Join GitHub today
GitHub is home to over 31 million developers working together to host and review code, manage projects, and build software together.Sign up
Default to non-interactive init. #5902
By default, running
Getting the old behaviour
Whether or not init is interactive is controlled by a flag in
By changing it to the value below, all
Please include the following checklist in your PR:
I'm also against this as the default.
Beginners don't usually start by making packages. Even if they do, the current interactive mode isn't that scary; most questions have defaults so it is possible to just "skip" the ones you don't understand (and still end up with something reasonable).
That said, I also want to appreciate how much effort @m-renaud has put into revamping
I tend to agree with harpocrates here. The interactive mode holds the hands of beginners through a lot of stuff, and I think it is a good experience. I think that it might be nice to more clearly highlight as part of interactive mode that there is a flag to skip it in the future, to streamline the process for them. But certainly interactive feels more friendly for newcomers, and non-interactive for people that already know what they're doing.
Thanks everyone for the input and for the recognition, it means a lot :) It an attempt to not scatter the discussion too much I've replied to hvr's comment in #5696 (comment) to speak to some of the points he raised as well as bringing in some of the other comments left here to have a more complete picture in one spot.
A few short answers (expanded in the other thread):
I don't think this is true because as soon as you try and do anything non-trivial you need to pull in some packages from hackage since haskell bundles very little in its base library. Also, using the
If the goal is to end up with something reasonable then not having to do anything and still ending up with something reasonable seems like the base case :)
To speak briefly to a couple of concrete suggestions here:
I really like this idea and is something to be done regardless of the outcome of this PR discussion, so I'll put together a PR for that when I get a chance.
I mentioned a poll in the other comment, but wanted to ask again here to see what folks thought. LMK and I can put one together, could also post on various discussion threads to see what other people think.
Lastly, I just want to state that I completely respect whatever decision is made regarding this change, I just want to make sure that the various trade-offs have been thoroughly considered and there is diversity of input. I know the release of cabal 3.0 is coming up very quickly so it would be great to have a decision by then, but if that seems unlikely I'd like to at least make sure docs and suggestions from the tool are updated appropriately to reflect the ability to change the default.