[Typo3-dev] Alternative to hooks - fake "$this"

[Frank Benkart [DIACC]]

Hi Martin,
i understand your thinking. But i don't understand why normilisation should
be so important.
Think first we should built a stable and well formed core engine, and than
we can setup a normalisation.

Think that. You will build a Formular 1 car. you don't beginn with the
chassis and the aerodynamic. you start with the engine.

We do things with typo3 were we need very very often core classes an core
features in our extensions. so we need a oop architecture and not that
pseudo oop.

Hope you unsterstand me?

----------------------------------------------------------------------------
----
live strong - live typo3

"Martin T. Kutschker" <Martin.no5pam.Kutschker at blackbox.n0spam.net> schrieb
im Newsbeitrag
news:mailman.1.1099035772.19519.typo3-dev at lists.netfielders.de...
> Frank Benkart wrote:
> >
> > And that is my thinking to make typo3 in the first way not normalised,
the
> > first should be a refactoring of the core engine. Using real oop. If
done
> > this, the second way is the normalisation.
>
> There you go.
>
> I knew we would end up with the OO discussion once we started the
> normalization. And if you start thinking about radical changes you'll
> find a whole lot of things you might want to change (some because they
> really would be better, some only because of an ideal model).
>
> I don't want to discourage anyone in doing all these things, but
> consider the following:
>
> OO-ifying and normalizing will take a long time
> It will break compatibility with current (and future!) extensions
>   (so you have to rewrite them or create new ones)
> Development of TYPO3 will continue
>   (so new features will not be in the "clean" code)
>
> So I ask again:
>
> What are the benefits of all the changes? Some possible gains:
> - easier to maintain (on a code level)
> - easier to read (good for new developers)
> - performance (beware: normalization and classes may bloat the code!)
> - ???
>
> Not much, and the performance gain I doubt. At least without aggressive
> refactoring (eg moving to PHP5 classes, reference only array operations,
> superduper caching of all and everything).
>
> Looks like hard work to me. At least half a year of broken and unusable
> code (depending of course on the number of devs and the devoted time).
> Probably more until the "new" TYPO3 will be up and running. Do you have
> the endurance to do it?
>
> Masi