[Catalyst] RFC: The paradox of choice in web development

[Octavian Râsnita]

From: "Matt Pitts" <mpitts at a3its.com>
> -----Original Message-----
> From: Octavian Rasnita [mailto:orasnita at gmail.com]
> Sent: Tuesday, February 17, 2009 7:56 AM
> To: The elegant MVC web framework
> Subject: Re: [Catalyst] RFC: The paradox of choice in web development
>
> From: "Ali M." <tclwarrior at gmail.com>
> > When Catalyst is not chosen I personally believe it the combination
> of
> > two things
> > 1. Perl is no longer perceived as an easy language, or language that
> > make development easier.
>
> More exactly,, Perl is considered a language hard to learn, that
> creates a code hard to maintain, a language that uses a strange OOP
> style (because I guess there are no books for Perl beginners that
teach
> about Moose or Mouse), a language which is too flexible and because of
> this it is not prefered by the large teams of programmers because each
> of them could have a different style.
>
> > 2. Catalyst perceivably doesn't offer enough added value for
> > developers who are not that much into Perl
> >    to make the sacrifice and use Perl anyway.
>
> If the programmers are "not that much into perl", this means that they
> don't know how to use DBIx::Class and Catalyst and possibly other few
> modules which are usually used by Catalyst developers, and in that
case
> they can't understand the power of Catalyst.
>
> If Catalyst wants to compete with RoR or other frameworks, it should
be
> as easy to install as those frameworks, and the simple apps should be
> also very easy to create.
>
> The comparisons between web frameworks are not based on the number of
> the requests they serve, or on the number of database tables they
> manage, or on the number of backend servers they are installed on, but
> on the number of web sites that use those frameworks, so those
> comparisons might show that there are 100 sites that use RoR and only
5
> that use Catalyst, but don't tell that 3 from those 5 sites that use
> Catalyst have 3 times more visitors than all those 100 sites that use
> RoR.
> And of course, the conclusion is that RoR is much better.
>
> I think that the success of other languages, especially Python is also
> due to the fact that they support better Windows than Perl.
> WxPython is better developed than WxPerl, there are even screen
readers
> that interact with the GUI of the OS in Windows and Linux, and
> finally... the number of programmers for Windows is bigger than the
> number of programmers for Linux.
> Most Perl programmers use to consider good to publicly despise Windows
> and those who use Windows, and also consider that Perl is a language
> for the web, while those who use Python or even Ruby consider them
very
> good languages for creating programs with a desktop GUI.

> Sad to say, but I completely agree with this. It's quite ironic how the
> drive of open source has only furthered the need for OS agnostic
> software and platforms, which in turn, has actually made life harder for
> things like Perl that have strong origins in *nix OSes.

> "Oh yeah, we love Linux as a platform for its [list of goodies], but we
> can't ask our day-to-day workforce to switch desktops, so we need OS
> agnostic platforms that we can build in Windows and deploy in Linux."
> Seems to be the credo echoing from the business world.

> I myself am currently trying to support multiple developers (content &
> perl) working on a Catalyst app from Windows desktops and it's been a
> bit of a process. Cygwin seems to be providing the best solution right
> now, but Cgywin Perl fork()ing breaks frequently for me in Vista, so no
> HTTP::Prefork, which makes development much, much slower.
>
> I really, really want to be able to "just run" my Cat apps in Windows,
> and I probably could get it going under ActiveState or Strawberry if I
> stuck to it, but I _need_ it to not be that hard. I'm sure I'm not the
> only one.
>
> In today's world of software that is cross-platform and OS agnostic at
> its core, Perl 5 is showing its age. Still love it though.
>
> v/r
> -matt pitts

As someone said it many years ago (but I don't remember who was), Perl is 
dead... or something like that was the idea.
With that ocasion came the idea of creating Perl 6 that should be totally 
different, but who knows when it will be ready.

A better native OOP support in Perl would be wonderful, but I think those 
other ideas about how Perl 6 should look like are more important, like to 
have a kind of virtual machine like in DotNet or Java, and to use bytecode 
precompiled binaries which are totally portable.

Octavian