Sehrope Sarkuni

@michaelshobbs's response is correct. The deploy step will skip installing supervisord if it's already installed you need to install it into the buildstep image. I do that step manually and...

+1 to exposing our own interfaces and let the end user bridge that to their own collector. That would give us the ability to make it match whatever level of...

That's not encryption, it's percent encoding (also know as URI encoding): https://en.wikipedia.org/wiki/Percent-encoding This isn't specific to this driver or even Tomcat. It's true for all URLs. Any character outside the...

The comment by `decode(...)` mentions it being twice as fast as the JVM built-in bytes-to-String as of JDK 1.4.2: ```java /** * Custom byte[] -> String conversion routine for UTF-8...

@bokken Very interesting analysis! In the interests of simplicity, why not just a local array up to ??? KB (with a property to configure the max) then defer to the...

Yes you can ignore that. It's unrelated. That error is because the `pg_terminate_backend(...)` does not actually wait for the other process to be killed. So it's possible (as in this...

In skimming through some of the older docs to verify API availability for JDK 6 I noticed that there are two methods to load a CRL from an InputStream: ```java...

@marschall Because that test has a race condition. It has one connection kill another and checks for the connection to be invalid. But since it does not wait it's possible...

Why is this necessary? `TimestampUtils` is local to each connection and your test is accessing the same connection from two different threads with no external synchronization. That type of non-synchronized...

@vlsi Wow I didn't know the docs said that! Yes that's blatantly false and should be removed from the docs.