[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[cross-project-issues-dev] issue with ICU4J on weblogic 8

Hi, Fellow Callisto teams,
 
I am not sure this is a cross team issue, but would like to share it, since we are all moving to ICU4J.
 
After moving to ICU4J in BIRT, we have run into resource loading issue on the Weblogic 8 app server platform, other app server platform works.  Searching on the ICU4J bug log, it seems others have run into similar problem.  Following are the 3 bug entries for ICU in this area...
 
http://dev.icu-project.org/cgi-bin/icu-bugs/incoming?findid=5089

and 
http://dev.icu-project.org/cgi-bin/icu-bugs/config_build?id=4560;expression=weblogic;user=guest
 
and 
http://dev.icu-project.org/cgi-bin/icu-bugs?findid=4346&go=Go
 
They points to resource loading issue withing ICU4J on app servers platforms, specifically weblogic 8.1..
 
Did any of your projects run into this issue?  Is there a work around for this?
 
Thanks
 
Wenfeng  Li
Eclipse BIRT Project PMC

________________________________

From: cross-project-issues-dev-bounces@xxxxxxxxxxx on behalf of Bjorn Freeman-Benson
Sent: Wed 3/1/2006 10:27 AM
To: eclipse.org-planning-council; Cross project issues
Subject: [cross-project-issues-dev] Demonstrated respect


Planning Council members, Cross project mailing list members, 
(especially John Duimovich, Randy Hudson, and Dave Orme),

Some of you have complained about the way I wrote my email yesterday and I accept that criticism and will try to write less inflamatorily in the future. At the same time, however, I may have obscured my real point, so at the risk of making things worse, let me try again:


1.	I was not complaining about the fact that GEF and VE were missing from the Callisto update site (well, I was a little, but that wasn't the real point). I know that in the real world things change - schedules change, staffing changes, priorities change, etc - and that sometimes deadlines just can't be met. Right now I have an auto-reply vacation message on my email that says exactly that, so believe me, I know this. 
2.	What I was complaining about is that the project leadership of those projects hadn't taken the time to communicate to the rest of the Callisto team. 
	

This Callisto Simultaneous Release is hard a problem. A simultaneous release is a hard problem just within a single company where everyone reports to the same VP Engineering. It's an even harder problem in open source where the projects are staffed by (effectively) volunteers and from multiple competing companies. The only way we are going to make this work is to keep all of our colleagues well informed of our status, our progress, our problems, and any potential schedule slips. Just posting to our own project websites or mailing lists isn't good enough - we have to reach out to our colleagues (I'm talking about the collective Callisto team) and actively keeping everyone informed.

If we cannot commit to doing that, Callisto will flop. You know that. I know that. The key to making Callisto work isn't going to be technology - the key is going to be the communication channels that we build between the projects. Frequent, active, accurate, and timely communication. And, as Tyler points out, respectful. Respect for the schedules and dependencies of others which includes letting the rest of us know when you can't make a deadline, meet a requirement, or attend a meeting.

That's what I meant to say,
Bjorn


<<winmail.dat>>