[Haskell] Making Haskell more open

[Gour]

Simon Peyton-Jones (simonpj at microsoft.com) wrote:

Dear Simon,
> Dear Haskell folk
> 
> One thing that hit me forcibly during ICFP in Tallinn, and the
> associated workshops, is that the Haskell community may not be as good
> as (say) the Perl community at engaging and involving the people "in
> the trenches" [PRL].  Haskell.org is centrally maintained by a couple of
> (excellent) people; GHC is still over-dependent on Simon and me; we
> don't yet have a good central site for offering libraries; and so on.

it is very encouraging to see that despite working on a capital Haskell
project, you are thinking about the global community as well.

> However, I still wonder if there are things we could do that would make
> it easier for people to contribute.  Here are two concrete suggestions:

[snip]

> Many of you will know much more about this kind of thing than I 
> do.   Share your wisdom with the mailing list.  

In the light of your suggestions, proposals, I'll (despite being a
Haskell noob) share my ideas¹, which are based on one assumption -
haskell.org can be used as a bed for the whole thing...

¹ I'll use letter chars to denote that points are not listed according
to priorities


a) convert the whole haskell.org site into a community portal 

by using some CMS (eg. Drupal - http://drupal.org/ - which is used on LTU &
Haskell Sequence sites).

By giving accounts to trusted users, creating of content can be made
much easier and we'll get a more material published in less time.

I'd like that we keep the present css style of the site (color theme)
which I find very nice and appealing, but use the Drupal's feature to
make some categories from the present homepage's links like:

	i)   A Short Introduction to Haskell can be made like About menu
	bar
	ii)  Definition of Haskell   - another category in the main menu
	bar
	iii) Books and Tutorials²   - another category
	iv)  Libraries and Tools² 
	iv)  Links to Haskell-related blogs & sites
	..	...
	xx)  Forums³

² those type of categories I'd made like e.g. http://gnomefiles.org/ & 
http://www.kde-apps.org/ sites where the books/tutorials/aplications
can be catalogized according to the software type, reviewed & rated by users,
and relevant links with homepage, screenshots, version histories etc.
can be provided.
		
³ I'll mention and (try to) justify them a little bit later

The rest of the haskell.org portal site can be used for displaying
latest forum topics, important news, quick links etc.

b) make a haskell.org repository for haskell-related projects

by adding some required features so that Haskell projects can move from
e.g. SF to haskell.org. Here I'm thinking about Trac - wiki and issue tracking 
system (http://projects.edgewall.com/trac/).

It is simple & powerful enough. It combines wiki (the present hawiki
entries can be imported (http://projects.edgewall.com/trac/ticket/2068)
with the ticket system handy for handling bug requests, feature
requests, setting roadmaps, timelines etc. and great thing that there is
a darcs backend available so every proejct can have darcs repository
avaialable. (this is why I prefer trac over e.g. bugzilla, RT...)

Several projects (ghc, gtk2hs) are already hosted on haskell.org, but
with the Trac it is very easy to provide SF-like features on haskell.org
with not so complicated setup & maintainance.

So projects (e.g. gtk2hs) can have their own style & features (gallery,
blog...), while other projects can use just a Trac and have a working
site with the darcs repo in practically no time (see e.g. opensync.org
site).

c) providing unified front-end for mail-archives & search

By implementing point b) several projects can move their mailing lists & 
mail-archives to haskell.org and we can setup front-end for
searching/browsing all the haskell-related lists like 
(see e.g. http://mail.gnome.org/archives/)


d) enhance the present wiki system

The present hawiki is very rich and provides lot of information for
haskell searchers, but it is, imho, too flat in the sense that it could
be enhanced by providing better Table Of Contents and having shorter
pages to find material more quickly.

Some of the wiki entries could find their place (link) on the portal's
homepage, in the Docs section etc.


e) forums

Several posts were regarding forums/newsgroups/mailing lists/irc and I
consider that forums can be very handy feature for the new Haskell
community portal site :-)

Let us consider the pro/cons of the present communication means:
	i)   mailing lists
		pro)	- very convenient interface
			- searchable archives
			
		cons)	- it is not easy to jump in the discussion after
			some time
			- it requires subscription

	ii)  #irc
		pro) 	- very interactive
			- quick solutions for many problems
			
		cons)	- it can be very time-consuming (discussion can go 
			astray or nothing is happening at all)
			- the same questions/answers are regularly
			  asked/answered
	
	iii) forum
		pro)	- provides ability to have sticky posts for FAQ,
			howtos, etc.
			- searchable archives
			- one can always jump in the thread by replying
			  to appropriate post (even much later than
			  original post is posted)
			- ability to have email notification when some
			  replies are posted

The potential of newsgroup was also mentioned - creating of
compl.lang.haskell, but I won't comment of it considering that the
newsgroup cannot be one & all solution, and, otoh, does not, imho, provide any
substantial advantage over the other three forms (we already have lists
& irc, and forums come 'for free' with CMS).


f) enhancing & opening GHC manual

Besides my vote for simple authoring tool for working on GHC manual,
Drupal offers so called 'Collaborative Book' (http://drupal.org/node/284)
feature which can even export to DocBook XML in case we want to preserve its
format.


Conclusion:

The above listed proposals can help to put Haskell comunity even more
together by providing central place for sharing ideas, news,
software...and by using ticket system and having concrete timelines,
roadmaps, tasks...it will become much easier to engage people with different 
skills contribute to the whole community in many ways: 

	i)   writing simple patches for certain project
	ii)  becoming developer		...
	ii)  providing translation 
	iii) testing
	iv)  writing documentation (manuals, tutorials...)
	v)   web design
	vi)  administration & maintenance (forums, lists...)
	vii) writing news articles...

The result of the above points would be that we can provide the strategy
for the further develpment of the whole community by: 
	i)  organizing bug-days and/or bug-week to squash bugs (ghc example) 
	ii) providing more man-power for the apps which are important for 
	promoting Haskell as general programming language further 
	iii) ...

Shortly, by having concrete strategy how to enhance/improve the whole
community (language, applications, libs, documentation) we can hope to
attract much more man-power (non-PhD users :-) & money-power (we will
also need some funds for maintaining hardware, (maybe) paying some
programmers to work full-time on certain projects etc.)

> Lastly, when it comes down to it, none of these things will happen
> unless some people volunteer to push them forward.  

David Roundy and darcs are nice example how he managed to attract lot of
people to contibute a little to the whole project.

otoh, I'm 'thankful' to Duncan for inspiring me & engaging my (not so developed) skills to contribute somehow (gtk2hs).

> Would any of you like to contribute your time and expertise?

Of course, I'm ready to help according to time & skills (not so much an
expert :-)

Huh, now it is enough, you canstart flaming me. I'm ready working on my
false-ego anyway :-)

Sincerely,
Gour

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