[XML-SIG] Re: Re: Maki: great!

Alessandro Bottoni abottoni@quadrante.com
Thu, 21 Feb 2002 11:46:00 +0100

----- Original Message -----
From: "Sam Brauer" <sam@webslingerZ.com>
To: <xml-sig@python.org>
Sent: Wednesday, February 20, 2002 7:45 PM
Subject: [XML-SIG] Re: Re: Maki: great!

[Sam Brauer]
maki comes with some example logicsheets.  One provides a set of tags that
allows you to embed SQL queries in your pages and inserts the results as
several nested elements in the page.  Another provides a tag set for
obtaining data from an HTTP request.
Example pages that use both of these are installed with maki.

I would like to include more examples, but nothing so elaborate as a

[Alessandro Bottoni]
Of course! The existing examples will be enough (even if I did not study
them yet...).

[Sam Brauer]
As other people have pointed out on this list, there are already
Python-based portal systems you could use.  To write a maki-based one
that's comparable would not be a trivial undertaking.  I'm sure it could
be done, but I'm not going to do it unless I need to create a portal site
for a client and find that the existing systems don't suit my needs.

If you write one, I would definitely check it out :)

[Alessandro Bottoni]
I cannot agree. I looked for something like ezPublish (or even the less
sophisticated PHPNuke) in the Python world for a while. There is nothing
like that.
WebWare is more an application server than a framework. You have to develop
the whole web app from scratch and this is not what solves my problems.
Zope is too "rigid" and complex. You have to study it in depth and this
require time, the time we do not have. Moreover, it is quite hard to change
or extend it.
No, sorry, there is any real alternative to ezPublish or PHPNuke in the
Python world, at the moment.

Deeveloping a Maki based system is a hard task, for sure, but it is the kind
of task we are payed for. I will largely prefer to develop a "archetypic
portal" like ezPublish once rather than re-write it several times, with just
small differencers. If you use ezPublish for a while you realize this: a
good system like that can meet the needs of the customer quite often and
change it is not very difficult. It is worth to show it to the customer and
search a meeting point between what it can do (with small changes) and what
the customer wants (normally, small UI toys and a gorgeous graphic). This
saves a lot of time and efforts.

[Sam Brauer]
I'm using maki right now to do a site that has all the tasks you listed
(article publishing, poll management, calendar, etc).  What I like about
maki is that it provides a great framework on which to build these
things, and it's generic enough that one could build almost any type of
webapp on it.  But writing webapps that are generally reusable for many
websites is hard.   The next time I have a need for an article publishing
system, I'll probably use the one I'm writing now as a starting point.
But I'm sure I'll have to make modifications so that it suits the needs of
the next project.  Though both sites have a concept of articles and a need
to publish them, each site may have very different concepts of what
constitutes an article, or how articles should be categorized, or how
articles can be related to other articles, etc.
Gee... I'm getting depressed now.  On the positive side, at least there's
job security in knowing that there's always more custom work to do ;)

[Alessandro Bottoni]
I cannot agree, again :-)
Making a "standard" web app is not so hard and ezPublish is a good
demonstration of this concept. Most customers want the same features with
just small differences. If your company has a skilled salesman, he/she can
easily show a standard portal like ezPublish to the customer, discuss it
with the customer and bring you a customer request that is 80-90% identical
to ezPublish. Quite often, the most important change is the site graphics.
Most features are already presents and you just have to configure them. Real
code chenges are quite rare and, in any case, can be easily be implmented in
a resuable way. That is a good care for depression, IMHO.

See you

Alessandro Bottoni