You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@sling.apache.org by Felix Meschberger <fm...@adobe.com> on 2011/03/30 17:00:17 UTC
Changes in the Sling Engine
Hi all,
I noticed a subtle issue with how the Sling Engine buffers output
(response data) from servlets and scripts. See SLING-2039 [1] for
details.
I have now removed all output buffering from the Sling Engine. This
makes request processing and the RequestData and ContentData objects
much simpler.
Now this buffering dated back to well before Apache Sling was born when
we were running the pre-Sling stuff on an internal low-profile Servlet
Container which had no response buffering. Nowadays, servlet containers
all have response buffering and there is no use for Sling to duplicate
that....
So, I expect to run Sling requests slightly faster (no buffer allocation
and no multiple byte copying). Yet, if you happen to encounter some
strange situations, please report. Thanks.
Regards
Felix
[1] https://issues.apache.org/jira/browse/SLING-2039
Re: Changes in the Sling Engine
Posted by Carsten Ziegeler <cz...@apache.org>.
In addition I've cleaned up the per request data object handling in
SLING-2040 and SLING-2042. We could remove some object creation and
remove a per request stack.
Regards
Carsten
Felix Meschberger wrote
> Hi all,
>
> I noticed a subtle issue with how the Sling Engine buffers output
> (response data) from servlets and scripts. See SLING-2039 [1] for
> details.
>
> I have now removed all output buffering from the Sling Engine. This
> makes request processing and the RequestData and ContentData objects
> much simpler.
>
> Now this buffering dated back to well before Apache Sling was born when
> we were running the pre-Sling stuff on an internal low-profile Servlet
> Container which had no response buffering. Nowadays, servlet containers
> all have response buffering and there is no use for Sling to duplicate
> that....
>
> So, I expect to run Sling requests slightly faster (no buffer allocation
> and no multiple byte copying). Yet, if you happen to encounter some
> strange situations, please report. Thanks.
>
> Regards
> Felix
>
> [1] https://issues.apache.org/jira/browse/SLING-2039
>
>
--
Carsten Ziegeler
cziegeler@apache.org