regarding packages and https://pkgs.racket-lang.org/pkgs-all specifically - I wanted to write a script that would notify me of new packages but there seems to be inherent problem to the design of pkgs - there is no “date-created” property only “last-updated” or “last-edited”.

The same problem is visible on the pkgs index - the packages that got just updated are flagged as new

which doesn’t reflect reality as there are many packages that are months old yet they are always flagged as new due to some authors scripts updating the packages from local repositories on regular basis (even with no changes)

The way I would like to see pkgs working is having both “new” and “updated” flags

The current package system does need improvement and there are several features I would like to see be added to packages too.

As it happens I am interested in working on the package system to improve it. There have been prior discussions about various wishlist items that need to be worked on.

I will create a post in the users/developers mailing list to check whether there are others working on the same thing and see about coordinating our efforts.

Regrettably I won’t be able to attend RacketCon this year. That would have been a perfect time to get into a deep discussion with other Racketeers so I’ll just have to make do.

Though it’s not the same - last year streaming of RacketCon allowed for some basic interaction for those of us who couldn’t attend. I hope this year will be the same (or better)

I’m not sure but shouldn’t these two terms be alpha-equivalent?

Sry to bug you with this, but I totally assumed that test—> uses alpha-equivalence by default to compare terms.

I double-checked my binding forms but I can’t quite figure out why these two terms aren’t equivalent.

nope, test--> doesn’t use alpha equivalence by default

you can set default-equiv

that’s exactly what I wanted to know, thanks.

or pass #:equiv my-alpha-equiv

let me check …

http://docs.racket-lang.org/redex/The_Redex_Reference.html#%28form._%28%28lib._redex%2Freduction-semantics..rkt%29._test--~3e%29%29

works brilliantly! many thanks!!

This package needs a “this package needs code, author, anything” label :simple_smile: https://pkgs.racket-lang.org/package/autoff

Hmm also the 4 “niceware” pkgs

and the “pointless” pkgs

@greg yes, these are some of these “always new” packages I was thinking about.

@githree There’s a lot of discussion about package catalog wishlist items on the github repo

https://github.com/tonyg/racket-pkg-website/issues

in particular, these two issues:

https://github.com/tonyg/racket-pkg-website/issues/50

https://github.com/tonyg/racket-pkg-website/issues/53

@githree The autoff niceware pointless packages have bad meta-data (lacking even author or code). Not sure if someone bypassed authentication to create them maliciously. Or some data loss glitch. But regardless, I think they simply need to be deleted? (As opposed to valid packages that are “new!” gratuitously frequently.)

lacking both an author and a source should definitely be grounds for removal by curators, since it’s not even possible to figure out who to contact about removing it through other means

I’ll open an issue for that

@greg https://github.com/tonyg/racket-pkg-website/issues/60

@notjack Thank you and sorry — maybe I should have just logged an issue there in the first place.

no worries, I’m glad you brought it up at all :)

@githree: Re RacketCon streaming: Yes, we’re planning to have streaming again this year.

I’m sad, I don’t think I’m going to be at RacketCon this year.

Oh no. :(

Too far?

Well I thought I could combine Strange Loop and RacketCon into one long trip.

But that wasn’t working out.

And flying out and back to each was looking non-optimal, too.

@greg: I can sympathize with too much travel, for sure. I’m missing strange loop this year and that’s a bummer

@notjack I believe I know who the author of pointless and niceware is: https://gitlab.com/users/ether-ash/projects

@githree fantastic!

I love the comments in https://gitlab.com/ether-ash/pointless/blob/master/pointless-lib/pointless/cond.rkt

Is there a way to customize the reader so that curly braces mean #hasheq? I basically want {foo "thing" bar "other"} to be equivalent to #hasheq((foo . "thing") (bar . "other"))

@githree Do you happen to know who that person is? I forget exactly why, but I remember seeing one of their projects at one point (maybe it was pointless?), and I was befuddled at their seeming lack of internet presence. :p

@cfinegan You could add a custom entry for #\{ to the readtable.

@greg You will be sorely missed at RacketCon. :(

@lexi.lambda I am afraid not, but if my goog-fu skills are of any use here is my hunch: https://twitter.com/ether_ash/status/880159791414476800

Hey slack, the SICP book uses the character “Alyssa P. Hacker” (a lisp hacker). What’s a good character for “a racket hacker”?

I use Alyssa P. Hacker for everything, tbh, regardless of language :stuck_out_tongue:

@githree Oh, thanks. I think I’ve seen that person on twitter, so I’m surprised I didn’t make the connection.

@joergen7 I asked something similar a month ago and never got an answer… May I see your default-equiv?

@greg you should come to Seattle! I owe you a coffee and/or beer

I made a package for this called delimit-app. It’s an #%app macro instead of a reader extension though, so you’ll have to quote the symbols

Hey thanks for letting me know about this. It’s pretty much exactly what I’m looking for - quoting symbols isn’t a big deal.