[TYPO3-dev] typolink generation high parsetime

[Steffen Kamper]

"Popy" <popy.dev at gmail.com> schrieb im Newsbeitrag 
news:mailman.3520.1193387819.20721.typo3-dev at lists.netfielders.de...
> 2007/10/26, Steffen Kamper <steffen at sk-typo3.de>:
>>
>>
>> "Popy" <popy.dev at gmail.com> schrieb im Newsbeitrag
>> news:mailman.3454.1193336329.20721.typo3-dev at lists.netfielders.de...
>> > 2007/10/25, Steffen Kamper <steffen at sk-typo3.de>:
>> >
>> >> but still better than 400 single queries.
>> >> What is big, array of 500(0) integers?
>> >
>> > The query wich will return 500 integers (or more...)
>> > I really think that asking only if needed is better, so, if you build
>> > links
>> > pointing to 5 different pages, your cache array will have only 5
>> entries,
>> > with only 5 littles queries...
>>
>> I don't see this as great progress, as we talk about temporary "cache" 
>> for
>> one page generation.
>> If 500 links are links to 500 pages, nothing is won, still 500 (*3)
>> queries.
>> If we have the array of integers, it's one! query.
>>
>> >
>> > By the way, if the function t3lib_pageSelect::getPage has a cache, it
>> will
>> > reduce the number of performed queries for typolinks, rootlines, pid
>> > lists,
>> > menus, etc...
>>
>> this is another point ;-)
>>
>> vg  Steffen
>
> That's why i speak about this kind of cache :)
> It ensure that we'll never do the same request more than once. Sometimes i
> use this kind of query cache in my extension and it works well.
> Ex: 48 queries processed instead of 293 on my forum on cache generation.
> With cache it is a bit less fun, as I have 6 queries performed instead of
> 11.
>
> -- 

do you thought about using MySQL Query-Cache? This was also discussed by 
core and does exactly the same.

vg  Steffen