[web-devel] Re: consolidating Haskell's web-framework scene

[Michael Snoyman]

I was planning on writing a blog post called "call to arms," which would
have been something similar to what you said. I agree that there is far too
much fragmentation in the community; I debated starting Yesod at all when
there were already a number of frameworks out there (Happs, KIbro,
Turbinado). However, none of them hit the spot that I was aiming for, so I
created it.

At the same time, Jinjing came out with Hack, and I tried to go with that.
Unfortunately, there are some inherent flaws with Hack (you can review the
Hack-discuss mailing list to see some discussions over there), so I started
WAI. So it would seem that so far, I've done more fragmentation than
consolidation :(.

On the other hard, the other frameworks having had much development, Yesod
is still active, and it looks like Happstack will be joining WAI, so not all
is lost. I am very interested in trying to continue consolidation, but I
think aiming for the One True Haskell Framework is a bad idea. Python- the
community known for One Way To Do It- can't even consolidate around a single
framework, so Haskell- the language of We Can All Write A Monad Transformers
Library- doesn't stand a chance.

So where can we consolidate? On libraries. I have purposely released a
number of the components of Yesod as separate libraries that are available
to any framework author (or even someone *not* using a framework). Here are
some in particular to look at:

wai
wai-extra
web-encodings
clientsession
authenticate
data-object, data-object-json and data-object-yaml
failure and control-monad-failure
hack-handler-webkit

Some of these packages have very obvious room for future development, and if
someone (like Gour) is looking for somewhere to start contributing to the
Haskell web community, I would recommend one of the following:


authenticate currently supports OpenId 1 and RPXnow. I would like to have
authenticate itself support *all* sites that RPXnow does. The first step in
that process would be to get authenticate to use the openid package and
determine whether a server support openid 1 or 2. From there, Twitter,
Facebook, Windows Live support would all be nice.


hack-handler-webkit is, IMO, a really cool idea. It allows you to convert
any Hack application into a desktop app. It runs simpleserver and then opens
up a Webkit window that connects to it. This is much better than simply
running the simpleserver and then opening up a webbrowser, because:

1) The program terminates automatically when the web browser closes.
2) You can control menu bars, title bar, etc completely, making it *feel*
like a desktop app.
3) You can ensure that all of the functionality you want is present in
webkit.

Unfortunately, hack-handler-webkit has two problems:

1) It's written for hack instead of WAI. However, this should take all of 5
minutes to fix.
2) It only works on Linux, and even there I don't have any documentation on
setting it up.


Yesod is still missing a model layer. This is a project that I will most
likely be doing myself, since it will need to be designed to integrate very
closely with Yesod, but I'll mention some of the design ideas here. (By the
way, Chris Eidhof has a package on Github called Basil which looks very
interesting here.)

1) I want to use YAML quasi-quoting to define the models.
2) I don't want a super-complex "object relational mapping"; Haskell is
already much closer to the relational level than an OO language.
3) Utilize IO thunks (ie, lazy IO loading) so that we don't need to load
everything into memory at the beginning.
4) Not tied down to SQL *or* in-memory databases. It should support both of
those, but *also* support a system that is usable even in a CGI setup. I'm
contemplating now a database that uses multiple directories/files and uses
append-only to avoid concurrency issues.


Compile-time checking of templates would be great. The Bravo package looks
interesting, but I'm a little dubious about the approach used. I think
something based on typeclasses might be a bit more nimble. Once again, tight
integration with the model system I theoretically design would be nice.


So, in summary of a much-too-long e-mail, let's not bother trying to get
everyone to agree on One Framework. Happs has its place, Yesod has its
place, and I know others are using Loli and probably Kibro. (I'm not sure if
anyone every got started with Turbinado.) However, we can raise the bar for
*all* packages simultaneously by making a more solid ecosystem of Haskell
web packages available that any framework can cherry-pick from.

Michael
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.haskell.org/pipermail/web-devel/attachments/20100314/c9dc93b3/attachment-0001.html