xz piped into xzcat

[alokmenghrajani]

No description provided.

[jwilk]

Could be shortened to xz < file.txt | xzcat.

[wader]

Not sure what to do about some of the pipe-variants that are compress | uncompress etc hmm

[alokmenghrajani]

I would nuke the pipe section. My $0.02.

[wader]

Yeap maybe good to just keep it to single programs

@jwilk agree?

[wader]

@alokmenghrajani sent an invite so you can help out if you want

[jwilk]

Yeah, the compress | decompress scheme (and more generally foo --encode | foo --decode) is a bit… bland?

I'd reluctantly allow gzip (ubiquitous!) and xz (very short!), but probably not other compressors. Maybe. Dunno.

maybe good to just keep it to single programs

Is echo ',p' | ed -s file.txt one program or two? Because I'd hate to lose it.

Maybe discourage using the same (or nearly-same) program twice instead?

[alokmenghrajani]

@alokmenghrajani sent an invite so you can help out if you want

Thank you, I'm honored.

One way to keep things fun would be to define a specific operating system. E.g. a very specific Docker image, such as docker run --rm -it ubuntu@sha256:2cde79b4627d38d1448fc264f93e465f18b653bc9a62ee8ec85d99d4e8f39d4c /bin/bash. The number of boring cases will then perhaps be limited.

note: we shouldn't pick this specific Docker image since it's a little too bare (no gcc, no curl, etc.).