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] the quest for the perfect template language
Date Mon, 07 Apr 2003 13:48:01 GMT
on 4/6/03 4:17 AM Geoff Howard wrote:

> OK, I'm regretting sending that message - the tone is wrong and 
> I missed the point.  Stefano was talking about a revolutionary 
> idea and I just said "hey, you're talking about a revolutionary 
> idea".  He was asking why people weren't jumping in on it and I 
> wanted to explain why I wasn't and I think air a below the surface 
> concern I had, which is just that reinventing the wheel is an 
> "anti-pattern" - but then again so is continuing to drive on a flat 
> tire.

Well put.

I'm getting more and more concerned about the maintanance costs of
transformation-based templating compared to generation-based templating.

And I'm expressing my concerns.

I also stated that XSLT seems to be pretty close to providing a solution
for a generation-based template system that is efficient and easy to
write, but its syntax will very likely be to hard to use for the people
that will have to write the templates.

*thus*, I thought about creating a non-xml syntax that can be one-2-one
converted to XSLT.

I don't give a damn if I'm the only one who wants to do this. If I think
I need it, I'm going to do it anyway, even if I'm alone (but it seems
I'm not the only one who shares concerns about the XML xslt syntax)

Should I remember you that I was one of the first that wanted to use
XSLT for server side while almost everybody else at that time told me I
was nuts?

I rethink my strategies *exactly* when they become mainstream and I keep
a very open minded view about things.

I try to teach everybody I work with to do the same, but it's a very
hard job. One must have a pretty solid self-confidence to be able to
question his/her foundations continously.

I have learned the very hard way, in my life, not to take for granted
that everybody is able to do it.

But then again, no need to apologize, since I'm not offended anymore by
these things :-)

> I'm not bothered much by the xml syntax and I still think some of the 
> points from my last message apply but I definitely resonate with the 
> other core problems.  

RT are just that: random thoughts.

If they grow into more solid things, great, if not, well, darwin teaches
that most mutations are harmful to a genetic strain. but this never
stopped evolution to do its job rather magnificently.

> How many of these problems would be solved by some combination of:
> - Preprocessing xml to provide a structure more conducive to the 
> type of rule based processing that xslt excels at (like providing 
> count and position information).  I wriggled my way out of several 
> tricky xsl programming issues by simply providing more information 
> coming out of the generator.

Definately true. Look at my Paginator in the scratchpad. It does full
pagination and has a flexible 'pagesheet' that you can apply to
different streams. It could have been done with a stylesheet, but they
tended to become much harder to maintain than java code.

> - Creating a flexible, reusable library of xsl template rules to 
> perform the most common but non-trivial tasks.  Along these lines, 
> I have been needing a good crossbrowser dhtml menu implemented with 
> xslt and have found nothing, though there are many javascript-only 
> solutions.  I'd much rather have the heavy lifting of creating the 
> menu structure performed (and cached) serverside but it's just so 
> easy to plug in the javascript only version.

Imported XSLT become obscure black boxes and you end up rewriting them
yourself most of the time.

I have a tendency to think that XSLT is just like perl: write it for
your simple little needs and prepare to throw it away.

If you can do that, it shines. It's beautiful and very elegant.

But if you plan to reuse it, well, don't! simply use a different
language :-)

But this has probably something to do with my mindset, so if you
disagree, I really don't care, as long as I have the freedom to do what
I need, I'm happy.

I'm not the type of guy that will impose *my* way of reasoning on you
simply because I think it's the best I could find.

I will just talk about it and give you the freedom to use experience it
yourself, then I'll let you decide. And respect you along with your
differences.

Too bad this kind of attitude seems to be getting more and more rare
nowadays.

-- 
Stefano.



Mime
View raw message