You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@myfaces.apache.org by Gabrielle Crawford <ga...@oracle.com> on 2009/12/02 01:16:02 UTC
[Trinidad] remove application view cache in Trinidad 2?
Hi,
I'm working on state saving issues in Trinidad 2 (for JSF 2). I'm just
wondering if we really want to support application view cache going
forward.
The application view cache has some limitations that make me wonder how
commonly it's used, see the doc under "The Application View Cache"
http://myfaces.apache.org/trinidad/trinidad-1_2/devguide/configuration.html
Maybe more importantly, I'm not sure, but I think the reason it exists
is to avoid rerunning the tags? Is rerunning tags as much of an issue
with facelets? If not, maybe we should just say to move to facelets in 2.0.
Thanks,
Gabrielle
Re: [Trinidad] remove application view cache in Trinidad 2?
Posted by Matthias Wessendorf <ma...@apache.org>.
+1
On Thu, Dec 3, 2009 at 9:47 PM, Gabrielle Crawford
<ga...@oracle.com> wrote:
> thanks!
>
> https://issues.apache.org/jira/browse/TRINIDAD-1653
>
> Gabrielle
>
> Martin Koci wrote:
>>
>> Max Starets píše v Út 01. 12. 2009 v 20:42 -0500:
>>
>>>
>>> Gabrielle,
>>>
>>> I think the main advantage of using application view cache is that state
>>> saving/view root caching is done once for a particular page
>>> within an application (that only applies to pages displayed in response
>>> to a GET request).
>>>
>>> Since we have seen some issues with the current implementation, I would
>>> vote for not supporting application view cache
>>> in Trinidad 2. Partial state saving should make its benefits much less
>>> tangible.
>>
>> Yes, I did some profiling few moths ago before we migrated to JSF 2.0
>> state saving and I can confirm that same very complex view written:
>>
>> - as .jspx + trinidad state saving + trinidad components + application
>> view cache
>> - and as .xhtml + mojarra partial state saving + base JSF components
>> doesn't have performance problem (even no regression with .xhtml) in
>> both cases regarding state saving. +1 for removing application view
>> cache in trinidad 2.0
>>
>> Regards,
>>
>> Martin Kočí
>>
>> If there is demand for this feature in the future,
>>
>>>
>>> we can revisit it and try to address the issues we have seen.
>>>
>>> Max
>>>
>>>
>>>
>>> Gabrielle Crawford wrote:
>>>
>>>>
>>>> Hi,
>>>>
>>>> I'm working on state saving issues in Trinidad 2 (for JSF 2). I'm just
>>>> wondering if we really want to support application view cache going forward.
>>>>
>>>> The application view cache has some limitations that make me wonder how
>>>> commonly it's used, see the doc under "The Application View Cache"
>>>>
>>>> http://myfaces.apache.org/trinidad/trinidad-1_2/devguide/configuration.html
>>>>
>>>> Maybe more importantly, I'm not sure, but I think the reason it exists
>>>> is to avoid rerunning the tags? Is rerunning tags as much of an issue with
>>>> facelets? If not, maybe we should just say to move to facelets in 2.0.
>>>>
>>>> Thanks,
>>>>
>>>> Gabrielle
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>>
>
--
Matthias Wessendorf
blog: http://matthiaswessendorf.wordpress.com/
sessions: http://www.slideshare.net/mwessendorf
twitter: http://twitter.com/mwessendorf
Re: [Trinidad] remove application view cache in Trinidad 2?
Posted by Gabrielle Crawford <ga...@oracle.com>.
thanks!
https://issues.apache.org/jira/browse/TRINIDAD-1653
Gabrielle
Martin Koci wrote:
> Max Starets píše v Út 01. 12. 2009 v 20:42 -0500:
>
>> Gabrielle,
>>
>> I think the main advantage of using application view cache is that state
>> saving/view root caching is done once for a particular page
>> within an application (that only applies to pages displayed in response
>> to a GET request).
>>
>> Since we have seen some issues with the current implementation, I would
>> vote for not supporting application view cache
>> in Trinidad 2. Partial state saving should make its benefits much less
>> tangible.
>>
> Yes, I did some profiling few moths ago before we migrated to JSF 2.0
> state saving and I can confirm that same very complex view written:
>
> - as .jspx + trinidad state saving + trinidad components + application
> view cache
> - and as .xhtml + mojarra partial state saving + base JSF components
>
> doesn't have performance problem (even no regression with .xhtml) in
> both cases regarding state saving. +1 for removing application view
> cache in trinidad 2.0
>
> Regards,
>
> Martin Kočí
>
> If there is demand for this feature in the future,
>
>> we can revisit it and try to address the issues we have seen.
>>
>> Max
>>
>>
>>
>> Gabrielle Crawford wrote:
>>
>>> Hi,
>>>
>>> I'm working on state saving issues in Trinidad 2 (for JSF 2). I'm just
>>> wondering if we really want to support application view cache going
>>> forward.
>>>
>>> The application view cache has some limitations that make me wonder
>>> how commonly it's used, see the doc under "The Application View Cache"
>>> http://myfaces.apache.org/trinidad/trinidad-1_2/devguide/configuration.html
>>>
>>>
>>> Maybe more importantly, I'm not sure, but I think the reason it exists
>>> is to avoid rerunning the tags? Is rerunning tags as much of an issue
>>> with facelets? If not, maybe we should just say to move to facelets in
>>> 2.0.
>>>
>>> Thanks,
>>>
>>> Gabrielle
>>>
>>>
>>>
>>>
>>
>
>
>
Re: [Trinidad] remove application view cache in Trinidad 2?
Posted by Martin Koci <ma...@aura.cz>.
Max Starets píše v Út 01. 12. 2009 v 20:42 -0500:
> Gabrielle,
>
> I think the main advantage of using application view cache is that state
> saving/view root caching is done once for a particular page
> within an application (that only applies to pages displayed in response
> to a GET request).
>
> Since we have seen some issues with the current implementation, I would
> vote for not supporting application view cache
> in Trinidad 2. Partial state saving should make its benefits much less
> tangible.
Yes, I did some profiling few moths ago before we migrated to JSF 2.0
state saving and I can confirm that same very complex view written:
- as .jspx + trinidad state saving + trinidad components + application
view cache
- and as .xhtml + mojarra partial state saving + base JSF components
doesn't have performance problem (even no regression with .xhtml) in
both cases regarding state saving. +1 for removing application view
cache in trinidad 2.0
Regards,
Martin Kočí
If there is demand for this feature in the future,
> we can revisit it and try to address the issues we have seen.
>
> Max
>
>
>
> Gabrielle Crawford wrote:
> > Hi,
> >
> > I'm working on state saving issues in Trinidad 2 (for JSF 2). I'm just
> > wondering if we really want to support application view cache going
> > forward.
> >
> > The application view cache has some limitations that make me wonder
> > how commonly it's used, see the doc under "The Application View Cache"
> > http://myfaces.apache.org/trinidad/trinidad-1_2/devguide/configuration.html
> >
> >
> > Maybe more importantly, I'm not sure, but I think the reason it exists
> > is to avoid rerunning the tags? Is rerunning tags as much of an issue
> > with facelets? If not, maybe we should just say to move to facelets in
> > 2.0.
> >
> > Thanks,
> >
> > Gabrielle
> >
> >
> >
>
>
Re: [Trinidad] remove application view cache in Trinidad 2?
Posted by Max Starets <ma...@oracle.com>.
Gabrielle,
I think the main advantage of using application view cache is that state
saving/view root caching is done once for a particular page
within an application (that only applies to pages displayed in response
to a GET request).
Since we have seen some issues with the current implementation, I would
vote for not supporting application view cache
in Trinidad 2. Partial state saving should make its benefits much less
tangible. If there is demand for this feature in the future,
we can revisit it and try to address the issues we have seen.
Max
Gabrielle Crawford wrote:
> Hi,
>
> I'm working on state saving issues in Trinidad 2 (for JSF 2). I'm just
> wondering if we really want to support application view cache going
> forward.
>
> The application view cache has some limitations that make me wonder
> how commonly it's used, see the doc under "The Application View Cache"
> http://myfaces.apache.org/trinidad/trinidad-1_2/devguide/configuration.html
>
>
> Maybe more importantly, I'm not sure, but I think the reason it exists
> is to avoid rerunning the tags? Is rerunning tags as much of an issue
> with facelets? If not, maybe we should just say to move to facelets in
> 2.0.
>
> Thanks,
>
> Gabrielle
>
>
>