Convenient API for reading/writing sequence-style files incrementally

[well-typed]

Convenient API for reading/writing sequence-style files incrementally

[thoughtpolice]

The @dcoutts and @thoughtpolice method to Software Success™ is clear at this point:

• Think really hard, write a lot of code.
• Finally release that code after several years.
• Keep working for another year on it entirely.
• Never release it.
• Continue to add more scope forever and ever as things come up.
• GOTO POINT 4.

I have also used the new GitHub feature of multiple-assignees to assign this to both of us. Therefore, the total amount of responsibility has been diffused and split evenly between us, ensuring this will take twice as long to accomplish.

[bgamari]

I agree that this shouldn't be a release blocker.

[thoughtpolice]

I'm tentatively moving this into scope on the initial 0.2 release, because I imagine this use will pop up pretty quick and it'd be good to have something people can play with to start off.

[thoughtpolice]

(In all seriousness though, this would be useful to tackle soon-ish, so I might do this while @dcoutts continues working on #62)

[spl]

@thoughtpolice:

• GOTO POINT 4.

It seems like your BASIC-like program will crash because your bullets aren't numbered. Or is that dynamically-numbered programming?

[bgamari]

#123 makes this a bit easier.

[dcoutts]

New related draft standard https://datatracker.ietf.org/doc/draft-bormann-cbor-sequence