I see this as a business opportunity for competitors of US companies
to host their servers outside the US juristical access.
And implementing a strict secure communication from client to server (https for starters)

At one time there will be economic pressure to reduce (or make people belive so) the
sniffing on their private communications.


----

The main effect of observing and recording peoples behavior lies less in aquiering intelligence data to find bad terrorists
(or drug trading, or tax fraud, or pirating software, or mobbing schoolmates, or throwing away beer cans on the highway, or critizising the current govenment)

But more into making people change their behavior in their daily lives. (a broad feeling of intimmidation)
Beeing more careful and communicating less critical.

sorry, did not see that it was !=1
forget that code sample

Ok, if you had a lot of slots, this could be more managable:



In the end, your current if statment is faster.
If that would matter in your context.

Versions are important if you have to handle .. well more than one version.

So its needed when: 
-you give your build to someone to test it.
-When you release some version (in a store / website for example)
-when you make a backup of an important milestone

If you just work yourself on your own codebase / repository, and the project is
in its firsts steps, you dont need a version yet.

More important than the format, is that you actually increment the version...

A lot of developers are too lazy to do that,
ending up with different builds with the same version...

Have to deal with that daily

Its always good to cut projects into smaller Tasks to reach.
Much more motivating.

Like writing a little game:

I doubt Minecraft was optimized for Linux in the first place.
So there might be some tweaks to optimize it.
Question is, can you do that with the available (e.g. -> decompiled) code.

This also works, can be run without a main method.



or even



Would be a good way to make your teacher mad in a programming class.

Compare The Oracle JVM and OpenJDK.
Oracle works in this case faster on my Linux.

The biggest impact has changing to a propriatary driver compared
to the open source driver.
But this has also its problems.

Gimp - since 2.8 - sucks for pixel editing.
They messed up the brushes, very cumbersome now
and saving does not allow to save to png directly without an annoying popup.

on the Java4k.com site you will find many project with sample code.

These projects (although harder to understand since its a size-matters competitions) should give you some good ideas about programming.

Nice voxel landscape.

good stuff

Hi, Thanks for the feedback.

There are several more missions

For the scolling: I had "nonstopping" scrolling before, but this also had problems with scolling
out of scene when you move the cursor out of the applet (checking Skype etc.).
So I stop the scrolling when the applet looses pointer-focus.
Actually the best alternative is to add keyboard scrolling.

You can use the "X" key to stop the action, and still issue commands (a feature done for the old cumbersome keypad phone input)

In the code is actually group selection, but its still mapped to Phone keys  

If i find the time and mude, I can add more features and maps.  

Tip: dont create complexity for the sake of it. (An unnecessarily complex Class-Hierarchy / including
huge Logging libraries if you just need a "Log" class that can print lines to a textfile)

Else you create lots of boilerplate code, and loos oversight of the important parts of the program.

Loops and Things:

https://www.youtube.com/watch?feature=player_embedded&v=v9funsyB-XM

And advantage of "things with a price" is, that they have
more "value" to someone, than something free.

I think selling a game , even when its a very low price, will make the
player come back to play it more often than a free game.

The approach to directly detect collisions and alter terrain on an (Image)array
was used in the original Lemmings on the Amiga.
Thus the method should not slow down a modern PC that much.

It depends more on the resolution of your terrain (how many pixels is 1 lemming?)
if you need an optimization like a quad-tree.

That can be counter productive.
This approach works well with a business apllication that has a well defined set of functionalities.

But a game is only as good as the final product, and can not be estimated on paper.
Things might have to be changed a lot during devlopment.
Defining too much beforehand can end in inflexibility and frustration.

Its better to make a general overview, and then a prototype of the gameplay / Mockscreens.
Only if this prototype is fun, add the specific list of contents / assets / levels.

-> It boils down to be able to make protoypes and/or good mockscreens (if you are an artist)

Exception to that: If the game is in fact a clone (gameplaywise) the prototype can be the other game
to be copied.

Well in real life shoveling a cubic meter of sand is quite a task.
Did that before, and left me with quite some blisters.

You dont really need a team for this kind of game. Try to work on that first.
If its good, its much easier to show it others when you ask them to build a team (the have less "slots" open)

I have worked on several commercial games with companies, but it was a drag in the end (restricted licences, restricted creativity, restricted schedules)
And released a game on Android to.
Its not really paying off the effort economically.
So I rather just make games for fun and impressing myself with what can be done
Anyhow, If you want to get into game development
-either be ready to enter the "cold" business of professional production, where you are a small part of the bigger operation
-or just make something that makes yourself have fun - but also get a "normal" paying job.

You can select a unit/building, and then in the menu use "remove selected unit"

Unfortunately the highlight of building does not show up, but it selected when its name appears in the bottom.

----

Also patched that black cursor box. (did not show on Linux, so I noticed that quite late)

-Minimap can now jump camera by clicking on it
-Dialog will display full text by clicking it before exiting

Its more a sign of clarity.

When you read a code from another programmer you can quickly see which specific classes are referenced,
and if anything might be missing, old version etc.
So when I look at source-code at my company, I can quickly see what classes and packages are important to that class.

You basically know what is important and what not to that class.

Thanks for the feefback.

I actually use an "empty" custom cursor to make the normal cursor invisible.
In your case this must not have worked and used a filled (not a transparent) square instead.
What JVM and System do you have? (edit: ok, just see that this is in Windows for me too,)

Yes the ingame menu is not very optimal. It is this way because in the original version it was to be used
with the keypad / softkeys of a phone. Thats also why it pauses the game (much more hassle on a phone to make quick inputs)
I migh just make a qickexit when clicking pressing outside of it.

The Enemy-AI retreats when it is hurt too much.
I think its funny. Because normally Game-AIs keep on fighting until they are dead.
I will map the "S" key to a new state "keep position" or
just make the player-AI stop pursuing units when going to far from the last command-position.

Regards, Damocles

Alternative small-screen version online: http://rtsmaker.com
(you might have to refresh your browser)

Ok, thanks for the info, I might have overlooked checking for level boundaries there.  
(..and I`m a poor software tester, how come i didn't test that..)

cool that you made it to level 7  (not cheating I hope  )


some technical infos

Its actually not an isometric rendering, but based on 24x24 px square tiles. (that why manually creating the terrain is a pain)
Generally the same method used as in Starcraft (minus the nice editor)

Here the simple Mappy Tileeditor.
I made the level in 4 layers.
(terrain, units, pathfinding and triggers)

http://rtsmaker.com/MappyScreenshot2.jpg

   
old testrendering in WED:
http://rtsmaker.com/rendertest1.jpg
 

Made a bigger border (for less unforgiving scrolling) and use a custom mouse-pointer.
Map can also no longer scroll out of the screen.

Yes Starcraft might have been some inspiration 

I will work out how to make the borderscolling more convenient, (bigger border,
extra mouseborder outside of the rendering and additional cursorkey controls)

Did anyone pass "locked" missions yet? I think on some of the later missions
it can get quite rough. Its hard for me to see if it is out of balance then if I know the maps 

It worked quite well. The user had an accelerated cursor, that was operated with the keypad.
Thats also why there is a pause function (as using the keypad would be too slow to control many units in realtime)
There was also unit group selection, which is not reimplemented yet to PC "standards".





Thanks, Im not really graphics artist. The method I used for creating the 3D models
and rendering it creates this nice classic 2D-Iso look. The GUI elements are created in GIMP.
Ok, the Main menu is a placeholder and not very polished.

You will notice that many of the assets are color-indexed. This was done to preserve a small size in the
resulting jar (down to 350kb on the low-grade phone version)

Yes, the border is only 20 pixels.
I will see if a larger detection border is handling better.
I probably use an extra empty boundary around the drawing-area then.

Still dome work with the controls to be done.

I also want to implement unit-group assignment to number keys then.
(to keep the stupid repair trucks in the back) 

Tip:  use "s" to stop units from haggeling around in contested areas when they cant find a path to their desired position,
 so they go into attack/repair mode quickly.