: Thats oracle's problem (not ours). I would argue that if we release packages that fail to build cleanly because our build system:
a) is explicitly configured to try and build linkages against a URL that 404s (which causes javadoc to issue a warning) b) is explicitly configured to fail on any javadoc warnings ..then that is our problem. (particularly since we don't do anything to document / encourage users to override the broken URL if they try to build from source) : can use -Djavadoc.link. But we cannot enforce this: they might not be : using an oracle jvm, nor can we ship with the javadocs from oracle (as : the license.html that comes in the docs package does not look : compatible with apache to me) Agreed -- we shouldn't / can't ship a package-list from Oracle, but we should change the hardcoded javadoc.link value in common-build.xml to "" (ie: the empty string) ... that causes the build to succeed completely -- if users want happy/shiny/pretty links to java.lang.* classes then they can override -- but they shouldn't have to dig through the build.xml just to figure out how to get a basic build to work cleanly. And FWIW: I was mistaken in my last email when i said "ant javadocs" built all the javadocs correctly and then failed because of the warning -- i failed to notice that only the "All" version of thes docs were building, and then the build was immediately failure (before it recursed and did the individual contribs and the test-framework) ... and since the test framework is not included in the "All" copy of the javadocs, that means src users don't get any copy of those javadocs at all. So I'm ammending my vote of RC2 to a -1. -Hoss --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
