Re: [classlib] Build failure messages

2006-10-20 Thread Denis Kishenko
I have made fix[1] for unused category (only java code). If nobody has objections I could fix other caterogies. [1] http://issues.apache.org/jira/browse/HARMONY-1926 2006/10/20, Nathan Beyer [EMAIL PROTECTED]: I've been trying to work through many of the warnings, module by module. Most of

Re: [classlib] Build failure messages

2006-10-20 Thread Tim Ellison
Denis Kishenko wrote: I have made fix[1] for unused category (only java code). If nobody has objections I could fix other caterogies. [1] http://issues.apache.org/jira/browse/HARMONY-1926 See Nathan's comments about unused vars -- they cannot be fixed safely with the Java wizard. Unused

Re: [classlib] Build failure messages

2006-10-20 Thread Ilya Okomin
On 10/20/06, Tim Ellison [EMAIL PROTECTED] wrote: Denis Kishenko wrote: I have made fix[1] for unused category (only java code). If nobody has objections I could fix other caterogies. [1] http://issues.apache.org/jira/browse/HARMONY-1926 See Nathan's comments about unused vars -- they

Re: [classlib] Build failure messages

2006-10-20 Thread Denis Kishenko
Almost all unused variables are local. Only two of them are class fields and they are not using in native code. Build and all tests are successful. 2006/10/20, Tim Ellison [EMAIL PROTECTED]: Denis Kishenko wrote: I have made fix[1] for unused category (only java code). If nobody has

Re: [classlib] Build failure messages

2006-10-20 Thread Tim Ellison
Denis Kishenko wrote: Almost all unused variables are local. Only two of them are class fields and they are not using in native code. Build and all tests are successful. Thanks Denis! Regards, Tim -- Tim Ellison ([EMAIL PROTECTED]) IBM Java technology centre, UK.

Re: [classlib] Build failure messages

2006-10-20 Thread Denis Kishenko
Added patch for SUID wanings to the same issue.H-1926 2006/10/20, Tim Ellison [EMAIL PROTECTED]: Denis Kishenko wrote: Almost all unused variables are local. Only two of them are class fields and they are not using in native code. Build and all tests are successful. Thanks Denis! Regards,

Re: [classlib] Build failure messages

2006-10-20 Thread Paulex Yang
Nathan Beyer wrote: I added a compiler arg to the main 'build-java.xml' compile to disable all of the warnings. Thanks a lot, Nathan, but how about also add that arg to build-test.xml? I think the ECJ probably feels more unhappy with the test cases:). This should take care of the warnings.

Re: [classlib] Build failure messages

2006-10-20 Thread Nathan Beyer
I looked at adding the nowarn arg for the tests, but 'build-test.xml' doesn't have a single large compile statement like 'buid-java.xml' does. The arg would need to be added to the compile-test of every module's build, which can be done, I was just feeling lazy at the moment, sorry. :( Also, I

Re: [classlib] Build failure messages

2006-10-19 Thread Denis Kishenko
I also worried about a lot of warnings and I think warnings like these can be fixed last\modules\concurrent\standard\src\main\java\java\util\concurrent\ConcurrentHashMap.java [javac] (at line 26) [javac] import java.io.ObjectInputStream; [javac]

Re: [classlib] Build failure messages

2006-10-19 Thread Mikhail Fursov
+1 I never know if it's something important or not. On 10/19/06, Denis Kishenko [EMAIL PROTECTED] wrote: I also worried about a lot of warnings and I think warnings like these can be fixed last\modules\concurrent\standard\src\main\java\java\util\concurrent\ConcurrentHashMap.java [javac]

Re: [classlib] Build failure messages

2006-10-19 Thread Tim Ellison
Nathan has turned off all warnings. I suggest that we start to turn them back on incrementally and fix the warnings we see. Regards, Tim Denis Kishenko wrote: I also worried about a lot of warnings and I think warnings like these can be fixed

Re: [classlib] Build failure messages

2006-10-19 Thread Nathan Beyer
I've been trying to work through many of the warnings, module by module. Most of the warnings fall into the following categories. * serial - Serializable classes that don't have explicit serialVersionUID fields * unchecked - missing type variables, mostly collection references * unused - unused

Re: [classlib] Build failure messages

2006-10-19 Thread Geir Magnusson Jr.
coolio - thx Nathan Beyer wrote: I added a compiler arg to the main 'build-java.xml' compile to disable all of the warnings. This should take care of the warnings. -Nathan On 10/18/06, Geir Magnusson Jr. [EMAIL PROTECTED] wrote: Can we just shut them off for now? it's a good project for

Re: [classlib] Build failure messages

2006-10-19 Thread Geir Magnusson Jr.
You have patience I can only dream about... geir Nathan Beyer wrote: I've been trying to work through many of the warnings, module by module. Most of the warnings fall into the following categories. * serial - Serializable classes that don't have explicit serialVersionUID fields * unchecked

[classlib] Build failure messages

2006-10-18 Thread Mark Hindess
These messages aren't getting through on the commits list on due to the large number of warnings since switching to ecj. I've modified the summary process so that it ignores these for the moment. Hopefully this means that the failure messages that haven't been making the commits list will start

Re: [classlib] Build failure messages

2006-10-18 Thread Geir Magnusson Jr.
Can we just shut them off for now? it's a good project for someone to go and clean those up, but there's no need for all of us to see them. geir Mark Hindess wrote: These messages aren't getting through on the commits list on due to the large number of warnings since switching to ecj. I've

Re: [classlib] Build failure messages

2006-10-18 Thread Nathan Beyer
I added a compiler arg to the main 'build-java.xml' compile to disable all of the warnings. This should take care of the warnings. -Nathan On 10/18/06, Geir Magnusson Jr. [EMAIL PROTECTED] wrote: Can we just shut them off for now? it's a good project for someone to go and clean those up, but