[Typo3-dev] singletons instead of pre-loaded objects

[Kasper Skårhøj]

My Final Words:

> I see what you mean, and I agree. My point in my arguments was:  Typo3 is
> the way it is and won't probably change fundamentally, and thus we need
> very good documentation. You can expect Kasper to be open to ideas that he
> expects to bring general benefit to Typo3's future, but which doesn't open
> the need to rewrite major Typo3 parts and which ensure to be more or less
> compatible with existing extentions.

100% right.

But I do agree with many of Leenderts points, but the discussion just
reminds me of all the insufficiencies, rather than motivates for change.
I haven't had a single "Oh, I never though of that" moment in this
thread, only a very discouraging feeling that I rather want to leave it
all and do something else in my life. Remember; 5-6 years of dedicated
development on a single thing, having sacrificed a lot for the sake of
the project (=all of you); It is of course rewarding but no dance on
roses. It has worn me out, made me tired and subjectively sensible. Each
time I try to tell you all "Focus on extensions, focus on extensions" it
is because I want to divert you away from this itchy, negative spot and
focus on the area where the positive opportunities for the future are! 

If any of you think that we can basically "renovate" the whole core over
the next years to a totally OO design, then I don't think that is worth
the effort. If you disagree with this I want to remind you that I know
how big a job it has been until now to keep all of it in balance.

So the bottom line is:
- TYPO3s core will not change fundamentally. 
- New extensions *could* introduce more modern designs internally, but
this depends on the developers of those extensions. (This is
"possibility-land" - go, inhabit it!)

> 
> For some time I was thinking about how cool it would be to have a Typo3
> CMS (with such a nice backend and other cool ideas) properly OO-designed.
> Maybe someday someone will do this, but as Kasper noted: I can't imagine
> someone as motivated as Kasper in the last years in the near future to do
> this. But... Who knows??

Again, this is exactly right.

We could very well see the current TYPO3 system as the "prototype" for a
*complete rewrite*. This will give an enourmous base of experience from
which we can take all the success principles and discard all the
nonsense which *always* (to my defense!) will build up when you develop
something step-by-step into unknown land. 

That rewrite could be totally OO designed. The only mistake you could
make is to try to make it backwards compatible. It should *NOT* be
designed so! A rewrite would still have many things in common with the
old TYPO3 but the design should not be limited by backward compatibility
issues - then you gain nothing. Rather it should be a clear break, but
with some ways of converting data/sites to the new system.

Actually, making the rewrite is not a new idea. We had that on a meeting
in Düsseldorf last september. I was actually quite thrilled about it for
a day or two - just until I came home to the real-world problems. Then I
realized that such a project would take 1-2 years for 5 developers to
complete to a level where it would be worth using instead of the "old"
TYPO3 system. In Denmark we have a saying that "One bird in the hand is
better than 10 on the roof" - and therefore I'm personally staying with
our insufficient, poorly designed, old-fashioned but very capable and
extendable CMS "TYPO3 first generation".



With all respect for differing opinions but also for my time,


- kasper

--------
Please notice NEW EMAIL ADDRESS for 2004!! (due to SPAM-contamination)
	
"kasper2004 at typo3.com"