mirabilos

NaN and ±Infinity should just be serialised as literal “null” instead. I’ve been searching for about half an hour for how to get Jackson to do that, i.e. behave as...

Tatu Saloranta dixit: > @mirabilos What should be output is really debatable; `null` is one > @possibility but there are many others (throw exception; just output > @non-standard variant). Challenge...

“what ECMAscript does is not definitive in any sense of the word” This is not quite right. JSON is a proper subset of ECMAscript except that it allows two more...

On Sun, 8 Apr 2018, Tatu Saloranta wrote: > So while it is true that ECMAScript specs _also_ define JSON, that > does not somehow make JSON subset Agreed, **except**...

Yes, please!

The test doesn’t help, as its scope seems to be a speed test, not a test for the correctness of the transformations. No, we cannot always write PIC code, because...

I’m fixing that the assembly code cannot be used in PIC (such as `libxorgxrdp.so`) or PIE (executables) modules; this affects all ELF platforms where PIE is used for executables and/or...

No, current xrdp builds are also affected since GCC now defaults to PIE. I know about the non-assembly fallback (it’s used on all other architectures, like x32, anyway), but that...

Additionally, it would be nice to say: one blank line, but none above if the licence block is the first thing in the file, and never more than one blank...

@bot see https://github.com/mathieucarbou/license-maven-plugin/issues/78#issuecomment-240091273