On Fri, 19 Nov 2004 15:46:39 +0000, Simon Cozens wrote
> I'm not convinced that you make Maypole better by adding more classes.
It is argued that you don't reduce traffic congestion by simply building more
roads, you need to improve the public transport infrastructure. I think what
I've proposed is a justifiable improvement, but I could be wrong. I think I've
said already that I won't rush in to it. We should focus on the important
issues first.
> Every time you add another concept to the core, you increase
> complexity. It may look like abstraction, but it isn't. You weren't
> using the request object for anything else anyway, so you're not
> really abstracting anything.
I may not be using the request object for anything else, but I can't
necessarily say the same for anyone else. I think that there are threads on
this list that demonstrate that not everyone has a clear idea of the model,
view and controller domains. I hope we can go some way to improving that by
building on the manual and having clean examples.
> It's merely added complexity. And every
> time you increase the core complexity of Maypole, God kills a
> kitten. Make peripheral modules as complex as you like - Maypole's
> designed to allow you to do that. But keep the core small. Then you
> don't have to maintain it.
Agreed. And I'm not necessarily proposing a new class, there could simply be a
naming convention. I don't want any dead kittens on my conscience!
In fact, I wish I could remove a class. As useful as it is, I think
Maypole::Application does what Maypole.pm should do. It would simplify things
a lot, but would require turning the current inheritance tree on its head.
> needs to make their mark. Fine. I have a version of Maypole on my
> disk that does what *I* want in case things get too wacky out there.
Wise move :)
> But what I'm (not so) secretly hoping is that when Maypole gets
> cleaned up for a new release and the core implementation is polished
> a little and people WRITE LOTS OF APPLICATIONS IN IT BEFORE HACKING
> IT TO PIECES, we'll discover that the separation of concerns you're
> talking about really isn't that important, when it comes down to it.
You may be right. And I wouldn't mind being proved wrong.
--simonflk
_______________________________________________
maypole mailing list
maypole at lists.netthink.co.uk
http://lists.netthink.co.uk/listinfo/maypole
This archive was generated by hypermail 2.1.3 : Thu Feb 24 2005 - 22:25:57 GMT