laurent.orseau
2021-2-14 08:22:41

You can migrate packages instead of reinstalling, raco pkg migrate


laurent.orseau
2021-2-14 08:25:33

<https://gist.github.com/Metaxal/13c4effc483c5ef9f5797ace4950f97a|This script> can automate this in most cases


laurent.orseau
2021-2-14 08:26:20

But it’s only for migrating packages. For the racket distribution itself you’ll have to reinstall


cdimara
2021-2-14 15:22:51

@cdimara has joined the channel


samdphillips
2021-2-14 18:12:24

Also, on Windows raco pkg migrate is integrated in the DrRacket Package Manager


sorawee
2021-2-14 18:53:33

Probably on Mac and Linux too?


samdphillips
2021-2-14 19:08:45

Yes of course. For me Windows is a second class shell experience.


sorawee
2021-2-14 19:17:56

I haven’t been using Windows for very long time, but I thought WSL and/or Powershell is supposed to fix that?


samdphillips
2021-2-14 19:22:14

IME WSL is like being in a VM. I think Powershell is probably better.


dyllongagnier
2021-2-15 06:32:05

Is there any situation in which

(module ex racket (define x x)) does not result in a runtime error when you try to require it?


dyllongagnier
2021-2-15 06:35:24

I’m trying to understand why this isn’t flagged as invalid syntactically. Is it just because leaving it till runtime simplifies Racket’s implementation or can this code somehow be required without raising an exception?


dyllongagnier
2021-2-15 06:42:39

In theory, once you have expanded macros, it should be possible to find these errors without running the program. The LHS side of a define needs to syntactically be inside of #%plain-lambda or a case-lambda for the program to ever produce a non-error value.


laurent.orseau
2021-2-15 06:44:27

It may be related to optimizations performed for single expression modules, which are spliced into the requiring context, or something like that. There was an answer by mflatt about this not so long ago