Jeremias Maerki <dev <at> jeremias-maerki.ch> writes:
> See http://xmlgraphics.apache.org/fop/0.20.5/graphics.html#caching on
> how to reset the image cache in FOP 0.20.5 between rendering runs. Note
> that this function is not available anymore in the latest version since
> there the cache works differently. It has much better memory reclaiming
> abilities (using weak references) so a cache reset method should not
> be necessary if you use unique URLs. 

Hi

Thank you very much for the fast answer and your advice how to reset
the cache. That was very helpfull.

Anyway I have to say, that I think that this static image cache is 
not such a good solution (even not with weak references).

> so a cache reset method should not be necessary if you use unique URLs. 

Caching only with respect to the urls without considering if the url 
content has changed inbetween (e.g. modification date) seems not very 
proper to me. 
If done so, the user should at least have a chance to decide whether to
bypass this cache or to empty it. If I understood you right, in the latest
version it is not even possible anymore to empty the cache?

> On the other side, if anyone submits a patch to retrofit this method
> we will surely consider adding it. 

If you give me some hint how to access the cache and if I find some time
I would try to do it. But for the moment I think that I simply stick 
to the older version.

> However, I'd prefer an extension attribute on fo:external-graphic
> that signals FOP to bypass the image cache entirely.

I don't think that changing the fo:external-graphic specification to the
needs of FOP is something that is in your power or is it? 
I don't agree that this attribute would be a good solution. Since a 
writer of a FO-file usualy does not want to concern about cache bypassing.
FOP should better use the correct (cached or reloaded) picture in any case.

Those are just my thoughts about it. In any case I am very thankful for your
help and I appreciate your work. Thank you very much.

Greetz,
rolf


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to