there is a blog post that covers this in detail

https://blog.racket-lang.org/2017/09/tutorial-contributing-to-racket.html

A small thing that would make life better is to automate the --lookup part, by just looking at the next catalog in the list.

Changes to the racket repo are pushed, so let me know if you run into problems.

Thanks @spdegabrielle , bookmarked. I thought I remembered seeing this but forgot it was a blog post and couldn’t find the link on the home page anywhere.

"<https://github.com/racket/racket/compare/89831c01f6f9…0d279ca06dc4|999 new commits> pushed to https://github.com/racket/racket/tree/master\|master"

Yeah, I was wondering if this is because there are really 999 commits or GitHub can count only up to 999.

The latter would be rather surprising

I ran make cs-as-is in an existing checkout and it failed: cd ../../../cs/c/../../ChezScheme && ./configure --pb --machine=ta6le --disable-x11 --disable-curses CC="gcc" CFLAGS="-g -O2 -Wall -DELF_FIND_BOOT_SECTION -pthread" LDFLAGS="-pthread" AR="ar" ARFLAGS="rc" RANLIB="ranlib" WINDRES="windres" No suitable machine type found in "boot". Available machine types: * Since no directory in "boot" exists for pb, you can try using Racket v7.1 or later with racket rktboot/main.rkt --machine pb to create the boot files, and then try ./configure again.

Well, that’s as expected. make as-is means “don’t download anything”, and you need to download pb. So, it’s a feature that this failed.

If you were stuck without internet access, you could configure with --enable-racket to not need pb, but probably that’s not what you want.

If you want to just download pb, use make fetch-pb.

I just did make cs and it works.

Can anyone help me parsing:

> The first two arguments name a procedure in a module that is loaded by dynamic-require is specially designed separate place. ? I re-read it for like 10 times now and I still don’t know what is the main verb in this sentence

From https://docs.racket-lang.org/tools/drracket_module-language-tools.html?q=drracket%3Amodule-language-tools%3Aadd-online-expansion-handler#%28def._%28%28lib._drracket%2Ftool-lib..rkt%29._drracket~3amodule-language-tools~3aadd-online-expansion-handler%29%29

Best guess? That second is should be in a.

That makes much more sense

Someone needs to do more Quining: > The first two arguments name a procedure in a module that is loaded by “dynamic-require is specially designed separate place.” or > “The first two arguments name a procedure in a module that is loaded by dynamic-require” is specially designed separate place. Both make a lot of sense to me!

I just tried make both and got /bin/sh: 1: RUN_RACKET_MMM@: not found

Question for @spdegabrielle and @laurent.orseau: if I participate in the quickscript competition, can I host the code in a repo instead of a gist? It will be contained in one file. I simply dislike gist.

Absolutely! I was planning to bundle all scripts from the competition in a package once it’s over, but if you prefer it to be in your own package that works too!

Another question, and this one is extremely important: can you help me coming up with the entry name? :joy:

If you write your own package, you should copy/adapt this file so that your scripts’ directory is registered in quickscript’s library and is readily available

sure :smile:

Do you have an idea yet?

It’s implemented

Sounds like “implemented” is the name?

:stuck_out_tongue:

So, it’s exactly like your dynamic completion, except that it uses static analysis so that only valid identifiers are shown

nice!!

Doomsday-device

better-dynamic-completion?

Mind-control-ray

I want to add a word that adds “uncertainty” to the name

dynaplete?

Like “dubious”

Or “fishy”

Because it’s actually pretty fishy lolol

Dubious-dynamic-completion?

Actually, if I name it fishy-dynamic-completion, is it qualified for the standard fish competition as well? :joy:

:rolling_on_the_floor_laughing:

hahaha, very good

Dynalex completion?

I don’t think I will do that, because I’m not planning on publishing it to the package server.

What’s “Dynalex”?

That also works with raco pkg install <github-link>

dynamic lexical-based completion?

Ah, sound cool

But if we include it in the competition’s package, you don’t need to make it a package yourself

Right, feel free to include it in the competition package

I don’t plan on including info.rkt

Btw, you can also use gitlab snippets and pasterack, but they are based on the same principles as gist (a little worse actually)

I guess what I really want is a README file and shows things nicely. Gist probably can do that, too, but its UI is worse. Also Gist is easier to be “buried”

Does it provide only ids that are in-scope?

That looks familiar. I think it’s a problem with how configure is run, or re-run, or not re-run

Yes

very nice

Could as well replace quickscript-extra’s default one

It’s definitely a situation of variables not being substituted in the racket/src/build Makefile

You can add markdown comments after the file, so the main difference is that you see the whole file before seeing the ‘readme’.

(You can also have revisions, and you have permalinks to all revisions)

Can’t wait to try it :slightly_smiling_face:

Will finalize in a couple of hours. I am making demos right now :slightly_smiling_face:

and also fixing any remaining bugs

very cool. Don’t forget to include a line like (script-help-string "This is what my script does") after the require lines so that this help-string can be displayed in the library and elsewhere

Yep, I did!

It looks like continued-fraction doesn’t use contracts at all so it may be something else.

Now fixed, hopefully.

With your change, make both still had the same error, I believe because the configure script didn’t know it needed to re-run. I did a touch racket/src/configure, and then make both worked.

That makes sense. I only changed things to avoid creating the bad state.

hopefully no one else did my specific sequence of commands in the last few hours and thus it won’t matter

Quickscript Competition Deadline Extended till Tuesday Morning 4 Aug (uk time) https://github.com/Quickscript-Competiton/July2020entries\|https://github.com/Quickscript-Competiton/July2020entries

@laurent.orseau uploaded and entered

feel free to try it :slightly_smiling_face:

Also bad phrasing because the first argument is the module, but it is the second argument mentioned. “The first two arguments name a module and a procedure, which are loaded by dynamic-require into a specially designated separate place.”