You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lenya.apache.org by Michael Wechner <mi...@wyona.com> on 2004/12/20 15:14:25 UTC
loading sitetree of default publication and clustering
Maybe we can use XMLHTTP for improving the incremental loading of the
sitetree of the default publication and also make one step into running
the default publication on a cluster.
Also see
http://www.cmswatch.com/News/Article/?372
resp.
http://www.google.com/webhp?complete=1&hl=en
Thanks
Michi
--
Michael Wechner
Wyona Inc. - Open Source Content Management - Apache Lenya
http://www.wyona.com http://lenya.apache.org
michael.wechner@wyona.com michi@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Andreas Hartmann <an...@apache.org>.
Christian Stocker wrote:
[...]
>> Sounds useful, but IIRC XML HTTP is supported only by a few
>> browsers, isn't it?
>
>
> IE >= 5.0, Mozilla >= 1.0, Safari >= 1.2, Konqueror in the latest
> versions (AFAIK) and Opera also in the latest betas... That's not only a
> few, IMHO ;)
Good to know. Thanks a lot :)
-- Andreas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Christian Stocker <ch...@bitflux.ch>.
On 21.12.2004 10:08 Uhr, Andreas Hartmann wrote:
> Michael Wechner wrote:
>
> [...]
>
>> re the cluster we definitely need a repository layer which is able to
>> connect to a repository which supports transactions and the sitetree
>> needs to reside in there.
>>
>> Also the sitetree shouldn't be loaded anymore as a whole into memory.
>
>
> Wouldn't that be a concern of the repository implementation?
>
>> Re the sitetree navigation within the "Site/Info" area which should use
>> XML HTTP or something similar to load it incrementally and not as a
>> whole, otherwise we will have performance problems resp. memory
>> problems on the client side as well.
>
>
> Sounds useful, but IIRC XML HTTP is supported only by a few
> browsers, isn't it?
IE >= 5.0, Mozilla >= 1.0, Safari >= 1.2, Konqueror in the latest
versions (AFAIK) and Opera also in the latest betas... That's not only a
few, IMHO ;)
chregu
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Michael Wechner <mi...@wyona.com>.
Antonio Gallardo wrote:
>Hi Michi:
>
>Warning: I am just thinking loudly :-D
>
>
sure :-)
>I understand the repository layer is a must. I am thinking in some
>implementations of it.
>
>
I would prefer the filesystem as default, which should be possible with
Jackrabbit
>
>Is a good idea to use a database for this task?
>
you mean for the sitetree?
> That way we can use the
>SQL power to retrieve the data. A clear candidate as DB Engine could be
>HSQLDB. It is shipped with Cocoon. Or Derby. ;-)
>
>
agreed
>For compatibility, we can write an Ant task to migrate the sitetree of
>current applications to the database.
>
>
I guess this can be done quickly as soon as we will have the actual
implementation ;-)
All the best
Michi
>WDYT?
>
>Best Regards,
>
>Antonio Gallardo.
>
>On Mar, 21 de Diciembre de 2004, 5:32, Michael Wechner dijo:
>
>
>>Andreas Hartmann wrote:
>>
>>
>>
>>>Michael Wechner wrote:
>>>
>>>[...]
>>>
>>>
>>>
>>>>re the cluster we definitely need a repository layer which is able to
>>>>connect to a repository which supports transactions and the sitetree
>>>>needs to reside in there.
>>>>
>>>>Also the sitetree shouldn't be loaded anymore as a whole into memory.
>>>>
>>>>
>>>Wouldn't that be a concern of the repository implementation?
>>>
>>>
>>well, it should reside and being manipulated within the repository in
>>order to make use of the transactions.
>>
>>Of course one might want to consider a shared memory based repository
>>for performance reasons, but I guess even if there are a lot of
>>documents (e.g. a million) contained there won't be that many
>>modifications in the end.
>>
>>
>>
>>>This is certainly useful, but the whole repository thing is quite
>>>a lot of work. Can we break it up in smaller steps?
>>>
>>>
>>I guess we can, but need to do some more thinking and playing
>>
>>
>>
>>>Maybe implementing the topic map would be a good start, it might
>>>bring up further requirements to the site management component API.
>>>
>>>
>>agreed
>>
>>Michi
>>
>>
>>
>>>-- Andreas
>>>
>>>
>>>---------------------------------------------------------------------
>>>To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
>>>For additional commands, e-mail: dev-help@lenya.apache.org
>>>Apache Lenya Project http://lenya.apache.org
>>>
>>>
>>>
>>>
>>--
>>Michael Wechner
>>Wyona Inc. - Open Source Content Management - Apache Lenya
>>http://www.wyona.com http://lenya.apache.org
>>michael.wechner@wyona.com michi@apache.org
>>
>>
>>---------------------------------------------------------------------
>>To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
>>For additional commands, e-mail: dev-help@lenya.apache.org
>>Apache Lenya Project http://lenya.apache.org
>>
>>
>>
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
>For additional commands, e-mail: dev-help@lenya.apache.org
>Apache Lenya Project http://lenya.apache.org
>
>
>
>
--
Michael Wechner
Wyona Inc. - Open Source Content Management - Apache Lenya
http://www.wyona.com http://lenya.apache.org
michael.wechner@wyona.com michi@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Antonio Gallardo <ag...@agssa.net>.
Hi Michi:
Warning: I am just thinking loudly :-D
I understand the repository layer is a must. I am thinking in some
implementations of it.
Is a good idea to use a database for this task? That way we can use the
SQL power to retrieve the data. A clear candidate as DB Engine could be
HSQLDB. It is shipped with Cocoon. Or Derby. ;-)
For compatibility, we can write an Ant task to migrate the sitetree of
current applications to the database.
WDYT?
Best Regards,
Antonio Gallardo.
On Mar, 21 de Diciembre de 2004, 5:32, Michael Wechner dijo:
> Andreas Hartmann wrote:
>
>> Michael Wechner wrote:
>>
>> [...]
>>
>>> re the cluster we definitely need a repository layer which is able to
>>> connect to a repository which supports transactions and the sitetree
>>> needs to reside in there.
>>>
>>> Also the sitetree shouldn't be loaded anymore as a whole into memory.
>>
>>
>> Wouldn't that be a concern of the repository implementation?
>
>
> well, it should reside and being manipulated within the repository in
> order to make use of the transactions.
>
> Of course one might want to consider a shared memory based repository
> for performance reasons, but I guess even if there are a lot of
> documents (e.g. a million) contained there won't be that many
> modifications in the end.
>
>>
>> This is certainly useful, but the whole repository thing is quite
>> a lot of work. Can we break it up in smaller steps?
>
>
> I guess we can, but need to do some more thinking and playing
>
>>
>> Maybe implementing the topic map would be a good start, it might
>> bring up further requirements to the site management component API.
>
>
> agreed
>
> Michi
>
>>
>> -- Andreas
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
>> For additional commands, e-mail: dev-help@lenya.apache.org
>> Apache Lenya Project http://lenya.apache.org
>>
>>
>
>
> --
> Michael Wechner
> Wyona Inc. - Open Source Content Management - Apache Lenya
> http://www.wyona.com http://lenya.apache.org
> michael.wechner@wyona.com michi@apache.org
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
> For additional commands, e-mail: dev-help@lenya.apache.org
> Apache Lenya Project http://lenya.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Andreas Hartmann <an...@apache.org>.
Michael Wechner wrote:
> Andreas Hartmann wrote:
>
>> To provide a starting point, I added the outline of the
>> TopicMapSiteManager.
>
>
>
> within the trunk?
Yes, org.apache.lenya.cms.site.topicmap.TopicMapSiteManager
-- Andreas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Michael Wechner <mi...@wyona.com>.
Andreas Hartmann wrote:
> To provide a starting point, I added the outline of the
> TopicMapSiteManager.
within the trunk?
>
> -- Andreas
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
> For additional commands, e-mail: dev-help@lenya.apache.org
> Apache Lenya Project http://lenya.apache.org
>
>
--
Michael Wechner
Wyona Inc. - Open Source Content Management - Apache Lenya
http://www.wyona.com http://lenya.apache.org
michael.wechner@wyona.com michi@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Andreas Hartmann <an...@apache.org>.
Michael Wechner wrote:
> Andreas Hartmann wrote:
[...]
>> Maybe implementing the topic map would be a good start, it might
>> bring up further requirements to the site management component API.
To provide a starting point, I added the outline of the
TopicMapSiteManager.
-- Andreas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Michael Wechner <mi...@wyona.com>.
Andreas Hartmann wrote:
> Michael Wechner wrote:
>
> [...]
>
>> re the cluster we definitely need a repository layer which is able to
>> connect to a repository which supports transactions and the sitetree
>> needs to reside in there.
>>
>> Also the sitetree shouldn't be loaded anymore as a whole into memory.
>
>
> Wouldn't that be a concern of the repository implementation?
well, it should reside and being manipulated within the repository in
order to make use of the transactions.
Of course one might want to consider a shared memory based repository
for performance reasons, but I guess even if there are a lot of
documents (e.g. a million) contained there won't be that many
modifications in the end.
>
> This is certainly useful, but the whole repository thing is quite
> a lot of work. Can we break it up in smaller steps?
I guess we can, but need to do some more thinking and playing
>
> Maybe implementing the topic map would be a good start, it might
> bring up further requirements to the site management component API.
agreed
Michi
>
> -- Andreas
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
> For additional commands, e-mail: dev-help@lenya.apache.org
> Apache Lenya Project http://lenya.apache.org
>
>
--
Michael Wechner
Wyona Inc. - Open Source Content Management - Apache Lenya
http://www.wyona.com http://lenya.apache.org
michael.wechner@wyona.com michi@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Andreas Hartmann <an...@apache.org>.
Michael Wechner wrote:
[...]
> re the cluster we definitely need a repository layer which is able to
> connect to a repository which supports transactions and the sitetree
> needs to reside in there.
>
> Also the sitetree shouldn't be loaded anymore as a whole into memory.
Wouldn't that be a concern of the repository implementation?
> Re the sitetree navigation within the "Site/Info" area which should use
> XML HTTP or something similar to load it incrementally and not as a
> whole, otherwise we will have performance problems resp. memory problems
> on the client side as well.
Sounds useful, but IIRC XML HTTP is supported only by a few
browsers, isn't it?
> It probably makes sense to take a look at other systems (e.g. Daisy,
> Hippo, Magnolia, etc.) how (or if) they solved the problem.
>
> Also I think it makes sense to implement the Topic Map as another
> implementation during refactoring.
This is certainly useful, but the whole repository thing is quite
a lot of work. Can we break it up in smaller steps?
Maybe implementing the topic map would be a good start, it might
bring up further requirements to the site management component API.
-- Andreas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Jean Pierre LeJacq <jp...@quoininc.com>.
On Wed, 22 Dec 2004, Gregor J. Rothfuss wrote:
> Jean Pierre LeJacq wrote:
>
> > Yes. Its reuse at a design/architecture level. It should save us
> > time in development, documentation (since the API is already
> > published), and reduce the learning curve for new users since its
> > hopefully a known API.
>
> would you be willing to take this on? another API to look at for this
> might be jsr 170, and how it deals with nodes. may make more sense to
> use than topic maps.
Sure. I'll try to summarize the difference between the two.
--
JP
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by "Gregor J. Rothfuss" <gr...@apache.org>.
Jean Pierre LeJacq wrote:
> Yes. Its reuse at a design/architecture level. It should save us
> time in development, documentation (since the API is already
> published), and reduce the learning curve for new users since its
> hopefully a known API.
would you be willing to take this on? another API to look at for this
might be jsr 170, and how it deals with nodes. may make more sense to
use than topic maps.
--
Gregor J. Rothfuss
COO, Wyona Content Management Solutions http://wyona.com
Apache Lenya http://lenya.apache.org
gregor.rothfuss@wyona.com gregor@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Jean Pierre LeJacq <jp...@quoininc.com>.
On Wed, 22 Dec 2004, Andreas Hartmann wrote:
> Jean Pierre LeJacq wrote:
> >
> > Having a generic interface makes sense for a site map component.
> > But I do think that basing the API for this generic interface on
> > topic maps makes the most sense. It could be a subset, but a
> > consistent subset.
>
> That sounds very interesting. Up to now, the site management
> development in Lenya is more or less trial-and-error, it would be
> nice to implement a standard solution.
Yes. Its reuse at a design/architecture level. It should save us
time in development, documentation (since the API is already
published), and reduce the learning curve for new users since its
hopefully a known API.
--
JP
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Andreas Hartmann <an...@apache.org>.
Jean Pierre LeJacq wrote:
> On Wed, 22 Dec 2004, Andreas Hartmann wrote:
[...]
>>But the core code should rely only on the generic interface.
>>This allows to use custom site management concepts in publications
>>which are not included in the core.
>
>
> Having a generic interface makes sense for a site map component.
> But I do think that basing the API for this generic interface on
> topic maps makes the most sense. It could be a subset, but a
> consistent subset.
>
> As an example, consider the current lenya sitetree classes. It has
> its own traversal api which is incomplete. It would make more sense
> to base the traversal api on a standard such as the DOM traversal
> and range specification
> (http://www.w3.org/TR/2000/REC-DOM-Level-2-Traversal-Range-20001113/).
That sounds very interesting. Up to now, the site management
development in Lenya is more or less trial-and-error, it would
be nice to implement a standard solution.
-- Andreas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Jean Pierre LeJacq <jp...@quoininc.com>.
On Wed, 22 Dec 2004, Andreas Hartmann wrote:
> Jean Pierre LeJacq wrote:
> > On Tue, 21 Dec 2004, Gregor J. Rothfuss wrote:
> >
> >
> >>i see three areas that can be tackled independently:
> >>
> >>* integrating the repository layer (and make the sitetree object talk to
> >>that)
> >>* adding topic maps as a storage backend for the sitetree object
> >>* refactoring the sitetree js to use incremental updates
> >
> >
> > I'd go a little further and say that the sitetree object should
> > provide the topic map API.
>
> I think that will lead too far (can you say so in English? :) ).
>
> We can't support the details of all possible site management
> concepts in the core API.
>
> In Lenya 1.4, there is a generic site management component which
> is implemented by a TreeSiteManager or a TopicMapSiteManager.
>
> If publication code depends on a certain implementation, it has
> to do an explicit cast and throw an exception if the cast fails.
> Maybe it makes sense to use something like ECM's component
> selection mechanism.
>
> But the core code should rely only on the generic interface.
> This allows to use custom site management concepts in publications
> which are not included in the core.
Having a generic interface makes sense for a site map component.
But I do think that basing the API for this generic interface on
topic maps makes the most sense. It could be a subset, but a
consistent subset.
As an example, consider the current lenya sitetree classes. It has
its own traversal api which is incomplete. It would make more sense
to base the traversal api on a standard such as the DOM traversal
and range specification
(http://www.w3.org/TR/2000/REC-DOM-Level-2-Traversal-Range-20001113/).
--
JP
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Andreas Hartmann <an...@apache.org>.
Jean Pierre LeJacq wrote:
> On Tue, 21 Dec 2004, Gregor J. Rothfuss wrote:
>
>
>>i see three areas that can be tackled independently:
>>
>>* integrating the repository layer (and make the sitetree object talk to
>>that)
>>* adding topic maps as a storage backend for the sitetree object
>>* refactoring the sitetree js to use incremental updates
>
>
> I'd go a little further and say that the sitetree object should
> provide the topic map API.
I think that will lead too far (can you say so in English? :) ).
We can't support the details of all possible site management
concepts in the core API.
In Lenya 1.4, there is a generic site management component which
is implemented by a TreeSiteManager or a TopicMapSiteManager.
If publication code depends on a certain implementation, it has
to do an explicit cast and throw an exception if the cast fails.
Maybe it makes sense to use something like ECM's component
selection mechanism.
But the core code should rely only on the generic interface.
This allows to use custom site management concepts in publications
which are not included in the core.
The alternative would be to say "Lenya supports site trees,
topic maps, and nothing else". IMO this would violate the framework
approach.
-- Andreas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Jean Pierre LeJacq <jp...@quoininc.com>.
On Tue, 21 Dec 2004, Gregor J. Rothfuss wrote:
> i see three areas that can be tackled independently:
>
> * integrating the repository layer (and make the sitetree object talk to
> that)
> * adding topic maps as a storage backend for the sitetree object
> * refactoring the sitetree js to use incremental updates
I'd go a little further and say that the sitetree object should
provide the topic map API.
--
JP
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by "Gregor J. Rothfuss" <gr...@apache.org>.
Michael Wechner wrote:
> re the cluster we definitely need a repository layer which is able to
> connect to a repository which supports transactions and the sitetree
> needs to reside in there.
ok
> Also the sitetree shouldn't be loaded anymore as a whole into memory.
yes. for clarification, this concerns the server-side sitetree object.
> Re the sitetree navigation within the "Site/Info" area which should use
> XML HTTP or something similar to load it incrementally and not as a
> whole, otherwise we will have performance problems resp. memory problems
> on the client side as well.
ok. to do that requires some time with the javascript we use to draw the
sitetree (to make it XMLHTTP aware, etc)
> It probably makes sense to take a look at other systems (e.g. Daisy,
> Hippo, Magnolia, etc.) how (or if) they solved the problem.
sure
> Also I think it makes sense to implement the Topic Map as another
> implementation during refactoring.
i see three areas that can be tackled independently:
* integrating the repository layer (and make the sitetree object talk to
that)
* adding topic maps as a storage backend for the sitetree object
* refactoring the sitetree js to use incremental updates
lets file bugs for those.
--
Gregor J. Rothfuss
COO, Wyona Content Management Solutions http://wyona.com
Apache Lenya http://lenya.apache.org
gregor.rothfuss@wyona.com gregor@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Michael Wechner <mi...@wyona.com>.
Steven Noels wrote:
> On 21 Dec 2004, at 10:48, Michael Wechner wrote:
>
>> It probably makes sense to take a look at other systems (e.g. Daisy,
>> Hippo, Magnolia, etc.) how (or if) they solved the problem.
>
>
> We prepare the XML for the entire page (document content, fields,
> navtree) on the backend and send it as one big chunk to Cocoon. We do
> some caching of the navtree (as objects), but nothing else - it's fast
> enough,
what experiences did you make with re number of nodes, e.g. a million?
> and since our navtree nodes need to be checked for ACL access rights
> anyhow, caching won't help that much.
ok, but what about an incremental solution, meaning one only provides
resp. check the nodes which the user is interested in (similar to a thin
client solution or the "file system")?
> Also, I recall Bruno taking a look at the Plone sitemap thing lately.
I guess Plone is making use of Zope DB, right? Well, it's a common problem
and the architecture should be independent of the implementation language.
I guess I need to extend the question to OSCOM and beyond ;-)
thanks for your help on this
Michi
>
> Cheers,
>
> </Steven>
--
Michael Wechner
Wyona Inc. - Open Source Content Management - Apache Lenya
http://www.wyona.com http://lenya.apache.org
michael.wechner@wyona.com michi@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Steven Noels <st...@outerthought.org>.
On 21 Dec 2004, at 10:48, Michael Wechner wrote:
> It probably makes sense to take a look at other systems (e.g. Daisy,
> Hippo, Magnolia, etc.) how (or if) they solved the problem.
We prepare the XML for the entire page (document content, fields,
navtree) on the backend and send it as one big chunk to Cocoon. We do
some caching of the navtree (as objects), but nothing else - it's fast
enough, and since our navtree nodes need to be checked for ACL access
rights anyhow, caching won't help that much. Also, I recall Bruno
taking a look at the Plone sitemap thing lately.
Cheers,
</Steven>
--
Steven Noels http://outerthought.org/
Outerthought - Open Source Java & XML An Orixo Member
Read my weblog at http://blogs.cocoondev.org/stevenn/
stevenn at outerthought.org stevenn at apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Jean Pierre LeJacq <jp...@quoininc.com>.
On Tue, 21 Dec 2004, Michael Wechner wrote:
> Also I think it makes sense to implement the Topic Map as another
> implementation during refactoring.
+1
This is definitely the way to go.
--
JP
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by Michael Wechner <mi...@wyona.com>.
Gregor J. Rothfuss wrote:
> Michael Wechner wrote:
>
>> Maybe we can use XMLHTTP for improving the incremental loading of the
>> sitetree of the default publication and also make one step into
>> running the default publication on a cluster.
>
>
> i was thinking something similar this weekend, and looked around for
> good XMLHTTP widgets, havent found any so far.
>
> can you explain a bit more what the problems on the cluster are?
re the cluster we definitely need a repository layer which is able to
connect to a repository which supports transactions and the sitetree
needs to reside in there.
Also the sitetree shouldn't be loaded anymore as a whole into memory.
Re the sitetree navigation within the "Site/Info" area which should use
XML HTTP or something similar to load it incrementally and not as a
whole, otherwise we will have performance problems resp. memory problems
on the client side as well.
It probably makes sense to take a look at other systems (e.g. Daisy,
Hippo, Magnolia, etc.) how (or if) they solved the problem.
Also I think it makes sense to implement the Topic Map as another
implementation during refactoring.
Michi
>
>
>
--
Michael Wechner
Wyona Inc. - Open Source Content Management - Apache Lenya
http://www.wyona.com http://lenya.apache.org
michael.wechner@wyona.com michi@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org
Re: loading sitetree of default publication and clustering
Posted by "Gregor J. Rothfuss" <gr...@apache.org>.
Michael Wechner wrote:
> Maybe we can use XMLHTTP for improving the incremental loading of the
> sitetree of the default publication and also make one step into running
> the default publication on a cluster.
i was thinking something similar this weekend, and looked around for
good XMLHTTP widgets, havent found any so far.
can you explain a bit more what the problems on the cluster are?
--
Gregor J. Rothfuss
COO, Wyona Content Management Solutions http://wyona.com
Apache Lenya http://lenya.apache.org
gregor.rothfuss@wyona.com gregor@apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lenya.apache.org
For additional commands, e-mail: dev-help@lenya.apache.org
Apache Lenya Project http://lenya.apache.org