next »

[Gregg Williams]

Alright won't be much activity here for the next week or so. Need to wrap up the last bits of Gravitum and get it out of the door.

[DustyDrake]

Any suggestions for the editor's name?

I was amusingly considering M.I.N.T.E.R

SPEAR sounds like a good name if you can get a backronym outta it.

[Impmaster]

How about C.H.O.C.O.L.A.T.E?

Creative
Helping
Organization
Code
On
Lightweight
Abstract
Tiling
Engine

[Belimoth]

I dunno about editors but you plan an expansion pack called "Condition".
...
...
...
...
...

M.I.N.T. Condition

[Gregg Williams]

I dunno about editors but you plan an expansion pack called "Condition".
...

M.I.N.T. Condition

On another note, Kawe has started up some mockups of user interfaces, below is an early concept for a unit display.



I'm still tied up wrapping Gravitum, but hope to return to M.I.N.T as soon as possible, and with far less projects competing for attention.

[t-recx]

On another note, Kawe has started up some mockups of user interfaces, below is an early concept for a unit display.

<snip>

Looking good! What are those bottom icons?

[Gregg Williams]

On another note, Kawe has started up some mockups of user interfaces, below is an early concept for a unit display.

<snip>

Looking good! What are those bottom icons?

So the basic lineup goes like this it seems.

First: Action Points
Second: Health
Third: Extra actions that need to be recharged, or are one time use, etc.

[Gregg Williams]

Sorry for the lack of updates y'all. Gravitum is taking longer than planned, thanks to also needing to do crunching on some contract work. Thankfully crunch is done, and I hope to have Gravitum wrapped up within the next two weeks for the most part.

Then it will be back to M.I.N.T development without any other Retromite projects stealing time and glory.

[Gregg Williams]

So what I had hoped would take 2 weeks or so tops, has ended up taking nearly two months thanks to various other issues in life, but Gravitum development is 99% complete. and I expect to return to working on M.I.N.T this week.

[Gregg Williams]

Alright, all other projects are now wrapped up, at least from a production standpoint. After some design discussion, it looks like we're changing how map making will work.

Instead of going with a more free form painting and placement style system, we're looking at instead simply making various template style pieces that snap together. Similar to say the spacehulk board game, yet being city based. This should allow for super easy layout of maps, while also allowing us to enforce certain aspects to make sure the game feel and any technical limits are always addressed regardless of map design.

Also this likely allows us play with some other interesting aspects like the pieces being very much just layout based, and the game procedurally placing graphics as needed upon map load. In this fashion, every time you went to play a given map, it would have the same layout, but the roads, building types, and other bits of decor would be randomly chosen. This means we could also patch in new graphics for additional variety and maps would automatically be able to take advantage of them.

The possibility of an actual full fledged random map generator opens up as well with this change.

At any rate this change probably also means less time spent on map editor development and more game development happening. Still its early days for the project, and nothing is set in stone yet. (Well other than awesome mechs, multiplayer, action points and some other stuff.)

[Udderdude]

How would you compare the gameplay in this to Front Mission or Advance Wars?

[Gregg Williams]

How would you compare the gameplay in this to Front Mission or Advance Wars?

Quite a lot deeper than advanced wars, and individual unit based. It's been awhile since I've touched a front mission I think 3 was the last one I played, but from what I recall deeper in the tactical aspects. We're planning action points, line of sight, fire arcs, over watch and other elements. Front Mission might win out on customization, well see how that goes.

A huge distinction though I think in general is that M.I.N.T isn't being designed as a tactical RPG, but rather as a war game, and with a large emphasis on multiplayer. So gameplay is much more around selecting and customizing the units you want to field when you only have X unit slots and Y points to use in a given battle, and then it's all about tactics and strategy, there is no leveling up and so forth. (SP might deviate some of course)

My own hopes for it come closer to a mix of table top warhammer, battletech, and space hulk.

[Gregg Williams]

So progress has hit a bit of a halt, as we've ran into a serious issue with SFML which we use in our framework for window and input handling. On Mac (My primary OS) it seems to not recognize multiple mice being connected.

Previously I always did my development straight on the Macbook Pro, but after various RSI issues for the last few months, I've moved back to a desk setup and an external trackball. Which currently means our existing framework instantly crashes upon bootup due to problems in SFML. (The latest release of SFML doesn't seem to crash, just doesn't work.)

I've been attempting to debug the SFML code, but so far I've not been able to identify why the IOH API is failing to detect a mouse button being down.

[Gregg Williams]

Well after a long 10 hour day SFML is out, and GFLW is in, along with irrKlang.

Editor progress is also being halted for the time being, given our need to validate more design concepts. As such we're leaving tooling aside for now and focusing back on a playable early prototype to nail down more of the design factors.

We may also be switching from a staggered to diamond isometric map layout, as the new level design plan tends to make use of a lot more walls or buildings acting as walls, and the diamond layout tends to show walls and map edges really nicely.

[Gregg Williams]

So completed switch from staggered isometric layout to diamond based, and wrote up initial map loading support using the tiled map editor temporarily.
Next up is expanding to the rest of the map layers, getting buildings to be entity/tile hybrids, and unit depth sorting working.

Also started using irrXml which I really am liking compared to TinyXML2.

[Gregg Williams]

Units are now properly depth sorting behind buildings, though a few bugs still to squash here.

[Gregg Williams]

Squashed the depth sorting bugs. (I'm so glad I made sensible decisions to make this process much less painful, like splitting artwork for anything that is bigger than one tile, and ideally restricting units to a single tile width.)

Started to figure out how to utilize tiled for our purposes, the most annoying bit being its GID scheme, and that while there are object layers, you cant seem to actually attach an object to a given tile space in the map. Seems like the best way to accomplish that is to actually do it on a tile layer, with a real tile that just has a property set. Not exactly what I was hoping for though.

Also started to rework some of our tile strips, everything is currently in a single strip, but in reality it needs broken down into a strip per map layer, which looks to be 4 layers at the moment.

[Gregg Williams]

Well a long day of tile work. Needed to re-slice building graphics and make sure that nothing overlaps any pixel wise. This way when we turn the buildings transparent and dark, the alpha blending is consistent.

Still need to split out tiles into a few more strips, but should be moving forward more this weekend. Next major goals are really a full set of map layers, buildings as entity/tile hybrids, building transparency when needed,  unit selection, movement grids, and path finding.

[Gregg Williams]

Tile strips are made.
Map layers are correctly loading from a tmx file now.
Tile info database seems to be working correctly across the various strips, at least for height information at this point in time. Eventually this tile database will grow to cover a likely great many amount of properties.

The next task which is a potentially tricky one is actual building support. Basically buildings are composed of 3 layers of tiles (foundation, shadow, and structure) and are also an entity, as they can potentially be entered, destroyed, and so forth. The plan is essentially to have the building hold its own multi-dimensional array of tile data, while marking the main map's tiles as appropriate that a building entity exists at this location.

Then when the renderer steps onto a tile in a given layer that has a building entity, it can ask the building for an appropriate tile ID and eventually transparency state or not. The building just needs to be able to translate the global map coordinates, into its local coordinate space, and return the correct tile id from its array. This also allows the building entity to return different values for damage states and so forth. 

Once that is working, the next step will be to handle the various transparency aspects, so that if you select a unit that is hidden by the building, the building will do its transparency magic as needed.

[happymonster]

Ohh.. that sounds complicated! Good luck.