cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefano Mazzocchi <stef...@apache.org>
Subject Re: [RT] A Groovy Kind of Sitemap
Date Fri, 30 Jul 2004 00:05:37 GMT
Geoff Howard wrote:

> Why insert this stream of consiousness into this discussion?  I have a
> gut feeling that something in this discussion could lead to a solution
> along these or totally new lines that cures this "uneasiness", or
> could make it even worse.

I feel the same way: sitemap and flowscript sometimes feel like 
overseparating the concerns.

But like you, I don't know if a more scriptable syntax for the pipelines 
could allow us to solve this "uneasiness" or to make it worse... the 
"sin" that comes to mind is dynamic pipelines... but isn't this what we 
are doing anyway with selectors and matchers?

[I've been learning a pipe of new programming languages lately.

I think you can tell ;-)]

The sitemap *is* a programming language. It's a pretty kickass one.

I was talking to Sam (Ruby), Ted (Leung) and Fitz (Brian Fitzpatrick) 
comparing java and python (I'm sitting here at Guido's talk about Python 
's state of the union) and they were waying that java is terrible for 
xml programming while python is much better.

I said that the cocoon sitemap is a lot better than both.... but it is 
true that it's not a general purpose language.

cocoon has tons of reusable, first class, pipeline components... the 
only other thing that has such a thing is the UNIX environment.

the sitemap has no parallel in UNIX. Is this good or bad? I don't know.

If we had "first class pipeline objects" as operators in a syntax would 
be good or bad?

Say we had a syntax like

[*]
   <file($1)|
            |xslt(sheet.xslt)[foo:this,bar:that]|
            |cinclude|
            |xslt(sheet2.xslt)|
            |html>

would it be so hard to understand? would be so useful? I don't know.

it is true that no matter what, a new syntax for the sitemap is a lot of 
work.

And let's keep in mind that groovy is not even release final (I talked 
to James Strachan yesterday about this.. and he's obviously very excited 
and told me a way to implement continuations in groovy once "macros" are 
built into the language... which is due in the near future).

Anyway, food for thought.

-- 
Stefano.


Mime
View raw message