zstd-suport in GRASS >= 7.6: update manuals for variables and r.compress

[sbl]

The release news for GRASS 7.6: ​https://trac.osgeo.org/grass/wiki/Release/7.6.0-News , say that new default raster compressor (if available) is ZSTD.

Manuals for:

• variables (https://grass.osgeo.org/grass76/manuals/variables.html)
• r.compress (https://grass.osgeo.org/grass76/manuals/r.compress.html)
• ???

are not fully consistent/up to date in this regards and should be updated accordingly...

[mmetz]

In 74114:

update documentation with regard to ZSTD as default compression if available, see #3758

[mmetz]

In 74115:

update documentation with regard to ZSTD as default compression if available, see #3758 (backport trunk r74114)

[sbl]

Thanks. Found one little glitch (see attached diff for a suggestion). BTW., the manual for r.compress (​https://trac.osgeo.org/grass/browser/grass/trunk/raster/r.compress/r.compress.html?rev=74114#L98) says:

The NULL file compression is using the LZ4 method as being the best compromise between speed and compression rate.

Is that still accurate?

[sbl]

Removing last ZLIB-default remnant

[mmetz]

Replying to sbl:

Thanks. Found one little glitch (see attached diff for a suggestion).

Thanks, applied in trunk r74119. You can of course submit these improvements yourself ;-)

BTW., the manual for r.compress (​https://trac.osgeo.org/grass/browser/grass/trunk/raster/r.compress/r.compress.html?rev=74114#L98) says:

The NULL file compression is using the LZ4 method as being the best compromise between speed and compression rate.

Is that still accurate?

Yes, there is practically no speed penalty with LZ4, and NULL files are so simple that not much is gained with other compression methods.

[sbl]

Thanks for clarification.

Yes, next time I open tickets esp. on documentation I try to come up with a fix myself much faster. (But you make it hard to be fast enough ;-))

Closing as fixed.