SC2Mapster

Tried to get a jungle feeling as in the cinematic. An then some crashed ships and bones ...

The idea is very nice. But my main concern is, that map makers and players could abuse the tool to send you fake bank files. Map makers could do this in order to get their map to the top on your page. Players could fake their scores.

Quote from xSun: Go

I assume you mean gamescom, not blizzcon?

Yes, thanks for the hint.

At gamescom we (instarcraft.de, a german fansite) had a chance to do an interview with Jonny Ebbert, Senior Designer for Heart of the Swarm, and Bob Colayco from Blizzard's PR team. Here are some quotes which are related to maps and custom games.

You can read the full interview here: Jonny Ebbert interview (or read the german translation).

Quote:

Can you talk about some of the new features for the Galaxy Map Editor in the next expansion?

Jonny Ebbert: I'm not sure if I can. We are working on things for the mod community, but don't have to announce anything. We are constantly working on the editor, not only for the fans, but also for us. So it's constantly work in progress. We see the mods that are cool and so amazing and we do hear about the limitations and frustrations. We do listen to that and when it is possible we try to add this things to the editor.

Quote:

Features like Waaagh!TV where you can just jump into a game and watch it?

BC: (smiling) That stuff is always on our internal wish-list, but it is always a matter of prioritizing: What can we do? How may resources do we have for this instance? So it is a matter of prioritizing which things first. We can say that the Market Place related stuff is really improving the custom game search. Players [will] being able to weight their new custom games, making that whole process of finding a good mod better. That's at or near the top of the list right now. Everything else is just kind of aligning after that.

JE: What's really awesome on Battle.net is that we don't have to wait for a product to come out to release the improvement. So if something doesn't get in with Heart of the Swarm, we don't have to wait until Legacy [of the Void] to ship it.

Quote:

Can you give us some information about Blizzard DotA? Do you plan a release this year or with Heart of the Swarm?

I can't tell you about this. What I can tell is that we're working on Blizzard DotA. We will definitely have a lot to tell you at BlizzCon.

So do you think that Blizzard DotA could compete with DotA 2?

That's for you to decide. We will release it and then the fans will decide how awesome it is.

BC: One of the things that we can say is: If you guys were at BlizzCon last year and saw it and maybe played it, it's a lot different now.

JE: I would say completely. We completely wracked it. We basicly redid it after last years BlizzCon, so yeah, expect a lot of changes.

Quote:

Also between the missions, so that there will be something for the player to do?

Oh, definitely. There are kind of three things we are focusing for. First one, we want to make each mission as exciting ans unique as possible and have a tone-like mechanic that is special to that mission. The other thing we wanna do is to give you even more compelling choices in each of the missions than you had in [Wings of] Liberty. So Liberty had like Tech purchases and Mercenaries, it has research and in Liberty you have the ability to upgrade your units. Here you have the ability to split your units into new sub-species. You also have the Kerrigan powers that you can choose between each mission and there are possibly some other things we haven't revealed yet.

But no „Last Viking?”

BC: That game on the arcade machine...

JE: Oh... I'm not sure if the Zerg should have an arcade machine.

Maybe an infested arcade machine...

Yeah, an infested arcade machine. Actually I should write that down. (laughs)

Quote:

In StarCraft II the observation system with the stats and everything was a huge enhancement compared to your previous games. Can we expect further eSport enhancements for the community in Heart of the Swarm? For example the ability to watch you replay together with many other people is missing or the ability to follow a competition like it used to be with Waaagh!TV.

JE: We will not tell you about that until BlizzCon, like new features for eSports or things like that.

BC: It's kind of a similar question like what we talked about earlier: All of those things, they are on the list of things that we want to do. It's just a matter of prioritizing. Just because we have an expansion doesn't mean this is the only chance we have to cram these things in. We patched in a lot of improvements to Battle.Net and we added the chat rooms, we had some new content with the custom maps. So those kind of features can come in between releases and then post. So I think there is many more room to add some of those. At the top of the list right now is improving the custom game experience.

Quote:

A lot of people do criticize the ladder map pool. What are the reasons for rather making your own maps instead of using community made maps for the ladder pool?

BC: I think the way we look at the map pool and the ladder is that... I think a lot of the hardcore community, they want every map to be a macro map. They always want that third expansion to be easy to take and so every game has to get to 200 food. But that is not necessarily what - you know StarCraft has sold more than four and a half million copies - that is not necessarily what everyone wants. Sometimes you want to play rush maps, sometimes you want to play regular size maps and sometimes you want to play macro. If you look at the map pool as a whole it kind of gets all those points. The hardcore tournaments make their own maps and then they integrate that into their high level play and that's ok, that's ok form them to do that.

So might it be a solution to just increase the amount of maps in the pool and give the players more vetos?

JE: The map pool is not random. We cycle out maps we don't like for some reasons and we add maps for other reasons. And also the map pool has a size that is for a reason. We kind of consider them as an experience for players that pushes players in different reactions. If the map pool is gonna push you out of your comfort zone. Some people like to turtle, some people like to rush and some people like to hard-macro but we're gonna give a variety of maps and there is gonna be a certain number of each type of map. And the other thing is: We do take into account certain things in the metagame that are unfair across different races based on the maps. So if we wanted to use a player map, we need to make sure that it meets that criteria like “Can the Zerg defend all rushes?” ... The placement of the Watch Towers is also not random, it is very clear what is revealed with these Watch Towers and what's not. That's kind of the reason, but I also can't say we would never release a player map. If a player map is so awesome, maybe we would one day say “Hey, we're gonna add a player map”.

I think combined with the new patch notes this really shows that blizzard is trying hard to improve the custom game section of sc2 :)

There already is something like grimoire for sc2. Its calles Phyngal and you can get it here: http://phyngal.xgm.ru/ The hack also disables login to battlenet for the editor so I dont think that blizzard has a chance to detect it.

Last time I tried it it worked fine and it is quite easy to integrate an advanced parser into it as long as it can be run with one command. But I think it is difficult to integrate a new script editor into the original editor like TESH did for wc3.

You should try Hero Line Wars by Kueken: http://www.sc2mapster.com/maps/hero-line-wars/

It is somehow similar to Enfos.

@Nebuli2: Go

This video is not fake, it is made by Blizzard :D

The following code compiled:

unit foo();

int foo() { // function prototype not implemented
	string x;
	int z = 1.234; // wrong types
	if (false) {
		return x; // wrong return type + use of not initialized variable
		foo(); // dead code
	}
	bar();	// calling function which is not yet defined
	// missing return
}

int bar() {}

Quote from wangbot: Go

I wouldn't say Andromeda is bugged to the point of being unusable. Especially since I've coded two separate maps with it (and one of them uses Andromeda's more advanced OOP features quite extensively).

Hm, maybe it was just the way I intended to use the tool that was bugged ;) (updating the table)

But still I would not use it currently because the existing bugs might never get fixed.

Organize code: Mainly having namespaces, packages, modules or something like that. Ideally you could explicitly define which other packages are imported (usually import statements) and what is exported from the package (usually by private/public keywords)

Syntactic sugar: Things which do not really add something to the language but add new syntactic elements that allow to write shorter or cleaner code. A nice example is a foreach loop. A not so nice example (imho) are the pre- and post-increment statements.

Easy: How easy is it to learn and use the language? A language becomes more difficult if it has more features but there are also other things. A language is easier if it has no cryptic names or symbols and disallows statements which are hard to understand. It is also easier if the compiler can catch many possible errors at compile time. For example this can be archived by generics which are supported in andromeda.

Currently Galaxy+ + is the only language extension which is usable, I think. But I do not like it very much so I would wait for Grape or grim001s Languages or for gex to finish his master thesis and start working on andromeda again. I tried to make a table comparing the languages from my point of view:

How will overriding and overloading methods work? Will it choose the method based on the runtime type or based on the compile time type? I think using the dynamic type is the most intuitive solution but most languages I know use the static type.

Example with dynamic binding:

class A: 
	string foo():
		return "A"
		
class B inherits A:		
	string foo():
		return "B"

class C:
	static string foo(A a):
		return "A"
		
	static string foo(B b);
		return "B";
			
class Test:
	A a = new B();
	string x = a.foo() // x is now "B" ?
	string y = C.foo(a) // y is now "B" ?

I also thought a little bit more of exceptions... What if you have a function which can throw an exception in a parameter. Take the following example where c is a function which can throw an exception.

if a() and b(c(), d()):

So if c throws an expception d must not be executed because it comes after c. So to check for exceptions you would have to move c() in front of the if. But if you move c there it will always be executed, even if a() evaluates to false. To have it all correct this little if-statement would have to compile to something like this:

temp1 = a();
if (temp1) {
    temp2 = c();
    if (exceptionThrown()) {
        return; // if this statement is not in a try- block we can just 
                //exit the function, otherwise it would be even more complicated
    }
    temp1 = a(temp2, d());
}
if (temp1) {
    ...
}

I think it is quite tricky to get this transformation always right. Maybe the code generation should generate some kind of three address code as an intermediate step.

I really like the part where she sings "Dir ist alles erlaubt und alles gegeben" because when everything is allowed and everything is given, everything is very easy. So I always listen to this song before examinations and other challenges.

Quote from Vestras: Go

@peeeq: Go If you want, you can implement it; after all, the project is open source.

Maybe I could help a little bit with the code generation or just by writing some automatic tests. Can I just checkout my code from http://code.google.com/p/grape-sc2/ and commit my changes there or how is this organized? Do I need any tools besides Visual Studio to build the project (I have not tried it yet)?

Some more questions/suggestions:

1. Will it be possible to have functions as types so that you can pass functions as parameter?
2. If 1. will be possible: Will it also be possible to define anonymous functions (lambda expressions in python) or nested functions?
3. What happens if you take an item out of the list and cast it to a wrong type? Will there be a dynamic type check or is it just the responsibility of the scripter to get the types right? Will it be possible to cast objects to ints so they can be stored in a units custom value?
4. In your example are some argument_null_exceptions. Maybe it would be better to not have null in the language. This would make all these checks needless. For cases where null is needed there could be a special type. For example it could look like this

unit z = null // compile error
unit? z = null // ok
unit?[10] units //question mark = this type is nullable
//...
unit u = units[3] //compile error, wrong type
unit u = units[3].Value // use "Value" to get the actual value, exception when value is null
if units[3].HasValue :
	unit u = units[3].Value //now it is safe to get the value
	unit? v = u // automatic cast to nullable type


list a = new list()
list? b = new list()
// ....
a.add(1) // cannot throw nullpointer exception because a cannot be null
b.add(2) // can throw nullpointer exception
b?.add(3) // safe invoke, no exception is thrown if b is null, it just has no effect

See nullable types as in fantom: nullable types & safe invoke or nullable types in c#: http://msdn.microsoft.com/en-us/library/1t3y8s4s%28v=vs.80%29.aspx

This would make the language more complicated but a lot of functions would look nicer without null checks. For the first versions of the language you could just include the syntax as a kind of annotation for the programmer and later add the type checks. At least for non primitive types this should be no problem.