Re: adding 'zstd' as a compression algorithm

Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> On Tue, Feb 15, 2022 at 7:09 PM David Steele <david(at)pgmasters(dot)net> wrote:
>> This is even more true for zstd since it is not as ubiquitous as lz4. In
>> fact, it is not even included with base RHEL8. You need to install EPEL
>> to get zstd.

> Yeah, I agree. One thing I was thinking about but didn't include in
> the previous email is that if we did choose to make something like LZ4
> the default, it would presumably only be the default on builds that
> include LZ4 support. Other builds would need to use something else,
> unless we just chose to make LZ4 a hard requirement, which would be
> bolder than we usually are. And that has the consequence that you
> mention. It's something we should consider as we think about changing
> defaults.

Yeah. I'm +1 on adding zstd as an option, but I think we need to
move *very* slowly on changing any defaults for user-accessible
data (like backup files). Maybe we have a bit more flexibility
for TOAST, not sure.

regards, tom lane