Den fre 29 juli 2022 kl 16:29 skrev Branko Čibej <br...@apache.org>:
> On 27.07.2022 22:18, Daniel Sahlberg wrote: > > Den tis 26 juli 2022 kl 10:55 skrev Branko Čibej <br...@apache.org>: > >> On 15.07.2022 23:19, Daniel Sahlberg wrote: >> > Den fre 15 juli 2022 kl 10:46 skrev Thomas Singer >> > <thomas.sin...@syntevo.com>: >> > >> > Hi SVN developers, >> > >> > >> > Hi Thomas, >> > >> > Thanks for the detailed report! >> > >> > First, there are 2 classes of NativeException, one in >> > org.tigris.subversion.javahl package, the other in >> > org.apache.subversion.javahl package. They only slightly differ in a >> > constructor parameter. I recommend to use one class with 2 >> > constructors >> > instead. >> > >> > >> > I'm not experienced enough in the Java world to fully grasp this but I >> > assume the org.tigtis package is older and that it has been kept for >> > historical reasons (not to break an API used by older applications). >> >> The org.tigris package is obsolete but indeed kept for compatibility. We >> shouldn't even try to "improve" it by using some common implementation. >> > > When compiling, there are a bunch of warnings: > > [[[ > ... > /home/dsg/svn_trunk/subversion/bindings/javahl/src/org/tigris/subversion/javahl/SVNClient.java:1605: > warning: [dep-ann] deprecated item is not annotated with @Deprecated > public void diff(String target1, Revision revision1, String target2, > ... > ]]] > > I assume this is because the function is @deprecated in the docstring but > not formally annotated as @Deprecated. > > Would this be a reasonable improvement or should the code be considered > frozen? > > > It's frozen, don't bother with it. Nobody should be using it any longer. > Still it causes compile time warnings that could mask other warnings. /Daniel
