Re: Removing sitemap's check-reload attribute

2004-06-04 Thread Joerg Heinicke
On 04.06.2004 15:14, Sylvain Wallez wrote:
Working on the TreeProcessor, I noticed check-reload that exists on 
map:mount whose existence I nearly forgot about, and was wondering 
about its actual usefulness, furthermore considering that it defaults to 
true (i.e. reload when needed)

There are lots and lots of places in Cocoon where automatic reload 
occurs with no way to disable it: XSLT, XSP, JXTG, CForms to name a few.

So I propose to remove this feature, and have automatic sitemap reload 
always be turned on, as everywhere else.

WDYT?
+1 I never switched it off.
IIRC there is also such a setting in cocoon.xconf for the root sitemap. 
Does it make sense to switch reload on or off for each subsitemap 
individually or is the one setting in cocoon.xconf satisfying enough for 
all sitemaps?

Joerg


Re: Removing sitemap's check-reload attribute

2004-06-04 Thread Vadim Gritsenko
Sylvain Wallez wrote:
Working on the TreeProcessor, I noticed check-reload that exists on 
map:mount whose existence I nearly forgot about, and was wondering 
about its actual usefulness, furthermore considering that it defaults 
to true (i.e. reload when needed)

There are lots and lots of places in Cocoon where automatic reload 
occurs with no way to disable it: XSLT, XSP, JXTG, CForms to name a few.

Actually, you can turn off XSP reloading.
PS -1 on removing in cocoon.xconf, 0 on removing in map:mount
Vadim


Re: Removing sitemap's check-reload attribute

2004-06-04 Thread Ugo Cei
Sylvain Wallez wrote:
So I propose to remove this feature, and have automatic sitemap reload 
always be turned on, as everywhere else.

WDYT?
+1
Ugo


Re: Removing sitemap's check-reload attribute

2004-06-04 Thread Stefano Mazzocchi
Sylvain Wallez wrote:
Working on the TreeProcessor, I noticed check-reload that exists on 
map:mount whose existence I nearly forgot about, and was wondering 
about its actual usefulness, furthermore considering that it defaults to 
true (i.e. reload when needed)

There are lots and lots of places in Cocoon where automatic reload 
occurs with no way to disable it: XSLT, XSP, JXTG, CForms to name a few.

So I propose to remove this feature, and have automatic sitemap reload 
always be turned on, as everywhere else.

WDYT?
I like the fact that auto-reloading should be consistent and global, but 
I would also like to have the ability to turn it off, especially in 
production environments because this removes some performance overhead 
in the checking (even if it's asynchronous).

Comments?
--
Stefano.


smime.p7s
Description: S/MIME Cryptographic Signature


Re: Removing sitemap's check-reload attribute

2004-06-04 Thread Sylvain Wallez
Stefano Mazzocchi wrote:
Sylvain Wallez wrote:
Working on the TreeProcessor, I noticed check-reload that exists on 
map:mount whose existence I nearly forgot about, and was wondering 
about its actual usefulness, furthermore considering that it defaults 
to true (i.e. reload when needed)

There are lots and lots of places in Cocoon where automatic reload 
occurs with no way to disable it: XSLT, XSP, JXTG, CForms to name a few.

So I propose to remove this feature, and have automatic sitemap 
reload always be turned on, as everywhere else.

WDYT?

I like the fact that auto-reloading should be consistent and global, 
but I would also like to have the ability to turn it off, especially 
in production environments because this removes some performance 
overhead in the checking (even if it's asynchronous).

Comments?

There were some talks related to this when we discussed store cleanup, 
as reload is often related to compiling the contents of a Source (e.g. 
building a TreeProcessor, a form definition, a stylesheet, etc) and 
caching that compiled form for later use. Caching also possibly meaning 
purging no more used entries, contrarily to what happens today with 
private Maps to store these objects.

There's an initial attempt at this in the scratchpad (util.SourceCache) 
which we may build upon.

That may also be a service provided at a lower level by SourceUtil (but 
then relying on an ugly static setting) or by the SourceResolver itself.

Sylvain
--
Sylvain Wallez  Anyware Technologies
http://www.apache.org/~sylvain   http://www.anyware-tech.com
{ XML, Java, Cocoon, OpenSource }*{ Training, Consulting, Projects }