[tcaudilllg]

OK I've got something to show now. I'll PM a link to the zip to whoever is interested.

I'm not gonna just post a beta for everyone to see. No, some people have been a little too negative for that.

[tcaudilllg]

Added an RNG, for purposes of testing the actor cloning function.

[tcaudilllg]

So free motion time. Then, I think I'll be in position to begin work on that elevator action RPG my friend and I are trying to write. It'll be a good game once it's finished.

The free motion thing is gonna require a different design scheme than the tile based system. However I'm implementing the two alongside each other: some characters will be free movers, and others will be tile based movers at the designer's discretion. I don't think it'll be much trouble to have the two systems running in parallel. On the other hand, one does not simply change a tile based character into a free moving character, as the graphical requirements, in particular, are different. And, there is also that issue of multi-tile characters. I'm thinking I may have to put a table lattice over the character sprite graphic preview, so as to mark certain regions as physical and not.... Oh this brings all sorts of added complexity.... For example: how to tag a hit to the head vs a hit to the gut or a side sweep? Looks like I'd need a tag system just for this.

There must be a simpler way... perhaps shared variables between several actors?

[tcaudilllg]

Added support for multi-cell actors.

Which just leaves the free movement, elevation, jumping, and another round of debugging. And I guess I should start on the scrolling issues, too. But, that'll be in a different release.

[Verious]

I would like to see the progress.

[tcaudilllg]

Free movement is done.

Somewhat undecided about what to do about tile tags. I could make them activate on "cross over" events, or I could make them contact event triggers. Cross over events aren't actually implemented, so might just stick with contact events and see how that works.

Verious I'll send you a pm as soon as I've verified the free movement engine.

[tcaudilllg]

After two straight days of coding, the free movement system is now debugged. Was also able to debug the variable assignment system. Performance is good... however, I'm going to have to rework the input engine because the keyboard doesn't read presses fast enough to move the character freely at a reasonable rate. Also, as predicted the tag system needs work... again. I think instead of reading key presses, I'll read key ups and downs.

Edit: it worked. No more performance issues.

I'll see if I can get the tags working again, along with some more board testing.

[tcaudilllg]

Tags work again.

[tcaudilllg]

I realized that it was expedient to divide collision responses into actions and reactions. The actor which makes the contact has the chance to run a script on the next cycle. The actor who was collided into has a chance to run a reaction script pursuant to the tags of the collider. If both are moving at the time of the collision, both actors have the chance to react.

I'm going to offer two modes of event testing, "shallow" and "deep" Deep mode processes every event in turn, with no skipping. Shallow mode processes one event per actor per cycle. All others are discarded. I realize that it might be worthwhile to implement a "priority" variable to give certain tags event priority over others. A tag deactivation function would also be useful.

edit: Variables scoped. Tying up loose ends.

I'm optimistic about the scrolling version after seeing this:
http://ernestdelgado.com/public-tests/gifoncanvas/

[tcaudilllg]

OK I'm ready for a release now. I think the multi-board system is working and all, still, although I've not tested multi-board play in the last two weeks...

The game system appears to be at a state of evolution enough that it can reproduce The Legend of Zelda (using the board system to alternate between displays). The old-style arcade games it does with ease.

My focus now is on elevation, functional stairs, and eventually diagonal movement (not yet implemented).

Rather than use "deep" vs "shallow" event testing, I'm going to offer the capacity to test events at varying levels of depth on a per actor basis.

[tcaudilllg]

Added a whole slew of new input functionality.

[tcaudilllg]

Music and sound now work. However, only within the limits of HTML5...

Working on the projectile system.

[tcaudilllg]

Projectiles now work.

There appear to be some glitches in the actor registration system. Will have to fix those before release. Also for some reason the board switch isn't fully working.

[tcaudilllg]

Sporadic bugs, particularly hangs. They are buried really deep down though, and it's difficult to find out what's wrong.

[tcaudilllg]

OK, things seem stable now. Reduced the complexity of the event system... an alternative solution isn't obvious at this point.