If there’s a way to do this that plays nicely with compilation, I bet it would involve syntax-local-lift-expression so that the counter initialization code can execute once at module initialization time.

Is there a particular reason ~r raises an exception when given +inf.0?

(I always have to write a wrapper or deal with special cases inline)

@phanthero It’s not really set up for that. If you change racket-program to an ssh link, can emacs make-process run that remote and pipe input/output? I don’t know.

Oh, also, the Racket back end will be trying to open various files, and if it’s getting tramp ssh links it won’t know what to do with them.

There are a couple customization vars, racket-path-from-emacs-to-racket-function, and vice versa. I’d added them to support a scenario of using Windows Racket.exe from a WSL linux Emacs. But they might help here.

Like, maybe you could define a function that strips the tramp ssh prefix.

I guess the TL;DR is that currently this is an “off-label use” of Racket Mode. Maybe you could get it to work but I’m not sure.

If you want to open a feature request issue on GitHub, I’d be glad to look at it in a month or two when I might have more bandwidth. Or if you figure out it works with XY change to Racket mode, I’d be happy to look at a pull request.

Do we have a github repo that collects #lang/slideshow presentations?

Okay, thank you so much for the great extension and advice.

Have any Racket on Windows users seen a problem running Racket.exe or Raco.exe when the current directory is quite long (but AFAICT still within the Windows max path len)?

Context why I’m asking: https://github.com/greghendershott/racket-mode/issues/505

Or to skip most of a long comment thread: https://github.com/greghendershott/racket-mode/issues/505#issuecomment-724799416

I feel like there were some changes that might be relevant recently, so testing a new version might be worthwhile

I’m happy to test. My work laptop is Win10. I run Emacs 28 and Racket 7.9 BC.

Win ha ha of the day, it made the first directory, but complained about having to display it. mkdir xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Exception setting "WindowSize": "Window cannot be wider than 170. Parameter name: value.Width Actual value was 200." At line:1 char:281 + ... { $rawui.BufferSize = $bufsize; $rawui.WindowSize = $winsize; } else ... + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + CategoryInfo : NotSpecified: (:) [], SetValueInvocationException + FullyQualifiedErrorId : ExceptionWhenSetting Directory: C:\Users\knauthg Mode LastWriteTime Length Name ---- ------------- ------ ---- d----- 11/10/2020 11:58 xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

oh, it never actually made the first directory cd xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/ 11/10/2020 12:00:46 KNAUTHG-SB ~\xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx\xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx PS> pwd Path ---- C:\Users\knauthg\xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx\xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

That’s a good lead, thanks @rokitna

Didn’t get any output at all. Left comment on github.

Thanks @gknauth. I’m juggling a couple of laptops today and already a few git branches for Racket Mode, so, although I can/will get around to updating Racket.exe on that laptop, it was really nice of you to try that and report back.

No problem. Continuing conversation (for now) on github thread. I’m “accu-knauthg” there because I’m a different person when I’m at work (but not really :slightly_smiling_face: ).

#lang racket (define a 10) (match 'hi [(? a) 1] [(? number?) 2]) Currently, match doesn’t report which clause or which clause’s LHS is problematic. @mflatt is it possible to improve match so we can get more specific error messages?

Since even errortrace doesn’t give more specific information than the match form, probably the match macro needs to provide more source locations in its generated code.

I’ve pushed a repair; it was a really dumb buffer-allocation mistake.

As penance, I updated Racket CS to handle the Windows 259-character limit on traditional-form paths by converting to \\?\ form, like BC does. And I discovered a wonderful new twist, which is that a path used as a directory is limited to 247 characters, not 259.

@capfredf https://github.com/racket/slideshow/blob/0fd065a2ac995b6e692a20704528b59672a64fb9/slideshow-lib/slideshow/slideshow.png

Ok, so this is made by some japanese people for <http://vim-jp.org\|vim-jp.org>, which also has a slack: https://github.com/vim-jp/slacklog-generator

This allows even free tier slack to be generated and stored to HTML Thought it might be useful for a community like this. There have been some similar things made by the English speaking community but I’m not sure if they work

Yep, it seems all the archives for every month of each channel is located at this site: https://vim-jp.org/slacklog/

I was wondering if someone here more experienced with Go, Make, Python etc could create something similar for us :slightly_smiling_face:

@spdegabrielle I am confused. Why makes this image special?

Yeah, there is presumably an application created by the ? form that should get the source location of the a

Thank you! This is what broken windows policing should have been.