JDK8 Warnings Cleanup Day: Coding Guideline (original) (raw)
Stuart Marks stuart.marks at oracle.com
Thu Dec 1 00:27:34 PST 2011
- Next message: Request for Review of 7116890 (Warning Cleanup java.io)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Warnings Cleanup Day is here!
When you're cleaning up warnings today, we'd like you to focus on one thing:
==> simple, minimal, and risk-free warnings fixes <==
Most of the warnings can be eliminated with such changes. For example, take a look at Alan Bateman's recent changeset [1]. The vast majority of these are one-line changes. These should be close to obviously correct by inspection, so that we can push in as many warning fixes as we can with only minimal testing.
Along the way you'll inevitably run across opportunities to do other cleanups, do some refactoring, reformatting, fixing little bugs you run across, and so forth. Please avoid this temptation and stay focused on minimal warning fixes.
Another thing you'll probably run into is warnings that can be removed by changing public APIs. In some cases APIs aren't very well designed (particularly with generics and wildcards) so using them always generates warning messages. Unfortunately, because of compatibility constraints and processes for evolving Java's public APIs, any such changes are out of scope for Cleanup Day. If you run across such cases, please note them so that we can file a bug, and move on.
It's not a goal of cleanup day to eliminate every last warning. We expect that there will be a small number of warnings that are caused by poor APIs, or that are too difficult to fix or that require too much discussion to deal with in a single day. Again, make a note of these, keep things simple, and move on.
OK, enough talk. Let's code!
s'marks
[1] http://hg.openjdk.java.net/jdk8/tl/jdk/rev/7795c41ed54c
- Next message: Request for Review of 7116890 (Warning Cleanup java.io)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]