@KraZug as a tip, talk about how you use Mathematica, what domains you have expertise in, etc... It's technically "invite only", so maybe worth trying to convince them for the invite :)

Here's something fun: resources.wolframcloud.com/FunctionRepository

@user6014 @KraZug I can't imagine they'll be that picky though... pre-release testing is basically doing work for them that most companies legitimately hire people to do. A friend of mine who works at a well-known tech company was like "Wow... WRI is scamming people" when I mentioned the pre-release testing system.

On the other hand, we're all altruists here :)

Oh and fascinating. The equivalent paclet repository link: .../PacletRepository redirects to .../app/objects/resourcesystem/published/PacletRepository/index which suggests the paclet repository is actually coming into existence

Good to see

Oh...nevermind

CloudObjects["user:resourcesystem/published/FunctionRepository"]

CloudObjects::notperm: Unable to perform the requested operation. Permission denied.

$Failed

CloudObjects["user:resourcesystem/published/PacletRepository"]

CloudObjects::cloudnf: No CloudObject found at the given address

$Failed

@JohnFultz any possibility you can comment on any type of paclet repository as it pertains to open-sourcing of things in the language?

I just discovered a very odd parser glitch. Try the following

myList = {1, 2, 3, 4}

and then

myList[[{1, 3}]

No error message no nothing. When you try this with a command line kernel, it waits for additional input like when you type 1 + and press enter. However, this seems rather strange since ]] can never be used with a linebreak in it.

Which brings me to my last #dailyfun: It is rather strange that this works:

Having said all this, can you guess the value of i in this example?

(which btw evaluates as well without any error)

@halirutan I think this is because [[ can eat any whitespace. Like this works:

myList[
  [{1, 3}]
 ]

also for the second one I don't get to see that glitch :/ It won't evaluate for me.

@b3m2a1 Really? I'm on Linux with 11.3 and it eats it without any complaint. And that [[ can eat whitespace is awful, but that's probably a very personal opinion.

@halirutan yeah I don't like it. I just remember being slammed by it constantly back when I first started using Mathematica a few years back. Coming from more standard languages where [...] is the indexing operator I kept messing it up and accidentally inserting whitespace too

It was pretty nasty

Like this: myList[ [ 1 ] ] just looks terrible and is so much less readable

Honestly the [[...]] also looks bad in my mind, but I understand why it's necessary.

@b3m2a1 Yeah, Mathematica is very bracket-heavy. But I wouldn't know a better solution that doesn't look terrible after so many years with the syntax.

@halirutan I agree. Once you introduce a b as a multiplication you don't really have much choice

I wonder how the Mathematica parser works... it seems like it strips all non-string whitespace as one of it's early steps as it's fine with this: a [ x ] which surprised me

@b3m2a1 I can only comment on my parser and whitespace multiplication was a bit tricky. Especially, since it is not consistent.

@kirkus, @user6014, @b3m2a1, thanks. I might send an email then.

For anyone who uses Internal`Bag I took a few mins and wrote up a convenient interface for it to make it look more like List (although of course it isn't). You can find it here

Here's how it looks:

eb = ExprBag[];
Length@eb

0

AppendTo[eb, 1]; // RepeatedTiming

{0.000011, Null}

Length@eb

5655

b = ExprBag@Take[eb, UpTo[25]];

Take[b, UpTo[26]] // Length

25

b[[5]] = 5

5

Normal@b

{1, 1, 1, 1, 5, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, \
1, 1}

And we can see that AppendTo always takes the same amount of time on it, unlike with List:

testBag = ExprBag@Range[100000];

AppendTo[testBag, 1]; // RepeatedTiming

{0.0000107, Null}

lll = Range[100000];

AppendTo[lll, 1]; // RepeatedTiming

{0.0001, Null}

Weird that there's no built-in way to convert GeoPosition etc to WKT

WKT = docs.geotools.org/stable/javadocs/org/opengis/referencing/… ?

@b3m2a1 Is there a Internal`Bag counterpart to deleting from a bag (or list) without copying the list?

Additions / improvements to this are welcome:

@b3m2a1 they aren't that picky, after all, they let me in...

@b3m2a1 It's not scamming. Mathematica is a programming language. It is not just an end-user product like a word processor. People build other software on top of it. These people want to verify that a new version won't break things.

Imagine if Apple stopped giving developers a beta version of macOS, then when a new macOS is released, 10 times as many programs break as usual

MATLAB has a prerelease programme as well, though they usually give us only one version.

It's not merely about fixing your code to make it compatible. It's about WRI fixing stuff in Mathematica that broke your code. Before it's too late.

@b3m2a1 Some components of M are now open source here: github.com/WolframResearch

@ArnoudBuzing yes

well, maybe not

easier to read: en.wikipedia.org/wiki/Well-known_text

particularly the geometry stuff

@Szabolcs I meant in terms of an actual paclet repository. He mentioned pushing for an actual on in that post way back when.

I appreciate the stuff on GitHub, though. Useful to look through.

How do i view the defaults for some method in mathematica. Specifically, Arnoldi i want to know the maxiterations and tolerance..

i havent been able to locate it in the documentation