Bernd

Hm, I think this would be a good improvement, but I am not sure if all providers (especialyl external ones) are ready to deal with concurrent invocation of the do...

I don’t think there was ever a clear specification on locking/concurrency, so must provides would have checked what the code does. But having said that, I don’t think that area...

BTW thanks for using a branch, lets give us some time to discuss this, its painful to remebr all those undefined corners (again and again).

BTW I always try to use a VFS manager single threaded, I am not completely sure but most of the code looks like this was the intended use anyway :)

One of the most important checks a registry client is doing regularly are checking if a tag has changed (all other artifacts are immutable). So having a (HTTP compatible) way...

Two points, if you keep it in the spec I would strongly warn about using it for too large entities and also mandate that Expect: Continue should be sent/supported, especially...

As CAST5 is a 64bit block length cipher it really should be avoided. I guess the only reason it is given here is for interoperability (as it is a RFC4880...

> > What is wrong with the change? > > You'll need parentheses to make that work, changing it to use '|' would work too. Does that mean there is...

> > Does that mean there is no explicit test for DataInputStream endianess > > @ecki I think this is not strictly necessary as order of bytes is explicitly specified...

the accessibility not only fails in forced cases with Java 17 but also in cases where a method is Public, but only if accessed through the right type. For example...