Welcome to Gaia! ::

The problem isn't really the 2D nature, although the perspective will seem a little odd (Gaia avatars look pretty normal either from the perspective we currently have, or a straight out front view perspective), but the nature of the map. As it is now, each "room"/"square"/"sector" of the map ends when the avatar moves to the end of it. That is, you visually reach the end of the map sector when you actually reach its boundary. A front view mode would set that boundary really far back, or even almost infinitely far back. When you were near the "back" of a sector, it would be really hard to see the avatar, and moving "forward" toward the player would just make you larger, blocking up the screen. Other players or animated could intentionally or unintentionally block the view of stuff in the background just by standing over it closer up.

One alternative would be to have the map sectors as they are now, even with the visuals different. However, this would mean that the avatar would disappear at seemingly arbitrary points while walking forward or backwards, and the camera would shift into another zone that's just closer up to them (rather than being completely unseen before). That would probably be disorienting.

Another alternative is to use it with the camera following the avatar, panning through scenery items forwards or backwards. This would probably be graphically inferior to the current design, due to having to consider all the flat renders as images to flip through according to how far through the avatar was, and would require a lot of work.

The last alternative would be to only use it in limited cases. Of course, it would still take some work (the scaling of walking would be different, for example - how fast you move in a direction when you walk one way, then also the avatar size changes). The places where it could be used would be hallways and such without a lot of depth or the ability to approach too close to the "camera" (a narrow cliff path, where the camera view is perpendicular to the path is one example that comes to mind).

Maybe in some cases that would be useful for a specific reason, but given the difficulty compared to usage, for most instances I'd say it wouldn't be necessary.

*cuts 3D rant*
*cuts snarky comments*


Actually, there's no reason why this couldn't work! gonk It'd take some creativity to make it happen however.

Taking the corridor example, I seem to find it very possible to add "depth" to the game. I've seen it on other games before! But it'd look quite awkward if the avatars moved at constant speed around the screen.
This game footage (check until second 40) is entirely built with 2D environments. Everything you see is 2D, and the invisible "pathing" the 3D model of the MC is walking on is also 2D. Notice how as the main character moves away, his model reduces in size, and moves around slower on both axis. It's a trick to give you a feeling of depth, that the model would reduce in size when it reaches the top most part of the pathing.

I believe that the developers are VERY capable of doing this. Not just does zOMG!'s camera work exactly like this. They already did a trick like it in Gold Beach! They mapped areas with water depth, and a visual effect would happen depending on how deep the water was.

If we make a place like a winding stair way going up, and we map "Distance to the camera" instead of water depth, we could emulate the effect of being far away. The further up you go, the slower the movement in the Up/Down axis is, and the smaller your avatar becomes. The left/right axis movement is also slowed down, but only by a fraction of what the Up/down is.

But Quint! From a sinking avatar, to an avatar that shrinks and moves slower goes a long way of coding.
Actually, not that much, really.
As a starter, distance is already programmed in the game. Both the size and movement effects at that! Go the general's House. Next, go to EB's lair. You'll notice that your avatar is HUGE in general's house, and can move around the screen very quickly. However, in EB's lair, you take a awful lot of time to get anywhere, and the hair clips in your hair start becoming a indiscriminate mass of colorful pixels that could be of any shape when zoomed in.

It'd take an effort to figure out a way to make a distance transition, but other than that, the effect is confirmed to work perfectly well already.

Eh, I'm not so sure on that. There's the basic tech, as you state, but it depends on how it is implemented. If the quality "Scale" is just a set factor on maps, set for each section (or each overall map/instance), then they'd have to add in a new mechanic which would allow for equation based scaling rather than a constant just set in the info or whatever for the map or section.
Or scale might not exist, with map squares just being different sizes (so instead of saying "this map is scaled at 1, so the avatar is going to be 4 arbitrarylengthmeasurementunits (almu) tall, and this map is scaled at twice as large, so the avatar is now going to be only 2 almu tall", it might just say "this map is 100 almu tall, and this map is 200 almu tall, and the avatar is always 4 almu tall").

That's the main technical hurdle. If there is no location equation dependent scale feature, and there's so far been no reason for one to exist, it would have to be made specifically for this purpose. Isometric to Perspective switches aren't precisely easy... [will edit in a bit more later, must go now]

That crossed my mind. Yet, without any technical knowledge, I only thought "What if movement and avatar size are default values loaded along every map?"

It came up because I remembered how in SS, to see the Sea Lab in the distance, the developers had to add a different screen, that just momentarily puts you outside the ledge proper. Probably to both make it easier to not have to draw screens in between, while still allowing the view to fit the whole screen.

Regardless, what I meant by "distance" already being there stands. They must have a scale value, so they'd need to code a way to allow this to change. It could even be a secondary value that overrides the regular one under circumstances. Or doesn't even override anything! They could just make it so any flash object entering a region gets automatically scaled to a % of its size, the same way any object entering water will sink and be "hidden" below a certain height value contained in itself.

Still, considering how the perspective is done in this game, which is using actual 3D models and rendering pictures with the model viewer, then drawing over the render - just so that it is always as correct as possible, such work seems needless, no? - it's likely that this effect would never happen, as that would mean going out of their modus operandi for one screen only. There would have to be some really good reason to go ahead with such an exception.

It's not by any means a difficult change. It could potentially be a complicated one, though. You make a scale map or scale equation that calculates the proper scale at different points. You place constant values of this for all areas as they are now, to match their former properties. You remove the former property, and all references to it in all the rooms or maps. You test to make sure it didn't break something, and that everything seems normal and works out realistically without adding extra pressure on the computations. Or, you make a scale map or scale equation, then you put another function which either follows room/map properties or scale maps/equations where applicable. You remove references to how things decided which scale to do before (just by checking map properties). You test to make sure the function switches at all the right times, that nothing is broken, and it doesn't make things run slow in those specific spots.

That's the technical side. Then for the artistic design side, you take shots from a different camera angle (not too hard, usually), and arrange things carefully - making sure things don't obscure vision, that the horizon looks natural (not too much hard to see details in the distance, etc., but not too little detail either), etc. Since it would still be placed in an isometric map, you'd have to calculate out where to place things to make stuff look like they're in the right place in perspective (tricky for larger maps - instead of just modeling a map and then being able to divide it into rooms, they'd have to compress each part into perspective drawings from different points). Test some more, make sure the animations fit close enough to the perspective feel and the angle you chose (since animations are made for the angle of vision currently in use), use other effects as needed.

It's not impossible, but it's a nonnegligible greater amount of work than making areas with the normal point of view. With the limitations of it (can't be too expansive or we'd just lose sight of avatars and such, stuff blocking the view, etc. - the need for the angle of vision being somewhat close to the one we already have [so the animations and the avatars don't look too odd], etc.) it would need some impressive and/or important usages to make it actually worth it.