Well, looking at perl/python/PHP (as in LAMP), I would say the trick is to pick an application area and make it extreemely easy to use the language in that area, as well as trivial to install and maintain the language support. Of course the fact that perl/python/php all picked web-services (active web pages) just as the internet took off may have helped. So as is usual in computing - use your crystal ball to find out what the next big thing is, then make sure Haskell has a cool domain-specific-embedded language that is easy to use for that application domain ready, and then make sure everyone knows about it. Keean. Gour wrote:

azrael@demonlords.net (azrael@demonlords.net) wrote:

...

However if there is a desire to make Haskell more accessible to the majority of programmers then the first impressions that I, and those like me have is important.

Sure. I'm also pretty new to Haskell & ghc concerned to see Haskell used by more programmers.

Applications like darcs are very important enabling newcomers to see some concrete & succesful non-academia-open-source project.

Besides that, seeing the progress of projects working on the binding for gui-libraries (wxhaskell & gtk2hs) are VERY encouraging giving me hope that my investment in learning Haskell can be worthwile i.e. one is able to put together 'full-blown' application with gui interface.

However, these projects are mostly done by few individuals working hard and there is also "A Bus Hits Simon Peyton Jones" problem referred during the CUFP workshop.

Any idea how to make a (more organize) community effort to bring Haskell out?

Sincerely, Gour

Aaron Denney wrote:

On 2004-12-06, Gour wrote:

...

Any idea how to make a (more organize) community effort to bring Haskell out?

I'd rather it didn't until a few warts were fixed. OTOH, it may be too late already, barring a Haskell 2.

Does Python not have warts? Or Pearl, or Java, or C#? I don't think that a few warts prevent a language from becoming a "success". But you may be right that it is too late... Haskell is getting old! Sometimes I think that for a language to "succeed" it must do so in its infancy. Perhaps the thing to do is create a new language with a new name, but base it entirely on Haskell's semantics, then equip it with just one really good library to solve well just one important niche problem, and see what happens. If it is seen as a shiny new silver bullet in just one niche area, it might take off like a rocket. -Paul Time flies like an arrow. Fruit flies like an apple.

On Tue, 7 Dec 2004, Paul Hudak wrote:

But you may be right that it is too late... Haskell is getting old! Sometimes I think that for a language to "succeed" it must do so in its infancy.

This is also my impression.

Perhaps the thing to do is create a new language with a new name, but base it entirely on Haskell's semantics, then equip it with just one really good library to solve well just one important niche problem, and see what happens.

It seems important to me that this library has very simple functionality but which is needed every day. Word processors, spreadsheets, web browsers, cd players, games and so on are much more popular than mathematical software. :-] On the other hand Niklaus Wirth had not the success of Pascal with later languages, like Modula and Oberon. So he decided to return to the name ComponentPascal for the newest developments ...

Haskell is still a very dynamic language - there is still much room for it to grow and improve. In that sense, it is still young. Perl and Python and Ruby have hit middle age - no more growing, just more bulk accumulating around the middle. Matt On Wednesday, December 8, 2004, at 08:06 AM, Gour wrote:

Paul Hudak (paul.hudak@yale.edu) wrote:

...

Does Python not have warts? Or Pearl, or Java, or C#? I don't think that a few warts prevent a language from becoming a "success".

I agree.

...

But you may be right that it is too late... Haskell is getting old! Sometimes I think that for a language to "succeed" it must do so in its infancy.

Let's take e.g. Ruby - it's also over 10-year and is just gaining wider acceptance.

...

Perhaps the thing to do is create a new language with a new name, but base it entirely on Haskell's semantics, then equip it with just one really good library to solve well just one important niche problem, and see what happens. If it is seen as a shiny new silver bullet in just one niche area, it might take off like a rocket.

However, RAA (Ruby Application Archive) counts 1291 project in 218 categories.

Many programmers are switching from either Perl or Python (I had to unsubscribe from the mailing lists 'cause the traffic increased tremendously).

What is so special in Ruby in comparison with e.g. Perl & Python?

otoh, SF counts 91,889 projects (OK, many are dead) & 968,206 users. That is the whole army and I'm sure that filling the library-gaps, providing more documentation & printed books (what Pickaxe book did for the spreading of the Ruby, besides 20+ printed books in Japanese) covering (more) advanced features of the language (some kind of follow-up for your & Thompson's book) suited for the 'average-open-source-programmers' can bring lot of newcomers to the Haskell camp in order to solve general problems - "Haskell is a general purpose, purely functional programming language", isn't it?

What other alternative one has in the lazy-functional camp?

So, if there is already a wonderful one, why neglect the child? Let's help it to grow.

Haskell is getting old, but it is still too young .. :-)

Sincerely, Gour

-- Registered Linux User | #278493 GPG Public Key | 8C44EDCD

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

On 2004-12-06, Gour wrote:

However, these projects are mostly done by few individuals working hard and there is also "A Bus Hits Simon Peyton Jones" problem referred during the CUFP workshop.

Any idea how to make a (more organize) community effort to bring Haskell out?

I think the two main things to do that would be: 1. Write Haskell code that has a wide appeal (example: darcs) 2. Fix Haskell weaknesses Haskell weaknesses are things I've mentioned here before, and refer generally to documentation and breadth of the standard library (or ease of finding/installing additional packages). The Cabal people are working on the library problem. I am too, by writing a bunch of code and also integrating a bunch of other code. (Once Cabal/Hackage is more mature, this integration could probably be split out.) Others have written a lot of good code; the HaXML, XML-RPC, Peter Simons' projects, Ian Lynagh's code, etc. all contain a lot of code with wide practical use. All of these projects would welcome more coders, I believe. Three of us also have a very rough start on a hands-on, practical introduction to Haskell aimed at the experienced imperative programmer. -- John

On Wed, 8 Dec 2004 19:27:56 +0100, Gour wrote:

...

Three of us also have a very rough start on a hands-on, practical introduction to Haskell aimed at the experienced imperative programmer.

This is *very* important - bringing new people from the imperative (I do not mean they should be 'converted' at any cost :-) camp by showing them how Haskell can be very elegant solution for solving general programming problems and than one does not require to hold a Ph.D. in a CS to be qualified to program in Haskell.

A year or so a go I started a tutorial series with the intent of doing just this. However I only got past a very rough first draft of the first tutorial that, quite frankly, was rubbish, and I didn't have time to finish it. There is still the "why should you care" article available at http://www.haskell.org/complex but the rest of my ambitious project was put on ice. Anyway, I'm currently working on an article for a Swedish print magazine on Haskell (similar to the one linked above, but less argumentative) that's due out at the end of January 2005. Hopefully that will contribute to spark the interest of a few imperative programmers to try Haskell out. I think that's a pretty good way to go about it. If you calmly list the benifits of Haskell, with a few tutorial-ish examples, and get that published in a print magazine (which the mainstream programmers read) I think that could convince quite a few people to give it a try. /S -- Sebastian Sylvan +46(0)736-818655 UIN: 44640862

Ben Rudiak-Gould wrote <snip>

Your newer-is-better premise makes little sense. Haskell is a far "newer" language than Java; many aspects of Haskell's design are no older than Haskell, while nearly all aspects of Java's design have been around in other languages for decades. You might as well be arguing that Java is better because it's based on older, proven technology. Better yet, suppress the urge to compare Haskell and Java at all; after all, the more different they are, the more worthwhile it is to learn both! Once you're reasonably adept at programming in different languages, then you can start thinking about ways to combine the advantages of each.

-- Ben

Interestingly enough not once did I indicate that newer is better or that there was a problem with haskell language. All I said is that the compiler seemed archaic. I will say that it probably comes from my background. I'm sure that if I had started off with C/C++ and moved on to Haskell or that if I started with Haskell first then I would probably see things differently. Jason

On Mon, 6 Dec 2004, Sven Panne wrote:

azrael@demonlords.net wrote:

...

The original observation was that the compiler seems archaic. When asked, I gave some general comments. What I should have just said was that it was to much like a C compiler. Which, no matter how neat you think it is, is archaic.

Hmmm, using the number of files generated from a source program as a measure of the "coolness" of a programming language or its compiler is extremely strange. There's nothing I could care less about if the language itself fulfills my needs. Do you care about the strange intermediate files VisualStudio produces? The contents of you CVS or .svn subdirectories? I'm quite happy being able to ignore these things...

What's unfortunate here is probably that the files are lying around in the same directory as the sources. E.g. the build system of Modula-3 uses a directory structure like this: Project LINUXLIBC6 - object files and other generated files for Linux SOLgnu - object files and other generated files for Solaris src - sources, Makefiles and other data This way not only generated files are out of scope in every day programming but it is also absolutely no pain to develop for several platforms simultaneously. The disadvantage is clearly that a programmer has to conform to this structure, but this could also be considered as advantage. But one could consider it as an disadvantage that is more complicated to work with many but small programs.