Compress manual doesn't list lz4-v2

[jiquera]

Please fix the entry in the manual and explain the difference and the preferred setting. Right know the sample scrips are using lz4-v2 and the manual only mentions lz4.

Related to that... why do you support both? the difference is pretty much negligible. (one byte at most per packet)

[Gert Döring]

the difference is packet data alignment, which is better with the -v2 compression algorithms - which reflects into "less CPU usage, less power drain on mobile devices, possibly higher throughput".

And you're right, we need to document this.

[jiquera]

yeah you explained it the other day over IRC ;) which finally made me understand the code. This is basically the follow up: document it :)

[mmokrejs]

Why is this trivial change not yet present in 2.4.3 manpage?

[Gert Döring]

@dazo: while you're at moving the manpage over to rst, can you include lz4-v2?

[tct]

This is a low hanging fruit , I will submit a patch.

[Gert Döring]

bump?

[plaisthos]

It is already in the manpage:

The algorithm parameter may be :code: lzo , :code: lz4 ,

:code: lz4-v2 , :code: stub , :code: stub-v2 or empty.
LZO and LZ4 are different compression algorithms, with LZ4 generally
offering the best performance with least CPU usage.

[Gert Döring]

But that is missing why there are "lz4" and "lz4-v2"...

[Gert Döring]

bumping to 2.5.4, waiting for the patch promised 19 months ago

[tct]

The man page now has a good description of compression, including:

The lz4-v2 and stub-v2 variants implement a better framing 
that does not add overhead when packets cannot be compressed. 
All other variants always add one extra framing byte compared 
to no compression framing.

Details: ​ https://github.com/OpenVPN/openvpn/blob/master/doc/man-sections/protocol-options.rst
Scroll down to --compress algorithm