[Catalyst] Page layout w/Catalyst

Charlton Wilbur cwilbur at tortus.com
Thu Dec 15 16:31:56 CET 2005


On Dec 15, 2005, at 9:30 AM, john_oshea at wordbank.com wrote:

> On 15 Dec 2005, at 14:13, Matt S Trout wrote:
>
>> Seriously, though - this is perl, so TMTOWTDI applies. The reason  
>> I (and I
>> suspect a lot of people) first got addicted to Catalyst isn't so much
>> because it provides easy ways to do things as that it saves you a  
>> lot of
>> common web app pain but still doesn't get in the way of doing  
>> things your
>> way - or whatever way is best for the problem domain you're  
>> applying it to.
>
> I suspect a significant number of people don't actually have a  
> "way" when it comes to web apps. Or are aware that their way is  
> probably going to result in a complete mess as they realise they've  
> not gotten "that MVC thing" completely straight in their heads.  
> Which is where things like the Advent Calendar and people's sample  
> code definitely help.
>
> And is probably the reason why RoR does so well in attracting new  
> people, come to think of it.

There's also the flip side of TMTOWTDI, which is that people who  
don't put sufficient thought into the problem frequently wind up  
picking a poor way to do it.  This is why some frameworks and  
environments are successful:  for any given problem, they offer one  
clear solution, and other solutions are painful enough to steer you  
towards the one clear solution.

I think there are two major reasons that Ruby on Rails is popular  
right now.  One is that it offers a single right way to create web  
applications:  if you follow their conventions, you get a web app  
with minimal pain, and their conventions are chosen to be the right  
thing in the vast majority of cases.  Two is that all sorts of nifty  
AJAX things are built right into it, and not only do you get a web  
app with minimal pain, you get AJAX with minimal pain.

Catalyst, because of its Perlish nature, won't ever appeal to the  
people who want one true way of creating web applications.  At best  
it will appeal to senior technical leads, who choose Catalyst and  
then build their organization's one true way on top of Catalyst.  I'm  
not sure this is a bad thing; if we like TMTOWTDI at the language- 
elements level and at the modules level, we ought to be in favor of  
TMTOWTDI at the meta-language, framework, and environment level.  And  
there are ways of getting nifty AJAX things in Catalyst; from a  
marketing point of view, creating a dozen or so nifty and easy  
Catalyst-AJAX demos would go a long way, I think.  (Maybe such a  
thing exists already, and I just don't know about it.)

(I also think there's a third element at work: Ruby on Rails is not  
Perl, PHP, or Java, and so it provides a politically neutral place to  
go for people who are unhappy with their current Perl, PHP, or Java  
infrastructure.  A lot of the time, people who are using one language  
in that triad have a lot emotionally invested in *not* using either  
of the others, and so when their chosen environment seems inadequate  
it's much easier to make the leap to neutral Ruby than to the enemy  
language.  But this is not something that can be usefully addressed  
technically.)

Charlton


-- 
Charlton Wilbur
cwilbur at tortus.com






More information about the Catalyst mailing list