At 12:50 PM 8/10/00 -0700, Shannon -jj Behrens wrote:
>Chuck,
>
>We have all agreed that we are tired of duplicating effort. I would
>very much like to merge our two products (i.e. Aquarium and WebWare), as
>each has a lot of stuff the other doesn't (if you look closely, you'll
>find that Aquarium does indeed have stuff that WebWare doesn't).
>However, I do have some issues. Please feel free to respond to my
>issues:
Sounds interesting... Fenster's author, Dan Green, has also expressed a
similar interest although I think his product is young enough that he's
joining the Webware effort rather than asking for a merger.
Perhaps this week is the turning point of some consolidation of the
infamous "python web module authors".
I could definitely be interested in a merger with Aquarium. As you've
noted, there are things to discuss...
>The FreeEnergy framework - The FreeEnergy framework is best documented
>by FRD for the Java version of FreeEnergy
><http://java.apache.org/turbine/fsd.html>. It has helped up with
>numerous clients and is the basis of FreeTrade, an open source ecommerce
>framework written in PHP, which I am coauthor of
><http://working-dogs.com/freetrade/>. FreeEnergy has proven to be a
>successful approach to Web application development, and I will not very
>easily give up such an approach. Furthermore, Aquarium implements the
>FreeEnergy framework in such a clean, flexible, OOP manner that it would
>be a shame to give it up.
Admittedly, I haven't had time to check this out yet. I will, though.
>Coding style - I am an admitted anal perfectionist. If you've looked at
>my code, perhaps you'll agree that it's extremely clean and well
>documented (at least that's my goal). Part of my goal in creating
>Aquarium was that the code would be so cleanly coded that other
>developers would find it easy to hack the source. Afterall, an aquarium
>is a transparent structure of great strength. It yields a great amount
>of support, but it doesn't necessarily hide anything. This is the
>spirit of Aquarium.
I started reading one of the main scripts. I think it was the "engine". I
did in fact, notice lots of comments which was nice.
However, I also noticed the script was, er, "naked". That is to say it was
just a series of statements in a file as opposed to a class with methods.
Were it a class, I could subclass it and override a method to customize. As
it stands now, I would have to, as you put it, hack the existing source. In
the event that my hack is special to my needs or gets rejected for
inclusion in the main source, I have to rehack every time I get a new version.
That makes me shudder.
I didn't get much further than that so I'm not sure to what extent this
applies to other parts of the code.
As you say, you're big on clean, commented code. I'm also big on objects.
It's almost as easy to write as a class as not, and later you'll find that
you can subclass, customize, instantiate, cache, etc. to your heart's
content often in ways that are eminently useful but weren't originally
planned for.
"The Object is the Way."
Unfortunately, I run into to people who have been scarred by C++ or VB and
associate "object" with "bad". I was lucky enough to spend 6 years with
Objective-C and NEXTSTEP.
>Leadership - Naturally, it would be tough to yield the position of power
>that I have as project lead of Aquarium. However, I don't really care
>about leading half as much as I care about clean designs and
>ultra-clean, well-commented code. In any case, the most successful open
>source projects are not stiffled by an over-authoritative leader, rather
>they are led by the contributors--i.e. coding merit determines
>leadership.
Well, there isn't a whole lot more to say here. I do, in fact, consider
myself the chief architect for Webware and therefore exercise some
authority in terms of design and pushing forward. This isn't different than
many other successful OS efforts like Linux, BSD, Emacs, etc.
I'm glad to hear that you care so much about ultra-clean, commented code. I
do as well, although I occasionally let something slide while we mull over
the possibilities. There's always a balance to be struck between the
various factors of a project.
>Those are my concerns. Please let me say that I found Web Ware to be a
>very good project. Furthermore, I hope you are not insulted by anything
>I have said. If you are, please accept my apologies.
Not insulted. Honest dialogue is what it's going to take to collaborate on
anything.
Can you shoot us a brief list of where Webware and Aquarium overlap and
what Aquarium has that Webware doesn't?
BTW if you look through the Webware discussion list archives, I posted a TO
DO list just recently. Or I can send it to you....
Also, I've CCed the webware-discuss list.
-Chuck
|