[Zope] Java re-invents DTML :-)

[Tino Wildenhain]

Hi,


On Fri, 14 Feb 2003 07:51:22 +0100
Dario Lopez-K=E4sten <dario@ita.chalmers.se> wrote:

> From: "Ausum Studio" <ausum_studio@hotmail.com>
> > > Maybe you should write the inventor of this very complicated way of
> > > doing scripting and say "Hey, we tried that. It's a bad idea." :)
> >
> > Yes, for many applications. For many others DTML is a strong language t=
hat
> > will do the job that page templates couldn't. So my humble recommendati=
on
> > to newbies is don't forget to learn DTML. Some day you will rediscover =
it.
> >
> > http://www.zope.org/Members/ausum/CMFPortlets/Docs/readme_052/index_htm=
l#c
> >
>=20
> Sonce so many people are bashing DTML nowadays I feel the need to +1 on t=
his
> and on the referenced article.

at least -0.8 from me ;)

>=20
> When looking into customising Plone and CPS for instance, I find that many
> ZPT pages look to be as hard and tedious to follow as some of the my own =
bad
> DTML pages are.

I saw that. The problem is, the authors of said pages still do it the wrong=
=20
way [tm]. There is still much too much DTML bias in the code. Its biggest
manifestation is the often used tal:define.

Someone told me if he has to write all the code into seperate python=20
scripts, he would need a BtreeFolder to hold them all - but this is=20
pointless, because the code has to be looked up anyways and I bet a
folder is a better container then a page template.

>=20
> This is no bashing of ZPT mind you, I just think that if DTML is being
> misused, then it isn't DTML's fault. And I am sure that ZPT can be misused
> in a similar manner.

Unfortunately yes.

We need a get-values-or-actions-on-first-use-mechanism to get full use
of the abstraction of ZPT. The implementations that exists are
often too complex for most people so its rarely used. For example
ZPatterns.

Regards
Tino