On Dec 8, 2017, at 09:12, Ken Cunningham wrote: > I have suggested in the the past that depends_* could be a synonym for > depends_*-append. That would solve this easily. > > we could have a depends_*-reset if anyone really wants that (dont see it ever > being used, but ...)
I'm worried that this will be more confusing. The depends_* options would behave differently from every other option (notably except default_variants, which already behaves this way, though that might be considered a bug). sqlite3-tools is an example of subport that deliberately overwrites the dependencies of the main port. Though, I'll admit that while looking through ports, most of those I found that that overwrite depends_lib shouldn't be doing so. There are already many ports that clear the dependencies by writing e.g. just "depends_lib" (with no value after it). They do this when they detect they cannot install, so that the user gets the message before installing lots of unneeded dependencies. Granted it would be better to have a feature for this in base, where a port can indicate that it cannot be installed.
