[TYPO3-core] RFC #11358: Bug: typoLink() perfomance

[Michael Stucki]

Hi Rupi,

Rupert Germann schrieb:
> Ingo Renner wrote:
>> I've reverted that change from 4.2 as performance tuning is not part of
>> maintainance releases.
> 
> Ingo, 
> 
> please think a bit about the consequences of reverting this patch and also
> preventing 11317 from being applied to 4_2.

Actually the patch should never have been committed. The rule is that 
two core developers must give their approval, and so far there is only 
yours.

While I'm with Ingo that this should not go into a stable branch, I 
would have written this here to start the discussion. However, this has 
the risk that things are forgotten and the patch stays in TYPO3_4-2 
although it shouldn't. So Ingo did probably choose a more radical 
approach than I would have chosen.

In any case, I agree that such a minor enhancement is of the category 
"improvement" not "bugfix". I have to admit that I currently don't see 
how the code might fail, but in any case the risk to break something is 
always higher than without a patch.

If a stable branch must qualify as such, I consider every change as a 
potential risk.

> Do we really need more XCLASS-the-core extensions in TER?

There are many other performance improvements in Trunk like the treelist 
cache that Ingo wrote some time ago. I can't remember anyone writing an 
XCLASS extension just because it improves the performance. Even if, I 
still think it's a very bad strategy to do so ;-)

> and who says that "performance tuning is not part of maintainance releases"?

http://typo3.org/teams/core/resources/release-workflow/

> Even if such a rule exists - what I doubt btw. - it would be wrong.
> We're about to release TYPO3 4.2.7 in a version that is slower than needed
> despite we have the information and the patches how to make it faster.
> Come on, we know both how this discussion will end once it has hit the
> general mailing list.  

Everyone wants improvements, that is just natural. However, the majority 
of users prefer stability over improvements. There is no urgent need to 
improve the performance. It's just nice to have.

> This practice reminds me on commercial software companies who are holding
> back obvious improvements from updates to force people to by the next
> version. 

If I would be 100% sure that there is no danger in such a neat feature, 
I would probably agree with you.

I'm not holding it back because of forcing people to use 4.3. Instead, I 
need to assure that the quality is kept. See the difference?

- michael
-- 
Use a newsreader! Check out
http://typo3.org/community/mailing-lists/use-a-news-reader/