Robbie Gemmell

I also skimmed it and noted some things, replied to some of the comments on Tim's review. Even for a first phase it seems pretty raw/incomplete at this point, and...

Note the PR build is still failing, with: [INFO] Running org.apache.activemq.transport.http.HttpJMSMessageTest [ERROR] Tests run: 32, Failures: 0, Errors: 4, Skipped: 0, Time elapsed: 16.633 s

The value set by the property in ClassLoadingAwareObjectInputStream is a static, so updating it doesnt do anything if already set when the codepath has been first used. So the first...

AH, I somehow missed the last line of your previous message :)

I'll admit I gave up before the end as many of the comments seem to apply to a lot of the tests and I got bored seeing/saying the same things,...

> > @gemmellr thank you for your prompt review. Please characterize in more detail (or point to an example) of a _unit_ test as you described. > A unit test...

> Local MacBook Pro (total 32.535s): > Apache Jenkins CI (total 43.921s): > I'm taking it from the comments here and https://github.com/apache/activemq/pull/729#discussion_r866242620) that 32-44 seconds isnt seen as a that...

> I assumed we could merge this in to a 5.17.X release and then it would be fair game to remove in 5.18.0. Is there some written guidance on due...

Since the docs say this mechanism is essentially an alternative means of specifying static "failover:" transport usages with some unnamed options, and the code seemingly confirms that from looking like...

> @gemmellr > > > The other config type is "static": > > ``` > > > > ``` > > > > > > > > > > >...