Re: [PATCH] Add the "-info-os" command to MI

[Tom Tromey <tromey at redhat dot com>]

>>>>> "Stan" == Stan Shebs <stanshebs@earthlink.net> writes:

Stan> I can see the appeal of the decoupling, but we share argument parsing
Stan> like this because there's not a lot of practical purpose to having two
Stan> parsers - once the command is out there and being used, we're loath to
Stan> change it much anyway.  To me, the MI's main value is in avoiding
Stan> cases where CLI ambiguity and elision makes for a better user
Stan> experience, but in this case the arguments are data-driven by
Stan> platform-specific code, not much ambiguity involved.

Ok, I won't insist on it.

I thought this had bit us before but I couldn't find a case in the
archives.  So maybe my concerns are theoretical.

Tom