One of my goals for #mollytime is to avoid introducing a concept of errors into the language.

[lritter@mastodon.gamedev.place]

@rygorous @aeva which is at odds with the design goal of decreasing the levenshtein distance to making a meaningful change to the program.

i guess it boils down to: what's the worst harm a typo can do in your language?

funny then to consider the shell which has no types and permits casual mistakes of devastating proportions.

[aeva@mastodon.gamedev.place]

@Camwin oh yeah totally like, this is super obvious:

float Fnord(float A, float B, float C, float& D, void*** E, int F);

[darkyen@mastodon.gamedev.place]

@aeva @Camwin If this seems obtuse, wait until you encounter this monstrosity in a language without upfront types. The type constraints will still be there, but implicit, maybe poorly documented or hidden in asserts. No thanks 🙃

[morten_skaaning@mastodon.gamedev.place]

@lritter @rygorous @aeva can you make a checker that can give a correct good/bad answer on the validity of your program, when you insert random characters into the program source?

[lritter@mastodon.gamedev.place]

@morten_skaaning @rygorous @aeva isn't that a big part of every compiler?

[morten_skaaning@mastodon.gamedev.place]

@lritter @rygorous @aeva I talking about whether the outputs of the program are still useful, not whether syntax is correct

[aeva@mastodon.gamedev.place]

@morten_skaaning @lritter @rygorous can you check if it halts while you're at it?

[rygorous@mastodon.gamedev.place]

@aeva @morten_skaaning @lritter The Halting Problem is oversold.

1. It's not actually generally hard to prove whether any particular invocation will halt. (Just hard to give a single approach that works on everything.)
2. The HP is not even theoretically hard if you expand the set of valid answers to be { yes, no, not sure }
3. practically speaking, the answer is always yes because at some point there will be a sufficiently long power outage or, more likely, forced OS update

[jameswidman@mastodon.social]

@lritter @rygorous @aeva
...unless you're using nushell, in which case each command is type-checked before it is evaluated (and it won't evaluate if there are type errors).

[aeva@mastodon.gamedev.place]

@JamesWidman @lritter @rygorous that'll definitely save me from accidentally transposing the source and destination paths on various commands to exciting effect

[rygorous@mastodon.gamedev.place]

@aeva @JamesWidman @lritter see, if this was typesafe with proper constructors that could not happen:

cp(ThisIsTheSourcePath("foo"), ThisIsTheDestinationPath("bar"));