<SeanTAllen>
irc is the official real time support channel
<SeanTAllen>
we have a small enough community, thank your for the thought of doing a gitter Rishav but at this point in time, its just a source of confusion
Praetonus has joined #ponylang
<Rishav>
oh well. I understand the reasoning. But i feel that gitter is vastly better than IRC. doing things like pasting code is very hard on IRC :(
<SeanTAllen>
we do a community survey each year, the overwhelming majority of people wanted IRC not gitter or slack
<SeanTAllen>
I'd ask you to either label it as unofficial or delete it. Given that we are a small community, my personal preference would be for deletion until such time as the community as a whole asks for something other than IRC.
<Rishav>
ok. I deleted the room. Thanks for being patient with me. @SeanTAllen, is there any place where i can read/monitor the future roadmap of Pony?
<SeanTAllen>
there's no official roadmap at this time beyond "get everything stable and free of bugs for a 1.0 release", what that means is still up in the air. its something we need to address. One will come, but that's going to take some time.
<SeanTAllen>
in the end, pony development is driven in a large part by people scratching their own itch.
<SeanTAllen>
There are no paid developers so it's whatever people want to contribute in the way of bug fixes and RFCs.
<Rishav>
one last question. what comes after 0.9? 0.10 or 1.0 ? ^_^
<SeanTAllen>
Or 0.9.1
<SeanTAllen>
0.10
<Rishav>
>.>
<SeanTAllen>
Either of the two
<Rishav>
lol. thanks
<Rishav>
i am going through the tutorial. my mind is reeling from all the new stuff
<Rishav>
You have a great day and see you later!
<SeanTAllen>
you too
Rishav has quit [Ping timeout: 260 seconds]
Rishav has joined #ponylang
Rishav has quit [Quit: Page closed]
<ohir>
uh! slack, gitter, phpBB
* ohir
skimmed over chan
c355e3b has joined #ponylang
jemc has joined #ponylang
amclain has joined #ponylang
rurban1 has joined #ponylang
TwoNotes has joined #ponylang
<TwoNotes>
Is there any syntax for a literal empty Array[String]? It does not seem to like []
<jemc>
no, I think you have to just say `Array[String]` to create a new empty array
<TwoNotes>
Or I could use None, if it was legal to have a union type as a parameter to a behavior?
<jemc>
sounds legal to me - are you getting an error?
<TwoNotes>
If I use a 'type Foo is (A|B}' then deflcare a behavior taking Array[Foo] I get an error
<TwoNotes>
WHat I really would like is a behavior with a parameter of Array[( String | (String,Array[String]))]
<TwoNotes>
That is, an array of items, each item being either a String, or a tuple (String,Array[String])
<TwoNotes>
But it deems 'behaviors' are more restrictive than simple 'funs' as what their parameter types can be
<jemc>
the parameter types must be sendable (`val`, `iso`, or `tag`)
<jemc>
in this case you probably want to specify `Array[Foo] val` or `Array[Foo] iso`
<TwoNotes>
Yes, I left out the access modifiers
<TwoNotes>
But that is not what the message is about
<TwoNotes>
String val is not a subytpe of (String val, Array[String val] val): the supertype is a tuple
<TwoNotes>
Declaration is be speak_multiple( texts: Array[(String | (String, Array[String] val))] val ) =>
<jemc>
ah, I see - you can't use a tuple in the union like that
<jemc>
the tuple doesn't carry a type descriptor, so it can't be disambiguated in the union
<jemc>
however, I believe you *can* use a tuple union if they have the same arity
<jemc>
s/arity/element count/
<Praetonus>
I don't think that's the problem
<Praetonus>
Tuples should be allowed in unions since they're boxed in that case
<Praetonus>
TwoNotes: Do you have a minimal example of your problem?
<TwoNotes>
I will write one up. This code is currently buried in something else.
<Praetonus>
The first error is because you're trying to capture a val as a ref, which is forbidden
<Praetonus>
The default capability for Array is ref, so you'll have to explicitly state that you want a val when declaring the type of insert
<Praetonus>
The second error is because the type of the elements in an array literal is the most specific type possible
<Praetonus>
Here, your literal has type Array[String] and this isn't a subtype of Array[(String | (String, Array[String val))] val
<Praetonus>
To solve the problem, you'll have to specify the type that you want for your literal: [as (String | (String, Array[String) val)): "A", "B"]
<TwoNotes>
ok, I will try that
<TwoNotes>
Though this defeats the purpose of trying to make the notation as concise as possible..
<Praetonus>
Yeah, I think it would be possible to make the compiler smarter here to make use of the final use type when inferring the element type
bougyman has joined #ponylang
bougyman is now known as Guest52468
Guest52468 has quit [Changing host]
Guest52468 has joined #ponylang
_andre has quit [Quit: leaving]
Praetonus has quit [Quit: Leaving]
rurban1 has quit [Quit: Leaving.]
prettyvanilla has joined #ponylang
prettyvanilla_ has quit [Ping timeout: 265 seconds]
<TwoNotes>
What does "Value formal parameters not yet supported" mean?
<TwoNotes>
for item in a.values() do let key=item[0] Where 'a' is Array[Array[String]]
<TwoNotes>
It complains about the reference to item[0]
Guest52468 is now known as bougyman
TwoNotes has quit [Quit: Leaving.]
<doublec>
TwoNotes: you want item(0)
<doublec>
[...] is generic syntax and values in that aren't yet supported