Start Game ... Without an IDE

[Nostalgic]

A bit of a generic question, but with specific intent:

Does anyone have good instructions about building an SCI game without using either SCI Studio or SCI Companion? (i.e., Creating a game from scratch?)

Now I already know the immediate response: why would you want to do that?

Well, that's a good question. Here's the answer.

I figured out how to do this -- mostly -- with SCI Studio but it's so cumbersome that it's hardly worth the effort. So why the heck am I doing it?

Because I'm trying to get some developers and testers interested in game development as a means to teaching development practices and testing practices. As most of us here probably know, programmers often like to see how something is built up from the ground up. (That's the common complaint with using things like the MFC, for example, or the .NET Framework -- too much is hidden. That's great when you're an expert; not so great when you're learning.)

Part of this will also be a hope in getting people to look at crafting ancillary tools to work with SCI technology. That said, many of these people, myself included, are interested in how much you can build independent of any tool and then use one of the existing tools (or perhaps create new ones) to accomodate what you've built. (This is similar to how you can -- and should be able to -- write Java without having to use the NetBeans IDE or, alternatively, write C# without having to own Visual Studio 2005.)

Also, part of what I'm doing is writing up a new set of tutorials that are much more involved than anything I've seen before for SCI, including a much more in-depth explication of the language. Part of that explication will be based on an understanding of how an SCI game is built literally from the ground up, without the Class System in place initially.

So hopefully that explains my rationale a bit. Like I said, I think I found a more or less successful way to do this outside of SCI Studio -- have not tried with SCI Companion yet. It involves using the SCI Script Compiler (scc.exe) and a lot of moving of files around.

Thanks in advance for any thoughts ... even if those thoughts are just to tell me that I'm clearly an unbalanced individual who likes to make things hard on himself.

[Eigen]

I'm not sure I quite follow but here's some thoughts:

# First you'd need to write an app to add/extract individual resources from resource files.
# Then you'd need editors or convertors for each resource type
# Also you need a script compiler together with an editor

[Nostalgic]

Understood.

Actually, I wasn't clear at all. How I should have presented it was like this:

When you start SCI Studio, you can see that you have the option to create a blank game or one based on the template. Creating a blank game appears to lead you into a dead end. SCI Studio, while offering the option, apparently doesn't actually let you do much, begging the question of why the option is there. (I know I asked it, so I know others will as well.)

So what I wanted to do was get the bare minimum built -- outside of SCI Studio -- that could then be "imported" into SCI Studio. Ideally a full tutorial would actually not only show you how to build the basis of the class system on your own but to also build alternative class systems, which ultimately I think would be an interesting thing to shoot for.

In any event, I use scc.exe to compile the most basic Main.sc that I think you need to create. But this needs Vocab.000. So you have to have that, which is something that you're not told, you just have to futz your way through it. Then you have to copy the resulting files to your new game directory and try to import them into SCI Studio. I think that mainly works by making sure you import the vocab resources first (000, 996, and 997).

However, more often than not you get a resource error when you do this. I think I've found that if you delete one of the resource files and then let the game build again, it works. Sometimes.

It just seems very kludgy and that worries me because it suggests there's a heck of a lot going on behind the scenes (in SCI Studio) that isn't clear. And I like to separate out the SCI game creation part from the SCI Studio part, at least while learning. It also helps me see where improvements can be made in how SCI Studio does things.

So ... all of this clod-hopping around on my part led me to wonder if anyone has trod the same path before.

[Nostalgic]

Isn't it always the way? You post ... and then you figure it out.

So, I think I got it.

I'm able to create a game entirely from scratch and then import it into SCI Studio with the minimum of fuss.

Sorry for this thread. And the confusion. And the meandering thoughts.

[troflip]

I think you're headed towards a dead end :-)

In order to do animation and other things that you would want to do in a game, you'll need to use the sierra's interpreter's kernel functions.  And to use those you'll need to have a class system that has the same properties and methods as that in the template game.

That is to say, there is some amount of functionality that is built in to the SCI interpreter.  The objects passed to the kernel functions have to have specific properties and methods.

You could definitely design variation's of the class system in the template game, but much of it would have to be the same.

fyi, you could probably go look at the freesci  source code and determine what exactly is required of which object.

[Nostalgic]

In order to do animation and other things that you would want to do in a game, you'll need to use the sierra's interpreter's kernel functions.  And to use those you'll need to have a class system that has the same properties and methods as that in the template game.

Yup - that makes sense. What I want to be able to do though is "prove" that without having to just take it on faith. More to the point, I want to show how the template system, at least in its most basic form, came to be.

I agree, though, that once you have that idea down, trying to recreate the entire template system is largely a waste of time. (It would be like recreating MFC just to create MFC apps.) That said, I like to make sure that I investigate these things with any system I work with because I want to be able to prove statements about the system (to myself, even if no one else).

Beyond that, even if the template game did have to exist largely as it is now to work with an SCI Interpreter (which I agree with) it should be possible to abstract out the creation of that template without using SCI Studio. What prompted this on my part was that I was finding that SCI Studio didn't make this very easy. The fact that it offered the ability to create a blank game made me think that either (1) the author of SCI Studio was just being thorough or (2) I'm missing something completely obvious about how to start working with a blank game.

When all is said and done, I agree with you: this isn't a path I'm going to take too far.

[OmerMor]

Nostalgic,
I find what you're trying to do very interesting. As a C# programmer with a fond to old adventure games, this has high appeal to me.
Please continue to share your insights on the subject.
Thanks,
  Omer.

[Nostalgic]

I find what you're trying to do very interesting. As a C# programmer with a fond to old adventure games, this has high appeal to me.
Please continue to share your insights on the subject.
Thanks,
  Omer.

Will do. I'm actually writing a more involved tutorial for SCI. I'm combining elements from everything I can find, including information I'm gleaning from the FreeSCI specification.

In going through the current tutorial for SCI Studio, I find it a bit lacking and the same applies to the SCI Studio help. I'm not saying this in a combative sense. Clearly the author of the tool had his hands full just making the tool itself. It's hard to do that and write good documentation.

But I'd like to see more documentation for SCI as a whole and then show how that can be translated to tools like SCI Studio and SCI Companion.

Once I have enough of this new tutorial posted, I'll definitely provide a link. I would be appreciative of any commentary on it, even if it's just to tell me that what I've done isn't so great.