Is it possible to not have standard-style subcommands with the pacman ones?

[erin-desu]

Now, if you make a pacman-style subcommand, it allows using it the standard way. So we can have query, -Q and --query.

Is there a way to prevent query to work? I do not want to support that many alternative variants and have more unified interface.

[erin-desu]

Update: It isn't.

But it seems pretty easy to implement.
I just don't know if to implement it per-subcommand or globally (with local overrides ofc).

[erin-desu]

I just do not want to support multiple subcommands format. It is bit confusing. That's all.

Also implementing that doesn't feel really hard?

[erin-desu]

...by that I mean i am willing to implement it myself in.

[epage]

We are a bit cautious in the features we add because they affect

• Binary size
• Build times
• Discoverability: the more features we have, the harder it is to discover any features

We need to get a feel for what use cases are like for having both or only flags and enumerate the parsing and help behavior. For example, should we completely disallow named subcommands if flag subcommands are used? Granted, thats breaking change and we aren't quite ready to start on v5 but we could have short lived setting until then or develop this behind unstable-v5 feature flag.

Might also be good to see when flags were added to see if any of this was discussed back then.

If there is interest, then the first step is opening an issue.

[erin-desu]

Well, i considered just making a settings bitflag. Then i think it was like at most 20 lines of code from my little dirty prototype i did when i was bored.

[MCorange99]

This would be pretty nice, since it would allow pacman style of flags, which i really like, cause i dislike normal subcommands, and prefer flags. I dont really know anything about the codebase but i would think that it shouldnt be too bad adding a method to set if the normal subcommand should be available.