/haskell-cafe
>

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: How to understand such a newtype ?

2009-11-21T18:47:10Z

On Sat, Nov 21, 2009 at 06:20:40PM -0800, zaxis wrote:
> newtype X a = X (ReaderT XConf (StateT XState IO) a)
> #ifndef __HADDOCK__
> deriving (Functor, Monad, MonadIO, MonadState XState, MonadReader XConf,
> Typeable)
> #endif
>
> In `X (ReaderT XConf (StateT XState IO) a)`, X is a type constructor, how to
> understand `(ReaderT XConf (StateT XState IO) a)` ?

Well, “ReaderT XConf (StateT XState IO) a” is *the* type :).
It's a monad that is a Reader of XConf and has a State of XState.
This means you can use, for example,

ask :: X XConf

and

get :: X XState

> And why use `#ifndef __HADDOCK__` ?

Because Haddock used to have difficulties in processing some
directives, like that “deriving (..., MonadState XState, ...)”
which is part of the GeneralizedNewtypeDeriving extension.

HTH,

--
Felipe.
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

What's wrong with code.haskell.org ?

2009-11-21T18:31:02Z

Hi all,

I can't use `darcs get --partial http://code.haskell.org/gtk2hs/` to get
source code, and can't push patch.

What's wrong with code.haskell.org?
Rest for weekend? :)

-- Andy

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@...
http://www.haskell.org/mailman/listinfo/haskell-cafe

How to understand such a newtype ?

2009-11-21T18:20:38Z

newtype X a = X (ReaderT XConf (StateT XState IO) a)
#ifndef __HADDOCK__
deriving (Functor, Monad, MonadIO, MonadState XState, MonadReader XConf, Typeable)
#endif

In `X (ReaderT XConf (StateT XState IO) a)`, X is a type constructor, how to understand `(ReaderT XConf (StateT XState IO) a)` ? And why use `#ifndef __HADDOCK__` ?

Sincerely!

fac n = foldr (*) 1 [1..n]