Dev

[digikar99]

I guess the commits mostly cover all the changes; those that do not seem well mentioned include -

• travis now checks for both ecl and allegro
• sionescu fork for travis is a maintained one, with more recent implementation versions; these updates in turn demanded a need for the addons - perhaps, this should be taken care of in the travis script
• unix-opts-tests is renamed to unix-opts/tests; and (unix-opts/tests:run) runs the tests; there is some test refactoring as well; fiveam, local-nicknames, asdf:test-op are not supported on all implementations (clisp in particular); hence, cannot be used.

I am not sure about the versioning conventions, but the change on adding ambiguous-option condition might be significant enough for some people to demand a version bump.

[digikar99]

While I had attempted default, your implementation with the consideration of mutable args is probably better; as well as my implementation is incomplete missing out on a few aspects such as modifying the describe function.

[libre-man]

Hi thanks @digikar99 for the extensive fixes!

I will cherry pick/manually extract the features from it one by one if that is ok by you. The first one I've done was improving the testing in #27. In the next few days I'll extract the macro for the shorter definition of the options.

[digikar99]

Yes that's perfectly okay; if anything, the way it has been done with the recent PR is much better than what I had attempted.

[digikar99]

Looks like most of the things have been merged. Two of the things that seem to remain include ambiguous-option and expand-opts. Any opinions about them?

[libre-man]

I'm still planning on merging them, but I haven't had time for it just yet, also because it needs quite some work to split them up especially as the master now contains some of the functionality already of this PR. I'll probably have some time this weekend to get started on it again, so hopefully I'll be able to merge the expand-opts soon.

[digikar99]

I'd be willing to close this PR - this looks pretty dirty anyways - and perhaps issue a separate PR explicitly for the expand-opts.