[TYPO3-core] RFC: #11265: function getTreeList() doesn't use the caching framework <ot>

David Bruchmann typo3-team-core at bruchmann-web.de
Fri Jun 5 18:34:37 CEST 2009


----- Ursprüngliche Nachricht -----
Von:        Ingo Renner <ingo at typo3.org>
Gesendet:   Freitag, 5. Juni 2009 10:43:21
An:         typo3-team-core at lists.netfielders.de
CC:
Betreff:    Re: [TYPO3-core] RFC: #11265: function getTreeList() doesn't 
use the caching framework
> Rupert Germann wrote:
> 
> Hi Rupi,
> 
>>> It might be possible to use tags to clean the cache...
>> in other words: if you don't want to flush a complete table, tags are the
>> *only* way to clean seperate entries ;-)
> 
> seems so for now... (besides deleting by identifier)
> 
>> But the way how function flushByTags() achieves this currently will bring us
>> into severe memory or performance problems sooner or later.
>> same problem as in:
>> http://bugs.typo3.org/view.php?id=11175
>>
>> for the time being I will use flushByTags() at the missing places, and then
>> I'll take a look to the getByTags(), flushByTags() stuff in caching
>> framework.
> 
> great!
> 
> best
> Ingo
> 

Hi,

It's offtopic, but reading this thread I remembered a question or 
feature-request I have:

short: I'd like to stream content.

description:
TYPO3 fetches content first (saving it in arrays first) renders it and 
at the end the content is written.
In some special cases content is very large or/and needs much time 
beeing fetched. In those cases memory-using and the whole way of 
rendering are not able to serve the request (without AJAX) because 
memory is exeeded.

request:
Id'like beeing able to hold only the footer in memory while header and 
content are written directly.

consideration:
Ajax does allow requests after the page is send and will be better 
because the page is rendered ready before getting the content.
Nevertheless JS is required (AFAIK) at the clientside and I thought 
about a serversided solution.

Regards
David


More information about the TYPO3-team-core mailing list