|Think about Loose Coupling|
Thanks for the link to the paper. I found it interesting. I also found it somewhat biased and amusing.
I found it amusing because section 7.1 is labeled "Pull Strategy Violates Separation" - but he never treats the topic of whether "Push Strategy Violates separation."
There are also statements such as "Strictly speaking the URL structure of the site belongs to the controller." This is nice in theory and is true in ideal situations. After 10 years of work in non-ideal situations I've found that it "strictly" isn't true. Sometimes templates contain links that take you in and out of the controller loop.
The true bias is revealed when the last section of the paper is used to present his "new" templating system which does clear up some fuzzy boundaries - but doesn't provide anything truly remarkable over existing systems.
Formal papers are nice in an academic sense, but that doesn't mean his experience has any direct bearing on the experience of others. He is welcome to argue his point, but in the end - all of the various models have managed to get the job done for people who have used them. Existing solutions are certainly good enough for all practical purposes.
my @a=qw(random brilliant braindead); print $a[rand(@a)];
In reply to Re: Petal is pull-style... like TT/mason - Seamstress is push-style