|
|
|
Re: [jaxws] error creating web service from pojo [message #494956 is a reply to message #494195] |
Tue, 03 November 2009 11:39   |
Oisin Hurley Messages: 204 Registered: July 2009 |
Senior Member |
|
|
Hi Jeff,
Thanks for the positive comment What do you have in mind for the visibility stuff? The JAXWS code effectively calls out to a tool within the runtime's installation and delegates the generation to it entirely. What you are seeing here is a FAIL on the attempt to actually launch the tool. Of course, behind the scenes, we do build a classpath with which the tool must be launched, and it's most likely that classpath
that is ending up as overly long for poor old Windows. But I'm sure the Spring part updates the classpath too, so that adds more to mix. And if the file locations of the JARs have long pathnames, well then the classpath will be a monster, character-count wise.
So I'm not too sure that this is 'fixable' as such, since the classpath will be the target of other pieces of technology. But what might make sense is some kind of a warning at some stage, so that the developer doesn't get startled with this kind of cryptic fail. It's always worth putting in a bugzilla, just for tracking purposes, to see if anyone comes up with any bright ideas.
cheers
--oh
|
|
|
|
|
|
Re: [jaxws] error creating web service from pojo [message #570653 is a reply to message #494195] |
Tue, 03 November 2009 11:39   |
Oisin Hurley Messages: 204 Registered: July 2009 |
Senior Member |
|
|
Hi Jeff,
Thanks for the positive comment :) What do you have in mind for the visibility stuff? The JAXWS code effectively calls out to a tool within the runtime's installation and delegates the generation to it entirely. What you are seeing here is a FAIL on the attempt to actually launch the tool. Of course, behind the scenes, we do build a classpath with which the tool must be launched, and it's most likely that classpath
that is ending up as overly long for poor old Windows. But I'm sure the Spring part updates the classpath too, so that adds more to mix. And if the file locations of the JARs have long pathnames, well then the classpath will be a monster, character-count wise.
So I'm not too sure that this is 'fixable' as such, since the classpath will be the target of other pieces of technology. But what might make sense is some kind of a warning at some stage, so that the developer doesn't get startled with this kind of cryptic fail. It's always worth putting in a bugzilla, just for tracking purposes, to see if anyone comes up with any bright ideas.
cheers
--oh
|
|
|
|
|
|
Powered by
FUDForum. Page generated in 0.02785 seconds