Some anti-aliasing would go a long way.

Nice work.

Please remember that I am only releasing the source at your request, and that putting it on GitHub consumes even more time that could be spent on Guardian II.

I will at some point, but not until I see value in doing so.

I don't like major refactoring.

I carry on my project until I either finish of give up from lack of organisation*.

Any improvements to organisation of code are done in whatever next project I do.


*Most of the time I try to fix it up and get it good enough to be called a game.

Not yet, at least. Maybe in the future, when it's finished.

However, Guardian is now open-source. (should probably post this in the right thread)


It took me ~an hour to get Guardian open-source, with getting the license in all the files, creating the readme, deciding whether or not to put the art in the package etc.
It will be a while before I distribute a working project for people to tinker with.

Okay, campaign levels are delayed a little bit due to reformatting the levels.

Before everything was done by images with team data etc. in text files.

Now specific nodes can be placed in the level through the text files to help the flow, define gameplay etc.

So far there are only spawn and reinforcement nodes. Not much, but at least now you spawn and get reinforcements in different places if neccessary.

Maybe it'll even be possible to change where reinforcements arrive etc.

I'll do a bit more work on different nodes (for recon levels, capturing castles etc.) and then get to work on the first part of the campaign.

Also, recently: Removed invisible walls and instead made the camera stop at the level boundaries.

It's 2D voxels, not 3D pixels.

3D pixels are 2D objects rendered in 3D.

2D voxels are 3D objects rendered in 2D.

10% interpolation is a good idea.

Can we see the new code?

The code I provided takes the gap between the centres and checks if it's less than the sum of the bounds.

Your code is a bit weird. You're adding the width to the pos with one, and doing the same to the other and halving it. I think that explains the problem well enough.

Just make the player a 4D object, then add controls for movement in the fourth dimension.

Collision can be done with just a 4D AABB

1. Sorry. That was left in while 'experimenting'. That was there before I tried making it a file, and I accidentally left it in.

2. I need the path as a string to get files from the directory (and subdirectories)

The whole point was that it took too long.

No, it wasn't Github that drove me away from open-source.

I gave up on open-sourcing my code because it was a waste of my time that could be spent coding.

Github drove me away from Github. It was a waste of my time that could be spent coding.

I keep my source more or less undisclosed, but I'm happy to give away parts of it, and am working on some 'pure LWJGL' tutorials for beginners in my spare, spare, spare time.

There's no point in a novice sharing their code with the world anyway. It would be a waste of my/other peoples' time that could be spent coding.

I installed the Github application a while ago, and it stopped working and I ended up using the shell anyway.

Then I got sick of it and decided not to be open-source.

I have the time (& some/most/all of the skills) required to create such a site. I just don't have the time/motivation to manage and update it.

If you want me to get something running, I can.

Test Build 009

Fixed "user.dir" bug.

Now it should run fine on Linux, and if you for some reason want to launch the jar from the wrong directory.

Testers will have to redownload the launcher. I recommend waiting until version 010 is done, as there is nothing new that the player will notice.

Okay, I had a bit of trouble fixing Guardian II using the code mentioned above, and decided to post the fix I found in case someone has the same problem.



It returns the directory containing the jar.

Example level chain.


Squares are levels.
Blue represents the level to go when won.
Red represents the level to go to when lost.
A circle represents the target level.
Yellow means if you lose you go to a new chain.
Green means if you win you go to a new chain.
A dot with no line means if you win/lose then you have to redo that level.

Remove mainMob from the method parameters as it is useless, change to return mainMob, and when calling the function, do this:



Also, this code has too much of a brute-force approach just for simple animations. I recommend redesigning for future projects.

Okay, got back to work on this.

I'll be brief, as I'm working quite hard.

Paladin:


More info on what's going on.

In basic terms:

You have a lot of objects stored in RAM.
There are 'pointers' which store the position of the object in RAM.

When you call a function, you pass a copy of the pointer to the function, which then finds the object which it points at.
Any changes to the object affect the object, which will change what's being referenced by all pointers, but:
If you reassign the variable, it changes the local value of the pointer, not the value of the object, and not the value of the pointer in other locations.

Here is a workaround to help explain my point:


(Generics aren't neccessary, you can change it to just a certain class if you want)

Then you can change the value and it will change in all locations of the class.

DISCLAIMER: I do not recommend using that workaround. It was just to help explain the problem. Please just redesign your code.

Well, that depends on which provider you use.

Tumblr: Posts need to be approved by the admin before being published.
Wordpress: Has enough functionality that you can protect against spam while not punishing good people.
Blogspot: Don't think blogspot support multi-authoring. Can someone confirm?

Using Github could get quite messy.

Why not create a multi-author blog?

-Easy posting of code
-Feedback by comments
-Screenshots in posts
-Explanation of code in posts
-Sort snippets by tags
-New posts get more attention for a while.
-Admin can manually create pages for anything extra

It seems to be the optimal free solution.

Then when/if you want to get advertising involved, buy a proper domain and start hacking around in the html/php/whatever etc.

Actually, the best way to start an MMORPG is to jump in and get a server/client working.

If you spend forever talking about how something would work, and then when you get to actually making it, it all falls over, you waste everyone's time.

Ignore the ads part, ignore monetizing it, ignore all the stats, if neccessary ignore having different users and just let people post code.

JUST GET SOMETHING WORKING SO YOU DON'T WASTE EVERYONE'S TIME!

Stop dicussing this and get working.

Get a basic site where you can post code snippets first.

Then worry about the details afterwards.

Talking about how something will work when it will never be made is stupidity.

Usefulness is more important than complexity.

And usefulness can imply complexity in some cases.

What about most useful?

@theagentd: I would have appreciated that if this weren't an article.

Have an imaginary virtual medal!