@apg I turn some of my talks into packages; links here: http://www.eecs.northwestern.edu/~robby/talks/

@apg also @mflatt posted the source for all of his slideshow talks here: https://github.com/mflatt/talks

@samth: I’m about merge TR PR 540 (from @georges-duperon). I noticed that the merge button seems to default to “squash and merge”. Is that a repo setting you set?

FWIW, it’s not what I want here (squash + rebase + merge is what I want), but switching the default to “rebase + merge” on the racket repos would be nice.

@stamourv yes, that’s a setting

I think “squash and merge” means “squash + rebase + merge”

and “rebase + merge” doesn’t squash, but sqaush + merge without rebase doesn’t make sense AFAICT

Ah, found it.

Disabling merge commits for racket/racket.

Doesn’t seem to be an org-wide setting. Will write a script to do all of them later.

@mflatt Is it okay if I merge this: https://github.com/racket/scribble/pull/111

@leif I left a suggestion (version and history) in the PR

@mflatt: Could I ask you to take a quick look at this PR? It should be brief; the changeset is very small. https://github.com/racket/r5rs/pull/2

Someone else would probably also be able to review it, but it looks like you’re the primary maintainer for that repo?

@lexi.lambda Done. (Does anyone have a guess at why GitHub sends me email alerts only sometimes?)

Thank you!

@mflatt: I have email notifications turned off for watched repos but turned on for threads I’ve participated in. What do your settings look at for the Notifications section in your GitHub settings?

@mflatt: There was a point in time where not all our repos had email integration, but I fixed that a couple (few?) months ago.

I have everything checked except “Include your own updates”

I haven’t yet tried unchecking “web”; maybe N browser tabs on GitHub confuse it?

I have no idea what the issue is then. ¯_(ツ)_/¯

@lexi.lambda: From the git history, it looks like you’ve been involved with the homebrew packaging for Racket. Is that correct?

I think I touched it once to update it

That’s more than I have. :)

but it was a very easy change, IIRC

Any idea about https://github.com/racket/racket/issues/1680

?

ISTR we’ve had other reports, possibly about similar things, with homebrew before.

I’m not sure, but I could make guesses. I don’t (and never have) run Racket from a Homebrew installation

Do you know how the main person behind Racket’s homebrew package is?

I’m afraid I have no idea.

@stamourv: It sounds like Homebrew renamed the formula to minimal-racket, but whatever the installation is does not properly set the relevant prefix. I’m a bit surprised by this, since the formula appears to pass the right --prefix option.

Hmm, that does sound odd.

It looks like @leafac did the initial rename of racket to minimal-racket, so perhaps he knows more about this issue.

@lexi.lambda: Talking to @florence, it seems like there’s a cask for Racket, that uses our installers. So probably best to use that, instead of the recipe.

So we may be able to just ask for the recipe to be removed. I’ll post on the mailing list.

@stamourv I think some people have expressed a preference for formulas over casks

not that I understand this at all

I think most people also prefer working things to broken ones. :)

I think having the minimal-racket installation is a good idea if we can get it working. It mentions in the description to use the cask if you want the full installation.

Yep, would be nice to get it working. But unless you’re volunteering, it may be better to remove it. :)

@samth @mflatt: I can take a look at the broken test

@samth, @mflatt: it seems like the relevant file, psyntax.sch is generated from psyntax-input.txt? is that true? if so, how do I regenerate it?

I think by running psyntax :wink:

but we don’t want to be breaking that code, do we?

@samth: psyntax-input.txt has a note that it was modified from the original version, and the offending line is a modified line that uses [ ] instead of parens.

so my guess is that modification was basically a mistake and it should be replaced with parens.

ah ok

@samth: fixing psyntax-input.txt makes raco test -l tests/racket/benchmarks/common/psyntax pass, but it doesn’t change psyntax.sch, so I’m not sure if that’s a problem or not.

I believe the problem (based solely on the change and what I think might happen) is that psyntax.sch calls read to read in psyntax-input.sch

and your change caused it to fail on the [] there

so we shouldn’t need to change psyntax.sch at all

@samth: what confuses me is that psyntax.sch is clearly a generated file, and it also has a single pair of square brackets that appear to correspond to the single pair of square brackets in psyntax-input.txt

those are read by the implementation of read inside include in scheme/include (see psyntax.rkt) which will accept them

right, but I’m still left wondering how psyntax.sch was created in the first place

probably by running psyntax on a system with different treatment of []

such as racket/r5rs before your change, or chez, or something

what I’m really getting at is that it seems like maybe psyntax.sch ought to be kept in sync with psyntax-input.txt and I don’t know how to do that

but if you are okay with just changing psyntax-input.txt so that the test passes, I have a patch I can push

That’s my suggestion

alright, pushed

@robby @samth thanks for the links to slideshow talks. I gave a successful talk last night with slideshow (not on racket), but wasn’t quite happy with the way the slides looked. Was happy with the tool itself though.

OK, so I guess what I didn’t realize about slideshow, but now do, is that it doesn’t do much more than create the UI and handle keyboard events, and provide some very naive things. Basically, it gives you the minimal scaffolding in order to build custom slideshows with the whole power of pict.

now that I have that understanding, it’s even better.