M.K.

It would indeed be very helpful still. I do not see how these rules might conflict with each other; in fact, we do have a very crude way to enforce...

Could you make it a dict instead? That would provide a nice upgrade path for even stricter requirements, like a whitelist of accepted values. Edit: On the other hand, this...

If its "[for] FAT & MDO ONLY" then a whitelist check of the given username might help as well. Also a hint like "(automated bot message)" at the beginning could...

Sorry for not having read it properly a minute ago. :+1: for the threshold option.

I'll assume that for each length of input, gzip achives maximum efficiency if and only if all input bytes are the same. In this case, and using values from 0…127,...

I think it would be nicer to run the benchmark some few hundered times in the same nodejs process for warmup, only then collect measurements. v8 might be able to...

Would it help to deflate/inflate buffers instead of the stream? That way we could use zlib.deflateSync. ("Added in: v0.11.12")

I agree, we need an optimized path for applications that are basically signalling or chat, because that's how I will primarily use it. (Of course we shall still support large...

> compressing very small messages […] does not make an app faster but slower. Thanks for reminding. I should have expressed myself more clearly. I meant that even in scenarios...

I'm affected as well. When will a version with the patch be released?