cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jeff Martin" <>
Subject Re: RunTimeException in XSPProcessor
Date Thu, 03 Feb 2000 20:03:59 GMT
Sounds like the sort of thing that would cause this problem. It's possible
that the oracle DB drivers on are using a URL handler. It's working on my NT
box where it's the only thing running and not when I have multiple servlets
on the linux box.

As far as the exact problem I'm getting (There maybe other places where this
could become a problem). It looks like it maybe possible to remove the need
to use URL for the libraries loaded from the checking the
protocol before passing it on to the URL object for handling whould allow
the internal resources to be loaded without using the factory.

Something like this

InputStream is=null;
    is = this.getClass().getResourceAsStream(location);
    is = new URL(location).openStream();

-----Original Message-----
From: Donald Ball <>
To: <>
Date: 03 February 2000 03:16
Subject: Re: RunTimeException in XSPProcessor

>On Wed, 2 Feb 2000, Jeff Martin wrote:
>> I'm getting a RunTimeException everytime I try to access an XML page
>> under Linux but not NT. Can anyone shed and light on this.
>> I've been developing some pages using Cocoon 1.6.1 under NT JDK1.2.2.
>> I've just started to create a new enviroment on Linux using Blackdown
>> JDK1.2.2 (I've tried the sun release to) when I try and use any of the
>> example xml pages I get and RunTimeException Error loading logicsheet:
>> resource://org/apache/cocoon/processor/xsp/library/java/cookie.xsl
>> from org.apache.cocoon.processor.xsp.XSPProcessor.
>> I've had a poke around the code and this is the bit that causing me
>> problems. When the url is created using
>> resource://org/apache/cocoon/processor/xsp/library/java/cookie.xsl
>> this is throwing an exception because it doesn't know how to handle
>> the url. I wasn't aware the there was a resource protocol handler but
>> as this code works fine on NT there must be. I have managed to put a
>> hack in which uses this.getClass().getResourceAsStream(location) after
>> choping the resource: of the front of location.
>> Anyone got any comments.
>This may have to do with a serious flaw in the way the JVM works with
>URLFactories. I don't understand all of the details, but the gist is that
>only one URLFactory may exist in a JVM, whih essentially means that only
>one application may add support for additional URLs to the JVM. Is it
>possible that, on your linux server, another servlet has already created a
>As I understand it, Stefano, et. al. are _actively_ looking for an
>alternate solution. Do you have any suggestions?
>- donald

View raw message