[web-devel] Re: ANN: Salvia-1.0.0

[Gour]

On Mon, 22 Mar 2010 15:59:24 +0100
>>>>>> "Sebastiaan" == Sebastiaan Visser
>>>>>> <sfvisser at cs.uu.nl> wrote:

Sebastiaan> How do I put this delicately? I do not think WAI is the way
Sebastiaan> to go.

OK.

Sebastiaan> Secondly, this type of interface will degrade projects like
Sebastiaan> Happstack and Salvia to be mere `back-ends', which they are
Sebastiaan> not. 

I saw there is a recent thread in Happs mailing list entitled
"experimental port of happstack-server to wai", and although I cannot
comment on technical issues, it looks that they do not consider that
Happs will become degraded?

Sebastiaan> What I do like to see though, is something like the Ruby
Sebastiaan> Rack interface, which calls itself a "a Ruby Webserver
Sebastiaan> Interface". This is an interface between the actual
Sebastiaan> OS/network and the application server, not an interface
Sebastiaan> between an application server and the web application. Call
Sebastiaan> it WSI not WAI, this subtle difference actually means a lot.

Well, I'm not attached whether it is called WSI or WAI, just to see
that the web-development market is not so fragmented, so that we can
get some decent frameworks in order to make it easier to do web
programming in Haskell.

Sebastiaan> No unfortunately, there is currently only the examples and
Sebastiaan> the haddock documentation. We are working on a tutorial,
Sebastiaan> but that will have to wait for a while. Sorry for that.

Heh, this is common scenario. :-)

Sebastiaan> It can run on its own server (like Happs) or as CGI.

What about FastCGI?

Sebastiaan> Whether you can easily deploy it probably depends on your
Sebastiaan> hosting party, not on Salvia.

Well, I'm curious if one can deploy it on shared-hosting like
e.g. Webfaction since not everyone can afford or like to admin VPS.

Sebastiaan> It can run as CGI or behind existing apache proxies without
Sebastiaan> any problem. The codebase has some experimental support for
Sebastiaan> running Salvia itself as a proxy server, but this code is
Sebastiaan> not released yet because it needs some additional
Sebastiaan> coding/polish/testing.

Good. I use Cherokee on webfaction - that's why I'm asking.

Sebastiaan> No, this is a web server package not a database. Off course
Sebastiaan> you can use it to build database backed web applications,
Sebastiaan> using any datastore you like. The Salvia project will never
Sebastiaan> have built-in support for any form of data storage, I
Sebastiaan> prefer to keep these aspects totally separated.

Ahh, I read announcement where it says that "Salvia is a feature rich
web server and web application framework", so based on my experiences
with e.g. Python frameworks, there is usually some kind of ORM support
so that one can plug database easily. That's why I am expecting to
see the easy way to e.g. use HDBC or something.

Sebastiaan> I feel that Haskell is the true candidate for being the
Sebastiaan> perfect web language, but we still have a long way to go.

Heh, I've similar feeling...lot of potential, but (too) many different
solutions.

Sebastiaan> We should just keep on playing with new ideas and
Sebastiaan> abstractions until we all see the big picture. Some
Sebastiaan> standards will arise eventually.

Well, I know some people who object about HP thinking that it is not
good that some libs (from the same category) were preferred ove the
others, but I still feel that having HP or Haskell's 'batteries
included' is good stuff to consolidate things a bit and provide
assurance there is some 'standard set' of libs playing nicely together
and that it will be supported.

That's why I fail to understand why joining the forces and trying to
evolve some 'standard practice' for web development is considered
pre-mature?


Similarly to HP, nobody is forced to use it and there is always free
choise to choose whatever is suitable for someone, but HP is
appreciated by many who wants stable & supported development platform.

Although I'd like that it does not happen, it seems there is (some)
truth in words of John who wrote in another reply: "...i think it's a
fine idea, but it won't be widely used until the haskell community
actually goes through the sort of pain and suffering that forged the
resolve to use WSGI in python. We have to learn that lesson the hard
way."

Until then, Ruby & Python will continue increasing their web 'share' -
"Avoid success at all costs." :-D


Sincerely,
Gour 

-- 

Gour  | Hlapicina, Croatia  | GPG key: F96FF5F6
----------------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
Url : http://www.haskell.org/pipermail/web-devel/attachments/20100323/8928421e/signature.bin