[Mattias Gustavsson]

Work with entities in a three dimensional world-space. Translate to 2D through a simple projection. Job done.
_________________
www.mattiasgustavsson.com - My blog
www.rivtind.com - My Fantasy world and isometric RPG engine
www.pixieuniversity.com - Software 2D Game Engine

[tcaudilllg]

Yeah but it doesn't look as good. Plus it's one thing to write an engine for 3D processing and a completely different thing to make such an engine easy to make maps for. Requires an editor redesign I'm still working on.

[Mattias Gustavsson]

How to expose it to the user is a separate issue. But picking a good internal representation is key to making the 2D/3D problem manageable. If it feels like "an immense intellectual challenge", that is a clue that you have chosen your abstractions poorly. It should not be a hard problem.
_________________
www.mattiasgustavsson.com - My blog
www.rivtind.com - My Fantasy world and isometric RPG engine
www.pixieuniversity.com - Software 2D Game Engine

[tcaudilllg]

My abstraction isn't the problem. My problem is the approach itself. I've thought about it and concluded that the stairways should simply make use of the actor physics system, rather than the tile system. When the actor moves over a "stair actor", it gets the tile's slope and elevates itself according to the slope. I'd thought to conceal all that from the user, but it seems like I'd gone about it the wrong way.

[tcaudilllg]

Yeah I'm probably not gonna release ramps in this next version. Might release something simpler.

I've got the talk system finished, I think. Also mapped the gamepad to the keyboard. (via Gamepad API).

I think I have the design for the ramps finished... it turned out to be far more complicated than I anticipated, requiring separate routines for each direction in each case. There was simply no way to simplify the algorithm, although I wanted to.

Ultimately I needed to figure a number of elements of this system out regardless by the time I had moved on to the 3D system. For example, I had to calculate the point on the ramp whereat the actor would land given their X and Y coordinates. This required specific formulae.

I hate math.

[tcaudilllg]

Today's the day. Gonna release .995 in just a bit.

[tcaudilllg]

OK I realized that I could have handled the fonts by doing a drawImage crop, meaning all that pain was for nothing.

That smarts.

I realized my program design for the sprite editor bogged down when dealing with high zoom rates. Did some optimizations and localized the root of the problem to the mouse event trigger rate, which is apparently more than a thousand times a second. Installed a drawing queue + setTimeout and although the latency isn't as good in the best case, it's a lot better than it was in the worst case.

[tcaudilllg]

Tiletext finally works. Which means I'm ready for release. However this will be a closed beta, because my aim this time is to create professional relationships.

If you want to beta tests Gamestar .995, PM me.

[tcaudilllg]

Added SNES-style graphics layers (two per Z layer). Which, I think, brings Gamestar 1.0 to the point of completion.

[tcaudilllg]

Gamestar's development is nearing completion.

I've added functions for bitmapped fonts and particles (integrated into one system). However getting from the sprite/tile editor to the map editor remains cumbersome. I'm thinking to rework the tile system. Currently tile set design is cumbersome and boring: you have to first draw the tiles, then export them individually, then import them into the map editor/environment one by one. Need to make it so that you can edit tile sets on the fly and get that "one click and you're done" experience.

Power... all about DA POWAH!!!!

Of course I've not worked out the details of the 3D system, but as far as that goes I don't think casual designers will ever be ready for anything beyond Minecraft-style map editing. Professional 3D design is for the architects. If I can get to Xenogears level, I'll be satisfied.

[tcaudilllg]

New release... .995 or somesuch? Trying to finish the sprite editor. [s]There seems to be some kind of bug in Firefox CANVAS element that prevents it from scaling except immediately after having its width and height declared. This slows down the drawing substantially. Will have to bring it to Mozilla's attention.[s] my mistake... assigned canvas dimensions the dimensions of the containing div (which are string vars e.g. "0px").

In other news, Firefox is dying. It's not a tremendous deal... I can always port it over to Adobe AIR. Still, would lament losing Firefox. 10 years.... There's a possibility of course that a weaker Mozilla will be a better Mozilla. Google probably thinks this and that's why they are giving them cash, to keep them strong and ambitious (and bug-ridden).

I believe the app does work on Android. Not that well though. I've tried the sprite program and was able to engineer some scroll bars for the mobile version (mobile browsers don't use content scroll bars).

Debugged the tile system... I think. Also made a tile cutter function so tiles from RPG Maker and wherever else should be usable.

[tcaudilllg]

Firefox's strict security regime has brought continued development to a halt. Hence I'm moving Gamestar to Adobe AIR (and probably quitting Firefox completely).

[tcaudilllg]

I spent three hours tonight trying to get a . Tired of this. Tired of all the work. So many damn APIs... it took me years to learn Javascript... it just seems like like every "API" is just some bighead trying to force his way on everybody. Was reading tonight on what Adobe AIR "does and does not" support in HTML. It's a fine cherry pick. I'm tired of the cherry picking. Tired of the work. I want to create. Gotta find something I can CREATE with.

I've got no more will to struggle with Firefox's API. The internet is so intimidating... useless to create a browser of my own. I hate C++ and it's difficult to understand besides. I just want to throw up my hands with it. I'm proud of what I've made but there is so much evil in the world... it's like the light can manage to exist in spite of it but only thus. The vast ocean of evil squelches the light.

Of course I know these things come in cycles. The pendulum is simply swinging into the darkness at this particular point of internet history. There will come a time when it swings back into the light, once the darkness has annihilated itself. I hate that the darkness seized control of my favorite damn product, sunk its teeth into it. It's really depressing to see Firefox die, but die it shall.

[Mattias Gustavsson]

I feel your pain. Totally agree, I hate working with limiting APIs. I stick with C++ (or rather, mostly a C subset) even though I dislike it, just because it makes it possible (though still hard work) to remove most limitations.

I often find myself wishing for the olden days of coding with *no* APIs, just writing directly to memory and registers to make the hardware sing and dance *sigh*.

Some days, I miss BASIC.
_________________
www.mattiasgustavsson.com - My blog
www.rivtind.com - My Fantasy world and isometric RPG engine
www.pixieuniversity.com - Software 2D Game Engine

[tcaudilllg]

I persevered somewhat and managed to complete the transition to loading from data URLs for all my graphics. Well, really I had to read the graphics from a file and convert it to a data URL. It was really complicated. :P

In the process I managed to remove a couple bugs that reduced performance. The new build is up, now with a map-to-PNG conversion function.

I don't like Adobe AIR's strictness. It seems to be influenced by Google and I kinda hate that.