tomcat-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Rossbach ...@objektpark.de>
Subject Re: svn commit: r545127 - in /tomcat/container/tc5.5.x: catalina/src/share/org/apache/catalina/core/ApplicationDispatcher.java webapps/docs/changelog.xml
Date Fri, 08 Jun 2007 07:46:01 GMT
Hi Filip and Remy,

Sorry for the very late checkin, but I am little bit busy last two  
weeks. I have some trobule with today ApplicationDispatcher  
implementation :-(
Thanks, Filip that you revert my patch, we can discuss the change  
after tomcat release 5.5.24 is build.

> I see this particular patch is meant to call endAccess when there's  
> an exception invoking the request dispatcher, which could cause  
> problem in cross context. Well, maybe, because I don't see why a  
> single try/finally wouldn't do the same thing, and it does some  
> funky stuff like move some method calls.
Yes, this is the problem!

Remy, I very happy to see a simpler fix.

Peter

Am 07.06.2007 um 14:46 schrieb Remy Maucherat:

> Filip Hanik - Dev Lists wrote:
>> Remy Maucherat wrote:
>>> pero@apache.org wrote:
>>>> Author: pero
>>>> Date: Thu Jun  7 02:33:59 2007
>>>> New Revision: 545127
>>>>
>>>> URL: http://svn.apache.org/viewvc?view=rev&rev=545127
>>>> Log:
>>>> Fix correct ApplicationDispatcher forward/include handling after  
>>>> an exception is thrown. This patch fix a memory leak
>>>> as <code>STRICT_SERVLET_COMPLIANCE</code> system property is
 
>>>> enabled and that cluster  crossContext session replication
>>>> working correct after an exception is thrown at a  
>>>> RequestDispatcher.forward/include call.
>>>> But I don't really like this double try/catch implementation ;-(
>>>
>>> Yes, whatever you are trying to fix, this sort of last minute  
>>> patch is not reasonable to me. -1.
>> Peter, I need a response/action from you, as I am supposed to tag  
>> and generate our candidate binaries today
>
> I see this particular patch is meant to call endAccess when there's  
> an exception invoking the request dispatcher, which could cause  
> problem in cross context. Well, maybe, because I don't see why a  
> single try/finally wouldn't do the same thing, and it does some  
> funky stuff like move some method calls.
>
> This sort of last minute patch is bad practice :( I did it in the  
> past, and you usually have "good" reasons for it, but it should be  
> explicitly forbidden in favor of proposing delaying the release  
> (which may be accepted or not).
>
> Rémy
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
> For additional commands, e-mail: dev-help@tomcat.apache.org
>
>


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org


Mime
View raw message