Cody Allen
Cody Allen
For a non-interactive session UCM probably should avoid printing the big hello banner as well. Right now if you use `run.file` you see the following before the output of your...
Perhaps [queryTerminal](https://hackage.haskell.org/package/unix-2.8.1.0/docs/System-Posix-Terminal.html#v:queryTerminal) could be used to implement this? I'm not sure whether that would work in all Windows environments. But maybe solving it for posix-compatible environments is still a good...
> maybe non-interactivity shouldn't always be tied to the stdin/out Handle? @unorsk I think that you are probably right. Perhaps there should be explicit `--non-interactive` and `--interactive` flags and in...
Would the doc be evaluated at runtime? I tend to do `_ = do {{ my doc }}` to prevent unnecessary evaluation. > And then this Doc could also, optionally,...
> I think it's a major can of worms to try to do anything else. @pchiusano any specific concerns here? Do you think that the implementation would become difficult to...
> but then when you make another change to dequalify a symbol in the same function, suddenly your program doesn't typecheck anymore That's a really good point. I hadn't considered...
The runtime could potentially give a better error message, but there is indeed a type error in this code. `TypeMap.alter` is passing an `a` into a function that is expecting...
> Can we look at the specific situation where not having the full name is making debugging issues harder? Sure. Here is the original error message that I was trying...
Here is another example:  The supplied name of one of those is a suffix of the other, so it really isn't even providing a unique suffix. In this case...
@mitchellwrosen I am also having trouble reproducing the non-unique suffix issue with a minimal example. There have been a few changes in this area recently and it's possible that one...