- Haskell-Cafe - Haskell.org

distributed systems reading group
by Jason Dusek 24 Oct '08

24 Oct '08

I'm looking for good reading group for distributed systems. I've worked through a number of papers on my own and would like to work out some ideas with other folks who have an interest. I've read a little bit of Lynch, a bit more of Lamport and some of Chandra's work, as well as numerous slides, &c. * The impossibility results are sobering, but always leave a way out -- they make some assumption that one can dispense with in practice (for example, that processes start with different values). * The possibility results generally make untenable assumptions -- for example, that there is no link failure. I appreciate that this is not a matter of direct interest to the list, but I expect to find interested people here. -- _jsn

1 0

Sorry, but I'm in the mood (again)
by Achim Schneider 24 Oct '08

24 Oct '08

http://www.haskell.org/haskellwiki/?title=Special:Categories What are the mathematical properties of those Categories? Do they share common axioms? -- (c) this sig last receiving data processing entity. Inspect headers for copyright history. All rights reserved. Copying, hiring, renting, performance and/or quoting of this signature prohibited.

2 2

Re: universal algebra "support" in Haskell?
by Galchin, Vasili 24 Oct '08

24 Oct '08

Hi Russell and others, Please "cc" me on all responses to my post 'universal algebra "support" in Haskell'. I am interested in algebras and non-algebras like pre-order, post-order, etc. support in Haskell! Thanks, Vasili

1 0

A heretic question
by Achim Schneider 24 Oct '08

24 Oct '08

What kind of things, barring coding on Haskell-less platforms and library interfaces would you choose to do in C++? I'm asking 'cos I'm learning C++ and can't get the proper motivation to do any program I can think of in it: If I need abstraction, I'm thinking Haskell or Scheme, and if I'm thinking performance, C itself more than suffices. Plus template programming makes me shudder because of its atrocities against clear and straightforward FP, but that's a different matter. Coming to think of it, a compiler from a clean syntax to C++ templates sounds like a fun project... which I'd do in Haskell. -- (c) this sig last receiving data processing entity. Inspect headers for copyright history. All rights reserved. Copying, hiring, renting, performance and/or quoting of this signature prohibited.

17 30

Linking and unsafePerformIO
by Mauricio 24 Oct '08

24 Oct '08

Hi, If I have a Haskell wrapper (with unsafe...) over a function that's never going to return different values and is always side-effect free, but can change depending on compile time options of its library; my program is running, and then the version of my library is updated by my distribution smart instalation system, which does update versions of libraries in use; is it possible that I get a wrong behavior of my program? I do not understand enough about package management to understand how running programs or libraries are updated, and less about how linking works between Haskelll and libraries on other languages, so I don't know if my program is guaranteed to stay with a single version of a library for each run. (Sure this is a weird situation, but I do like to think about worst cases.) Thanks, Mauríco

13 20

Pandoc questions
by Andrew Coppin 24 Oct '08

24 Oct '08

There doesn't seem to be any option to make Pandoc produce actual MathML output. Is there a reason for this? (The only option I can see is to spit out raw LaTeX plus a 70KB JavaScript program to transform this into MathML at the client end --- which seems a little silly to me. There's also no way to style the raw LaTeX differently in case JavaScript is unavailable.) Also, while Markdown *almost* does what I want, there are a few small constructs it doesn't have. For example, I'd like to have some way to denote a "term" the first time I use it. I could just use italics, but I'd prefer some way to visually indicate that this isn't just an emphasised word, it's a new technical term. In HTML, I'd use a style class, and in LaTeX I'd define a new command. But I can't see a way to do something that will still allow Pandoc to generate correct LaTeX *and* correct HTML from a single Markdown source... Any hints?

4 8

Re: External Sort: Sort a 10-million integer file with just 256M of ram.
by Ben 23 Oct '08

23 Oct '08

the problem with the current externalSort is that it is not recursive -- it tries to do everything in one pass. basically instead of attempting to merge all of the chunks together it should only merge chunksize chunks together, producing a new list of chunks. then those get merged, etc....there is a discussion of this in knuth i think. it shouldn't be hard to make this work, i just didn't have time when i wrote it. ben

1 0

Re: [Haskell-cafe] Fwd: enhancing type classes with properties
by Alberto G. Corona 23 Oct '08

23 Oct '08

It seems that some of the goals are not so hard. here I publised my progress. http://docs.google.com/Doc?id=dd5rm7qq_165rshp74gf&pli=1 I show how to define a Ring, such a mathematical structure in Haskell, how to instantiate the class Num as a Ring , how to (possibly in other moment of space-time) instantiate a new class as Num and how to test the axioms for the new class. All of then is something like a sophisticated "assert" mechanism, but , I think, much more flexible and elegant. 2008/10/22 Alberto G. Corona <agocorona(a)gmail.com> > I guess that the namespace thing is not necesary. Maybe it can be done in > template haskell. I never used TH however. it is a matter of inserting code > here and there and rewrite rules somewhere at compile time. It´s a nice > project. I´ll try. > > > 2008/10/22 Mitchell, Neil <neil.mitchell.2(a)credit-suisse.com> > >> Hi Alberta, >> >> It's a lot of work, but I wish you luck :-) Many of the underlying tools >> exist, but there definately needs more integration. >> >> Thanks >> >> Neil >> >> >> This material is sales and trading commentary and does not constitute >> investment research. Please follow the attached hyperlink to an important >> disclaimer >> *<www.credit-suisse.com/emea/legal>* >> >> >> ------------------------------ >> *From:* Alberto G. Corona [mailto:agocorona@gmail.com] >> *Sent:* 22 October 2008 4:23 pm >> *To:* Mitchell, Neil >> *Subject:* Re: [Haskell-cafe] Fwd: enhancing type classes with properties >> >> Hi Neil, >> >> I see the contract type mechanism and safety check techniques reflected >> in the above paper are a good step, But I think that something more >> general would be better. What I propose is to integrate directly in the >> language some concepts and developments that are already well know to >> solve some common needs that I thing can not be solved without this >> integration: >> >> To make use of: >> >> -Quickcheck style validation. By the way, Don Steward recommend to add >> quckcheck rules close to the class definitions just for better >> documentation >> >> - Implicit class properties defined everywhere in the documentation but >> impossible to reflect in the code (for example the famous monad rules: >> return x >>= f == f x etc ) >> >> - The superb ghc rewrite rule mechanism (perhaps with enhancements) >> >> - object style namespaces, depending on class names. >> >> To solve problems like >> >> - code optimization >> >> - code verification. regression tests for free!! >> >> - The need for safe overloading of methods and operators (making the >> namespaces dependent not only on module name but also in class names) . Why >> I can not overload the operator + in the context of a DSL for >> JavaScript generation if my operator does what + does? >> >> - strict and meaningful rules for class instances. >> - to make the rewrite rule mechanism visible to everyone >> >> >> >> 2008/10/22 Mitchell, Neil <neil.mitchell.2(a)credit-suisse.com> >> >>> Hi Alberto, >>> >>> Take a look at ESC/Haskell and Sound Haskell, which provide mechanisms >>> for doing some of the things you want. I don't think they integrate with >>> type classes in the way you mention, but I think that is just a question of >>> syntax. >>> >>> http://www.cl.cam.ac.uk/~nx200/ <http://www.cl.cam.ac.uk/%7Enx200/> >>> >>> Thanks >>> >>> Neil >>> >>> >>> This material is sales and trading commentary and does not constitute >>> investment research. Please follow the attached hyperlink to an important >>> disclaimer >>> *<www.credit-suisse.com/emea/legal>* >>> >>> >>> ------------------------------ >>> *From:* haskell-cafe-bounces(a)haskell.org [mailto: >>> haskell-cafe-bounces(a)haskell.org] *On Behalf Of *Alberto G. Corona >>> *Sent:* 22 October 2008 1:43 pm >>> *To:* haskell-cafe(a)haskell.org >>> *Subject:* [Haskell-cafe] Fwd: enhancing type classes with properties >>> >>> I´m just thinking aloud, but, because incorporating deeper mathematics >>> concepts has proven to be the best solution for better and more flexible >>> programming languages with fewer errors, I wonder if raising the type >>> classes incorporating axioms can solve additional problems. >>> >>> At first sight it does: >>> >>> >>> class Abelian a where >>> (+) :: a -> a -> a >>> property ((+))= a+b == b+a >>> >>> >>> >>> this permits: >>> 1- safer polimorphism: I can safely reuse the operator + if the type >>> and the property is obeyed. The lack of ability to redefine operators is a >>> problem for DSLs that must use wreid symbols combinations with unknow >>> meanings. To use common operators with fixed properties is very good. the >>> same aplies for method names. >>> >>> 2- the compiler can use the axions as rewrite rules. >>> >>> 3- in debugging mode, it is possible to verify the axiom for each >>> value a generated during execution. Thus, a generator is not needed like >>> in quickcheck. The logic to quickcheck can be incorporated in the debugging >>> executable. >>> >>> 3 guaranties that 1 and 2 are safe. >>> >>> >>> >>> >>> >>> a type class can express a relation between types, but it is not >>> possible to define relation between relations. >>> >>> ============================================================================== >>> Please access the attached hyperlink for an important electronic communications disclaimer: >>> http://www.credit-suisse.com/legal/en/disclaimer_email_ib.html >>> ============================================================================== >>> >>> >> ============================================================================== >> Please access the attached hyperlink for an important electronic communications disclaimer: >> http://www.credit-suisse.com/legal/en/disclaimer_email_ib.html >> ============================================================================== >> >> >

1 1

Re: [Haskell-cafe] Crash!
by Andrew Coppin 23 Oct '08

23 Oct '08

D.V. wrote: > On Thu, Oct 23, 2008 at 10:30 PM, Andrew Coppin > <andrewcoppin(a)btinternet.com> wrote: > >> Really? Well I've never ever seen GNOME crash. Hell, I run a GNOME desktop >> on a 0.01 GHz Amiga 1200 and it works! (Sure, it's slower than MS Vista on >> > > Now that's cool. I wish my Amiga still worked. > Hint: If you ever get it to work again... don't install Debian. Or rather, don't install Debian if you enjoy programs that open on the same calendar day as you click the icon. o_O It's really neat that you can make Linux run on anything, but that doesn't necessarily mean you _should_...

1 0

Re: [Haskell-cafe] Detecting unused read handles? (was: File handles and pipes)
by Thomas Hartman 23 Oct '08

23 Oct '08

>> Specifically I'm trying to concatenate the output of two system calls >> into the input of a third. The following code does not get the job Can the left-fold enumerator described by oleg in http://okmij.org/ftp/Haskell/fold-stream.lhs and with more references at http://okmij.org/ftp/Haskell/ be adapted for reading the output from arbitrary system calls? I.e., to deal with pipes? My intuition says yes but I'm still trying to wrap my head around the whole left fold enumerator approach to io. The text around "We can now test reading from two streams with interleaving:" in the first reference seemed particularly promising to me. > > > > > 2008/10/19 Stephen Hicks <sdh33(a)cornell.edu>: >> On Sun, Oct 19, 2008 at 1:44 AM, Brandon S. Allbery KF8NH >> <allbery(a)ece.cmu.edu> wrote: >>> On 2008 Oct 19, at 1:37, Stephen Hicks wrote: >>>> >>>> I'm trying to understand how to get pipes working in Haskell, in >>>> particular with the revamped System.Process (though I've tried similar >>>> experiments with System.Posix.IO and Control.Concurrent). >>>> Specifically I'm trying to concatenate the output of two system calls >>>> into the input of a third. The following code does not get the job >>>> done: >>> >>> Pipes are perhaps a bit misnamed: if you want to combine the output of two >>> pipes and funnel it into a third you can't simply "plumb them together", you >>> need to provide code which reads from the output pipes and writes into the >>> input pipe. If you don't care about order, forkIO a thread for each output >>> pipe which reads from the output pipe it's passed and writes to the input >>> pipe. If order is significant, use mapM/forM to run the output-to-input >>> code on each handle in turn. >> >> Thanks a lot - that seems to work very well, and even scales to large >> amounts of data nicely (and quickly, with Lazy ByteStrings). >> >> I've got one more question now. Suppose I want to do the same thing >> on the other side, with two processes *receiving* the data. Is there >> a way to tell whether the first process wants input, and if not, wait >> for the second process to do anything? >> >> That is, suppose I have something like >> >>> do (Just inh1, _, _, p1) <- createProcess (shell "echo 1") { std_in = CreatePipe } >>> -- wait for p1, possibly feeding it some input? >>> (Just inh2, _, _, p2) <- createProcess (shell "cat") { std_in = CreatePipe } >> >> Is there a way to figure out that the "echo 1" process never wanted >> any input, and therefore not give it any? I looked through all of >> System.IO and everything seemed to indicate that inh1 was open, even >> after the process ended. The only indication otherwise was that >> "hflush inh1" failed with "resource vanished". I guess what I'm >> asking is I want to "wait until the process p1 is waiting for input >> and/or terminates", and if it's the latter, move on to the next >> process in line, before actually doing any hPutStr into the read >> handles. Is that possible? >> >> Thanks, >> steve >> _______________________________________________ >> Haskell-Cafe mailing list >> Haskell-Cafe(a)haskell.org >> http://www.haskell.org/mailman/listinfo/haskell-cafe >> >

2 1