when JK is used, it uses the Channel (network ip IO) way to connect.
Channel is not as forgiving as sockets.
this is the piece of information that maybe missing to understand how
the stream is interrupted.
it is this that causes the broken pipes.
it is not from the client side.
Thanks Erwan!
Jacques
From:
Author: erwan
Date: Sun Jun 13 20:01:15 2010
New Revision: 954295
URL: http://svn.apache.org/viewvc?rev=954295&view=rev
Log:
adding sysinfo messages to bluelight theme
Modified:
ofbiz/trunk/themes/bluelight/includes/header.ftl
Modified: ofbiz/trunk/themes/blue
Hi,
I've just read your conversation and I want to add my two cents:
When I started using Ofbiz I stumbled across the Javolution stuff inside and
was really fascinated
because that time I was interested in object pooling during research for a
chapter of my pattern book.
Javolution seems to be r
[
https://issues.apache.org/jira/browse/OFBIZ-3563?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12878388#action_12878388
]
Jacques Le Roux commented on OFBIZ-3563:
Of course I prefer the second solution :)
[
https://issues.apache.org/jira/browse/OFBIZ-3563?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12878382#action_12878382
]
Scott Gray commented on OFBIZ-3563:
---
product ids are always prepended with a p_ when gene
Thanks Jacques,
I've only ever seen the exception being thrown when Tomcat is fronted with an
Apache web server, my guess is that Tomcat handles prematurely closed
connections differently when it's a direct request. It could be that an output
stream is being used instead of a writer, I don't k
Scott,
OK, I see your point now, and I agree we should revert all anyway.
Something I'd like to do before is reproduce the ClientAbortException locally. So I launched a long process (Artifact Info) and
closed my browser, but the process continued and the error was not reported. I also tried wit
Also, last time I checked, we don't back port "tests" or theories to the
release branches. We back port solid confirmed fixes (well, in an ideal world
that is what happens).
Regards
Scott
On 13/06/2010, at 9:37 PM, Jacques Le Roux wrote:
> I'm just trying this to see what will happen on the d
I would like to see it reverted. It was an unnecessary change and serves no
benefit, BJ may think it helped him but he is wrong.
Like I said on the user list a ClientAbortException is nothing for OFBiz to
worry about and what ever was wrong with the instance that exception had
nothing to do w
I'm just trying this to see what will happen on the demo server. For now I get errors in the log that are not workable. Has it seems
it helped BJ, I want to try and I don't want to put to much time in this.
No theory here, pragmatism. IT is not a sound theory anyway, we all know that,
without te
10 matches
Mail list logo
