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

[Ingo Renner]

bernd wilke wrote:
> On Thu, 25 Jun 2009 17:15:38 +0200 Ingo Renner wrote:
>
>> Marcus Krause wrote:
>>
>> Hi all,
>>
>> I think you provided some good arguments. However, I'd also to cite
>> wikipedia:
>>
>> "A software bug is the common term used to describe an error, flaw,
>> mistake, failure, or fault in a computer program that prevents it from
>> behaving as intended (e.g., producing an incorrect or unexpected
>> result)." [http://en.wikipedia.org/wiki/Software_bug]
>>
>> This is also the way I see it and why I do not consider performance
>> tuning as bugfixing and thus I would not like to see that in
>> maintainance releases.
>
>
> "behaving as intended."
> so it is never said explicit nearly everyone assumes that this includes
> 'done in the best time possible' (in balance to the available resources).
>
> If a website needs 10 minutes to respond in perfect XHTML nobody would
> assume this as "behaving as intended" so the website is done in perfect
> XHTML.
> And not every TYPO3-installation is done on a high-performance server.
> Therefore a speed-up can be a great 'bugfix'.
>
>
>>> The system works for every user. A performance "bugfix" will only make
>>> it run faster. That's good but it does not fix an error, it just
>>> improves usability. But we do risk that something is seriously broken
>>> afterwards.
>>
>> exactly.
>
> every change can lead to another bigger break on other places.
> We all assume that patches are tested for the most common usages. I think
> nobody expects to get a warrenty for 100% stability and security.
>
> I would like to have this patch as 'bugfix' included in next maintainance
> release.

Honestly even I would like to have it personally. But being the release 
manager/maintainer brings more responsibility with it than just thinking 
about one owns preferences. As Michael already mentioned, and I agree 
with him in that point, there are other people out there, customers and 
users, who are hugely underrepresented in this list. We as the release 
managers and in the end responsible guys have to make sure that a stable 
system is released and also that some important principles of software 
development are adhered to.

On the other hand I would like to treat everyone the same, now if I let 
Rupis patch through I also  have to let other performance improvements 
through. Now with Rupis patch it might be easy as I had a look at it and 
I regard it harmless, but now if I let it through I might have to let 
other patches through concerning performance improvements which are less 
harmless.

You see the points? It's not always as simple as it looks in the first 
place and many considerations need to be made.


all the best
Ingo

P.S. If there's going to be a xclass speed up extension I'm probably 
also going to iunstall it or even patch the core myself... So please 
don't get me wrong, but understand that it isn't that simple.

-- 
Ingo Renner
TYPO3 Core Developer, Release Manager TYPO3 4.2