[TYPO3-dev] Are performance improvements part of maintaince releases?

[Steffen Ritter]

Hi,
I just read the whole thread.
I'm completely with the majority that we should have such performance 
improovements in patchlevels.

But as I figured out in Core Discussion and in addition in this thread, 
this is not a democratic descision. There last-words-people which are 
against it. I do not think that their arguments are "not good" enough, 
but if I outweigh the situation, I'm just the other way around.

We might discuss this endless, there won't be any agree!

So what possibilities do we have:

1. it lasts all as it is, the discussion will be stopped, nothing 
changes, some may be upset and bugged (and probably contributing less)

2. we take the democratic drifts and introduce these features,
others will upset, may not feel good with this as they have to take 
responsibilties for this... If anything might probably break it would be 
like "I said it... So look for at and fix it, I won't do it"

Both possibilities will some kind "break" the harmony in the dev-rounds 
for a longer time, since it is a hot topic and everyone needs to be 
satisfied. We are still humans... so this can't be disowned.

So what?
We need a third possibility.

What might this be... To me this is an ugly solution, but it might be 
the only way out...

We'll release 4.2.7 *WITHOUT* performance improovements.
Right beneath the sources we offer a second archive, containing the 
patched classes to extract, so that all users without knowledge about 
patching might benefit. We should add a text like "These additions can 
boost your system by xy%, it is well tested, considered worth to be 
published, but since there is no garantuee that nothing may break in 
special cases, use it with caution"

And afterwords refer to a, the discussion b, a page which sums up point 
of both sides to make clear interested users why this is handled in this 
way...



regards

Steffen