|
For me, it went something like this: A Connection object and a Statement object are tightly coupled to each other. The statement 'belongs to a connection'. They clearly have references to each other so that they can stay in synch (close on connection should close any statements left open under it for example). So we started going down the path of just letting the statement get at the internals of the connection for whatever it might need. But later, we start seeing heavily threaded applications, applications sharing transactions (via JTA support sitting on top of JDBC), etc. So we synchronize everything in connection and synchronized everything in statement to stop two threads from messing up a single object, but the problem that remained was that threads doing stuff to the statement are 'free' to modify the connection because they didn't call a method on the connection, they just changed what they wanted to. This caused some inconsistancies that were very hard to track down. The 'big hammer' approach isn't good, because we don't want everything locked to the connection or only one statement could do work at a time... so we had to go through the process of breaking apart the places where these objects could manipulate each other directly so that we could synchronize them correctly. Its a pretty hard thing to do after the fact. Combined with the fact that nearly all of our time (performance) is spend doing 'real' work, allowing the objects to have direct access to each other was a mistake in the first place... I probably didn't explain that very well. The real issues were, at times, the result of very complex interactions. And, of course, they were rarely easy to recreate so you could analyze them effectively (you're always in trouble when you turn on debug and the problem goes away). The lesson I learned (at least sort of learned... I'm a long ways from perfect about it yet) is that you need to write code from the start like you expect a lot of people to use it or risk having a lot of nasty issues on your hands trying to make it work for them later. Setter/getter methods are a prime example of a good practice to protect yourself (there are dozens of others). Richard D. Dettinger iSeries Java Data Access Team Democracy's enemies have always underestimated the courage of the American people. It was true at Concord Bridge. It was true at Pearl Harbor. And it was true today. Rochester Post-Bulletin Tuesday September 11, 2001 |---------+----------------------------> | | "David Morris" | | | <David.Morris@plu| | | mcreek.com> | | | Sent by: | | | java400-l-admin@m| | | idrange.com | | | | | | | | | 04/09/2002 12:23 | | | PM | | | Please respond to| | | java400-l | | | | |---------+----------------------------> >----------------------------------------------------------------------------------------------------------------------------| | | | To: <java400-l@midrange.com> | | cc: | | Subject: RE: Java Style Question | | | | | >----------------------------------------------------------------------------------------------------------------------------| Richard, I agree that performance is a non-issue and getters/setters offer an opportunity to add your own hooks into a class. I am wondering why a getter or setter for a class variable changes the multi-threading issue. Most multi-threading issues I have run into are fixed by passing them as parameters and avoiding either direct access or access via getter/setter. David Morris >>> cujo@us.ibm.com 04/08/02 10:22PM >>> ...2) Later you start multi-threading something that you never worried about threading issues for initiallly. You start making various objects safe from the outside users and that is pretty easy, but you have all kinds of thread safety holes where two objects clobber each other's internal state from two threads that are allowed to execute at the same time... Richard D. Dettinger _______________________________________________ This is the Java Programming on and around the iSeries / AS400 (JAVA400-L) mailing list To post a message email: JAVA400-L@midrange.com To subscribe, unsubscribe, or change list options, visit: http://lists.midrange.com/cgi-bin/listinfo/java400-l or email: JAVA400-L-request@midrange.com Before posting, please take a moment to review the archives at http://archive.midrange.com/java400-l.
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.