> On 23 Apr 2020, at 00:47, Jonathan Gibbons <[email protected]> > wrote: > > > On 4/22/20 2:35 PM, Pavel Rappo wrote: >> 1. We should check that this change passes the tests yet leaves the OpenJDK >> API documentation unaffected. The documentation shouldn't change, nor do I >> expect it to do so. > > Pavel, > > Thank you for your suggestion. It had an unexpected outcome! Yes, the docs > are different, for one file, which caused a minor initial panic, until I > realized the root cause. > > I can't post screenshots here, so I'll point you at the existing > documentation, and leave you to imagine how it is now displayed...
I built it just to have a look at the result. Impressive! > https://download.java.net/java/early_access/jdk15/docs/api/jdk.jfr/jdk/jfr/consumer/EventStream.html#setReuse(boolean) > > I think this case should be treated as a separate issue, as to why it is not > causing a more serious error when the docs are created. This is indeed surprising since we should see an error error: unterminated inline tag * If reuse is set to {@code true), an action should not keep a reference ^ which we do not. We should have a look at how `make docs-jdk-api` is executed. Specifically at how it treats errors and warnings. On a related note. Doc comments in jdk.jfr could benefit from proofreading. For instance, some of those comments mistake {@code} for {@link}. I'll create a followup issue for the jfr team. -Pavel
