Speeding up Cocoon start at runtime?

2003-12-20 Thread Antonio Gallardo
Hi: I am learning while reviewing the Cocoon build system. I found an interesting feature that we can take advantage while creating the jars: http://java.sun.com/j2se/1.3/docs/guide/jar/jar.html#JAR%20Index Since we are supporting VM 1.3+ then is OK to turn this feature on? The change is very

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Bertrand Delacretaz
Le Samedi, 20 déc 2003, à 09:59 Europe/Zurich, Antonio Gallardo a écrit : http://java.sun.com/j2se/1.3/docs/guide/jar/jar.html#JAR%20Index Sounds interesting, and as it is based on an extra file in jars (IIUC) I don't see how it could cause any harm. Did you do any measurements with and

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Antonio Gallardo
Bertrand Delacretaz dijo: Le Samedi, 20 déc 2003, à 09:59 Europe/Zurich, Antonio Gallardo a écrit : http://java.sun.com/j2se/1.3/docs/guide/jar/jar.html#JAR%20Index Sounds interesting, and as it is based on an extra file in jars (IIUC) I don't see how it could cause any harm. Did you do

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Antonio Gallardo
Bertrand Delacretaz dijo: Le Samedi, 20 déc 2003, à 09:59 Europe/Zurich, Antonio Gallardo a écrit : http://java.sun.com/j2se/1.3/docs/guide/jar/jar.html#JAR%20Index Sounds interesting, and as it is based on an extra file in jars (IIUC) I don't see how it could cause any harm. Hi Bertrand:

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Bertrand Delacretaz
Le Samedi, 20 déc 2003, à 11:17 Europe/Zurich, Antonio Gallardo a écrit : ..The change is already in the CVS. Please check it. build works, cocoon servlet works. Didn't do any performance check. Ciao, Bertrand

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Jorg Heymans
the IBM 1.3 jre seems to stumble over this feature. http://blogs.codehaus.org/people/avasseur/archives/2003_09.html#000139_aspectwerkz_has_support_for_ibm Is the IBM runtime supported by cocoon? jorg Bertrand Delacretaz wrote: Le Samedi, 20 déc 2003, à 09:59 Europe/Zurich, Antonio Gallardo a

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Antonio Gallardo
Jorg Heymans dijo: the IBM 1.3 jre seems to stumble over this feature. http://blogs.codehaus.org/people/avasseur/archives/2003_09.html#000139_aspectwerkz_has_support_for_ibm Is the IBM runtime supported by cocoon? Don't know. Are you using it? If yes, please test the lastest CVS and report

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Jorg Heymans
I am not using it but will have a go at it, downloading now. I guess the websphere installations use IBM jre.. rgds Jorg Antonio Gallardo wrote: Jorg Heymans dijo: the IBM 1.3 jre seems to stumble over this feature.

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Jorg Heymans
I downloaded the wsdk_51 for windows which comes with build 1.3.1, J2RE 1.3.1 IBM Windows 32 build cn131-20030618 Doing build results in errors both for jdk 1.4 and ibm jre 1.3.1 BUILD RESULTS IBM jre mail-compile: Copying 1 file to

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread bernhard huber
hi, snip/ D:\src\cocoon-2.1\src\blocks\mail\java\org\apache\cocoon\mail\MailContentHandlerDelegate.java:384: cannot resolve symbol symbol : method getDeletedMessageCount () location: class javax.mail.Folder addAttribute(deleted-messages, String.valueOf(folder.getDelet

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread Jorg Heymans
ok i removed the mail.jar from IBM and now the build is successful. I did get an out-of-memory exception but the build continued so I guess all is OK. Note that the JDK 1.4 build is still failing. IBM jre output: == Constructing Javadoc information... javadoc:

Re: Speeding up Cocoon start at runtime?

2003-12-20 Thread bernhard huber
hi, as far as i understand your posted log: javadoc fails due to memory shortage, you might want to exclude javadoc generation in the block.properties, or local.block.properties, failing build test is due to missing midi-jars, again edit the block.properties, or local.block.properties disabling