[TYPO3-core] Thoughts about mid-term strategy

[François Suter]

Hi all,

Now that the official announcement about TYPO3 CMS 6.2 being the new LTS 
version is out, I think we should have a strategy about how we want to 
reach that important milestone and on what we should focus.

I would place stability as the foremost goal. The Core went through 
quite some changes in recent times and although we have always kept good 
on the promise of backwards-compatibility, there have always been some 
edge cases which made life difficult for one particular extension or 
another. Such obstacles are not impossible to overcome, but we should be 
careful about not piling too many of them on the road to 6.2, so that it 
is not too hard to upgrade sites from 4.5 to 6.2, because there's going 
to be a lot of that.

Thus there should be an evaluation of big changes, to decide whether 
there are okay to go into 6.1 or 6.2 or if they should somehow postponed 
to 6.2+1 (as it is code-named for now), perhaps existing in some fork 
until the time comes to merge them back into the Core.

The other big goal that I see is to finish work that was started and 
clean up pending stuff. With the later I mostly refer to reducing the 
number of pending patches. If we want a rock-stable LTS version, it only 
makes sense that the number of pending patches be reduced to (near) 
zero. Finishing work is about things like using the new Logging API in 
the Core, making better use of system collections, improve system 
categories (important patches pending for a long time), in general make 
thorough (but sensible) use in the Core of technologies that were added 
but are not used to their full potential.

Performance is also a goal, especially since it has come under quite 
some fire recently. It has been mentioned as a goal for nearly every 
recent release, but the numbers seem to prove otherwise. Perhaps setting 
up a reference server might help in ensuring the validity of the test 
results.

I hope this can help in the discussions to set some vision for the near- 
to mid-future and have real goals for the next LTS.

Cheers

-- 

Francois Suter
Cobweb Development Sarl - http://www.cobweb.ch