>> POSIX doesn't even mandate having a gecos field at all.
>
> Urk!  Right you are: http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/pwd.h.html
>
> That's an argument for not including it in this base *POSIX* SRFI,
> especially when combined with the complexity of what it might contain

Indeed, it can be a reasonable argument to tip the scale slightly on the
side of leaving it out.

It's not really that complex though - the subfield format is a gross
hack to be sure, but it has been quite uncontroversial over decades.
Just split at commas, take the first field and replace ampersand (if
any) with capitalized username. Any fields beyond the first are not
useful except for the now obscure finger.

> If we do, I currently advocate using the two nouns "comment" and
> "gecos", one for its name, the other in the description, I'm not
> particular on which for which.

The thing is, we can all secretly hope the gecos field will die in the
future to be replaced by a normal "pw_fullname" field :) It seems like
the POSIX spec writers did. For that reason I would like us to pretend
that Unix just gives us a fullname. Hopefully, through some mutual
agreement to phase out gecos, it will one day be so.

> See my previous comment on read-link being insufficient for backups,
> and for getenv we should add a requirement that either you're using an
> R7RSsmall Scheme or implement SRFI 98 from which it got
> get-environment-variable, since we're assuming you have such a
> procedure so that you can query your PATH.

Indeed getenv is probably a moot point, as a Scheme that goes to the
trouble of providing all the complex OS stuff but leaves out getenv
would be quite absurd :)