Coding Issue: Java 1.1 vs. 1.2

[JimC]

re differences between Applet,Frame

I know what the differences are - I've outlined how to get around them in one of the above messages.

I'll give a brief example - say a class wants to load a file. It should call, say container.loadfile(FILENAME) where container is an interface which will be implemented differently for the Frame and the Applet versions.

I've used it before, it works.... and like I said, it'll save redesign time if we screw up.

Jim

[F_Smith]

Jim:

Agreed, we can do some things in duplicate on a class-by-class basis. But I'm referring to the architecture.

This is a big project. We have got to decide on an architecture before we get started. We can *not* make major architecture changes in a project this big later, without seriously impacting the success of the project. That's one big problem with 'scope creep'.

We really must decide now what our vision is for the functionality of the game. That will dictate whether we go with applet or app. I've tried to point out some of the functions I believe clearly point us in the direction of applets. Do you agree with those points?

To reiterate:

* build-in robust multiplayer support

* easy one-stop upgrades

I believe that you can not achieve either one of these with application code, therefore there is no reason to continue to spend time on application functionality. Do you agree?

P.S. -- of course we will be using standard 'object oriented design' methods of black-box methods.

[JimC]

Well, I have a demo Mark's path-intensity ideas up and running now....

Problem is, it stubbornly refuses to run in a browser-based environment, despite working beautifully using Sun's Appletviewer.

It also works fine as an application.

Luckily, I'm sticking to my principles of making the design as flexible as possible.

If anyone wants to attempt to run it from the web, it can be found here.

Stubborn Applet

If it works on anybody's system, or if anybody has any ideas as to why it isn't working, please let me know.

I'll distribute the class files for the application version to the coding team ASAP.

Jim

[Blade Runner]

F_Smith,

If I understand well the Java system, when we use an applet, it can't read or write something on the drives of the slave machine. How can somebody use a scenario or modify the rules of the game for him/herself? If we can write the best AI in the game programming history, the real selling point of this kind of games are the scenarios and the add-ons. I'm affraid we need to stick to a traditional application version. I know the applet version has a lot of advantage, but the modification possibilities are very important, practically this can be (must be) the strongest part of the game.

I read an interesting thread in the CIV III forum. Somebody propose a client/server method to produce a game program. Can we develop the Clash program with client / server architecture? This way we have one advantage: easy to implement the multiplayer version.
The AI(s) and the user interface(s) (clients) use the same interface of the server.

Blade Runner

[Blade Runner]

The thread in the CIV III section. The title is AI (surprise!!!) and Rong sended it.

Blade Runner

[Blade Runner]

JimC,

It is work for me.
AMD K6-II 300
Win95 OSR2
Netscape 4.51

Blade Runner

[JimC]

REALLY????

Well, it loads for me, but when I click on 'start search' it manages about 1 path every 5-10 seconds.... (this is using IE4)

In Sun's Appletviewer or in application mode it manages about 10-30 paths every second.

What kind of speed can you get?

Hmmmmmm - maybe it's Microsoft's fault, since it runs fine using Netscape and Sun's Appletviewer and Java VM.

Anyway, if it works for anyone else, please let me know what kind of speed you get when you click 'start search'.

Temperamental Applet

If things seem OK, I'll have the demo fully functional within the week.

Jim

[Blade Runner]

JimC,

When I try to check the speed, the applet sended an error message and wouldn't like to work. Anyway the applet was not so slow, but I really can not remeber the speed.

Blade Runner

[Mark_Everson]

Jim:

Tempremental Applet worked great over the web with IE5.0. It took about 10s to transfer and 30s to run on default settings.

WoooHooo I can Run something!

-Mark

[Mark_Everson]

Jim:

Tempremental Applet worked great over the web with IE5.0. It took about 10s to transfer and 30s to run on default settings.

WoooHooo I can Run something!

Is there an echo in here?

-Mark

[This message has been edited by Mark_Everson (edited May 26, 1999).]

[F Smith]

Worked fine for me using both IE4 and Netscape. Took a while, but not too bad.

RE: the saving of files. (Loading local files is no problems -- getDocumentBase() does the trick).

There are several options.

1) Best: a client/server architecture which uses a browser's 'save' function and cookies. By downloading the applet to a local drive, you can read from the local machine without a problem. This is how Yahoo's "Games" works.

2) Second best: we can sign the applet, and gain full file i/o capababilities. This is how Microsoft's "Zone" works. It isn't as difficult as some people make out.

3) Third best: we can use copy / paste for saving files (since they'll all be in text). I've seen this used smoothly in several amateur designers. Look in the 'Java Games' thread for examples.

P.S. -- work is getting pretty heavy. Might not be able to respond much until next week.

[Mark_Everson]

F_Smith:

Sorry to hear real life has swallowed you up. We'll take you when we can get you!

-Mark

[Blade Runner]

F_Smith,

Thankx! I think in this case the best to use APPLET format.

Blade Runner