Hi Carl,

I added the following section to our faq yesterday:
Is OJB ready for production environments? 
Depends on your production environment. If you want to program an aeroplane
autopilot system you should not use Java at all. (according to the official
disclaimer).

But I assume we are talking about enterprise business applications, aren't
we? And for such applications it's a clear yes. OJB is used in production
application since version 0.5. We have about 6.000 downloads each month (and
growing) and a large user base using it in a wide spectrum of production
scenarios.

We provide a regression test suite for Quality Assurance. You can use this
testsuite to check if OJB works smoothly in your target environment. (see
supported platforms documentation)

We also provide a performance testsuite that compares OJB performance
against native JDBC. This test will give you an impression of the
performance impact OJB will have in your target environment. (see
Performance testsuite documentation)

OJB is also the persistence layer of choice in several books on programming
J2EE based enterprise business systems. (see our links and references
section)



Here are some user testimonials from the user mailing list: 

"We're using OJB in two production applications at the Northwest Alliance
for Computational Science and Engineering (NACSE). One is a data mining
toolset, and the other is a massive National Science Foundation project that
involves huge amounts of data, and about 20 or 25 universities and research
groups like mine.
In fact, I've begun making OJB sort of a de-facto standard for NACSE
java/database development. I've thrown out EJB's for the most part and I've
tried JDO from Castor, but I'm sticking with OJB. Maybe we'll reconsider JDO
when the OJB implementation is more complete." 

"We are planning a November 2003 production deployment with OJB and WE LOVE
IT!! We have been in development on a very data-centric application in the
power industry for about 5 months now and OJB has undoubtedly saved us
countless hours of development time. We have received benefits in the
following areas:
-> Easily adapts to any data model that we've thrown at it. No problems
mapping tables with compound keys, tables mapping polymorphic relationships,
identity columns, etc.
-> Seemlesly switches between target DB platforms. We develop and unit test
on our local workstations with HSQLDB and PostgreSQL, and deploy to DB2
using the Type 4 JDBC driver from IBM. Works great!
-> Makes querying a breeze with the PersistenceBroker API
Overall we have found OJB to be very stable (and we've really tested it out
quite a bit). The only issues we've got outstanding at the moment is support
for connections to multiple databases, but I've noticed in CVS that the OJB
guys are already fixing this for OJB 0.9.9." 

"We've been using it in "production" for a long time now, from about version
0.9.4, I believe. It has been very robust. We don't use all of its features.
We've only see to failures of our persistent store in about 9 months, and
I'm not sure they were due to OJB." 

"So yes, we have made a quite useful mediumsized production website based on
OJB (with JBoss, Jakarta Jetspeed, Jakarta Turbine and Jakarta Jelly, three
Tomcats, OpenSymhony OSCache and for the database MSSQL server, all running
on Win2000.) It is attracting between 600 and 9000 (peak) users a day, and
runs smoothly for extended periods of time. And no, I can not actually show
you the wonders of the editorial interface of the content management system,
because it is hidden behind a firewall. 

I feel OJB is quite useful in production, but you certainly have to know
what you are doing and what you are trying to achieve with it. And there
have been some tricky aspects, but these could be solved by simple
workarounds and small hacks. 

The main thing about OJB is that AFAIK it has an overall clean design, and
it far beats making your own database abstraction layer and
object/relational mapper. We certainly do not use all of it, only the
Persistence Broker parts, so there was less to learn. We love the virtual
proxy and collection proxy concepts, the criteria objects for building
queries, and the nice little hidden features that you find when you start to
learn the system." 

"My Company is building medium to large scale, mission critical applications
(100 - 5.000 concurrent users) for our customers. Our largest customer is
KarstadtQuelle, Europes largest retail company. The next big system that
will go in production (in June) is the new logistics system for the
stationary logistics of Karstadt.
Of course we are using OJB in those Systems! We have several OJB based
systems now in production for over a year. We never had any OJB releated
problems in production.
Most problems we faced during development were related to the learning curve
developers had to face who were new to O/R mapping." 
 

> -----Original Message-----
> From: Sjoquist, Carl [mailto:[EMAIL PROTECTED]
> Sent: Thursday, June 05, 2003 4:40 PM
> To: OJB Users List
> Subject: Sunglasses for OJB (bright future)?
> 
> 
> We're trying to get a reading on OJB regarding its development status,
> degree of industry acceptance, roadmap, its future prospects, 
> etc... All the
> good things you'd want to think about before committing a significant
> project.   We've heard that Thomas Mahler's project at 
> Karstadt/Quelle and
> about a project at Los Alamos Nat Labs, but we'd be 
> interested in hearing
> from many others.

You'll find several more reference in our mailing archive.
Some references from the web:
http://www.openemed.org/ 

http://www.pansoft.de/objectrelationalbridge.jsp

http://www.ebi.ac.uk/intact/
 
> We also interested in contrasting an OJB approach vs. JDO 
> strategy, so it
> would be great to hear from anybody who has gone through that 
> decision.
> 
> We'd appreciate any feedback from people who have already made this
> decision, and how this has worked out so far.  Also, anyone 
> who can comment
> on where OJB is on the release roadmap, why it is not yet 
> reached v1 (still
> RC3?)...  
Don't get fooled by version numbers. 
If OJB was a commercial product the version would be 5.0 already!


> And when OJB v2.0 could be expected?

When all the work is done ;-)
Hey, this is open source development: No fixed dates, No fixed nothing!
The more people contribute the more work gets done in a shorter period of
time.

cheers,
Thomys


> Thanks much,
> Carl 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 
> 


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to