NetGore

Wow, huge is right.

Is this in NetGore? If so, did I add a map snapshot feature or did you assemble that on your own?

I like how you can't remember if you made one or not

If I know skye right, he's making the map as one huge png file, then splitting that into probably 512x512 textures, and then making the map up out of those :3

I really hope not, since that is probably the worst way to design maps.

Well I cheated...

This was an area made up of 6 vbGORE maps. Which I had textures for and glued them together along with some adjustments.

The way I make maps is:

1. Select a theme & the most notable features of the area
2. Draw a quick doodle of a possible layout on paper
3. Draw a sketch of the definite layout on paper
4. Scan in the sketch
5. Overlay with Earthbound graphics
6. Cut the area into textures (not necessarily 512x512s but just big enough to hold e.g. cliffs, walls, ... with as few "empty" spaces)
7. Put the whole thing together in a mapeditor

For preexisting maps (e.g. Earthbound ones or ones we had in the old SVO and I'm too lazy to glue together the separate textures) I do it as follows:

1. Search for a sketch I drew 2-3 years ago
2. Check if I can find the old textures for it, if so, reproduce the map with the separate textures

As with this map, I'm planning to just make 512x512s because I really can't be bothered to cut up the map again and trying to fit the pieces together, it would take an immense amount of time where putting the 512s together is just a couple of minutes work with minimal effort in the mapeditor.

So, no, you didn't add a snapshot function, but it certainly would be handy

- a desert.
- a Dock.
- ruins.
- some random dirt on the floor (tloz
- imperfections in the walls
- maybe an isolated area that can only be reached using a magic portal or something similar.

1. Nah, it's not a desert area
2. Nope, the river doesn't go anywhere, there's however a hidden entrance to a secret base behind the waterfall
3. That's not a bad idea
4. As I said, grass, stones, trees etc will be added later on
5. Yeah I know there are weird bits and pieces in the walls and the water shade, no one playing will really notice though and if it's a big error just let me know so I can fix it
6. See the area on the right

Lol since the maps in my game are 1v1 and 3v3 arena's only thats how the level artist is doing levels. He draws them on a 24inchx24inch canvas, sometimes paints some stuff and then scans into photoshop. He has a program that can cut them into tiles.

Not sure I exactly understand what you are talking about here. Does NetGore need an image like that chopped up and fit back together inside the editor?

No, that is just the way he did it. I'd recommend you never actually build maps that way.

Are the rivers some sort of orange coloured water, or are they lava or something?

Just because lava doesn't do reflections/shadows.

The area is called Orange Falls, it's orange water ^^

So using an image the same size as what he had there would be fine as an imported map?

I think Syke chops something like that into 512x512 bitmaps (or pngs, or whatever format) and puts them down in the map editor. (that way he doesn't have to go andput every individual map tile down)

What's the advantage to that versus just importing the full image in itself?

Well, if you imported the whole thing, that would be one huge image. Imagine what that would do. If you cut it into 512x512, the whole map won't be in one image, so your texture memory doesn't die... (unless NetGore handles grhs differently than I think)

Edit: Actually, the image is only 750kb... The image is pretty big though, not sure if textures can be that big. (at least in XNA, not sure about SFML)

Its more of an issue of design and productivity than anything. But yes, the whole texture has to be loaded into memory, even if you were to only draw a tiny, tiny piece of it. You "could" automatically chop up the image at runtime, allowing you to use a single giant image as long as the image container supports, but that would be quite a bit of work to do something you shouldn't be doing in the first place.

width*height*bytesPerColor = 512*512*4 = 1 MB for 512x512 image.

I assume you did the calculation based off of 3 bytes per color. Though I am quite sure SFML loads as A8R8G8B8.

Definitely an acceptable size. Modern commercial games will be loading textures even up to 8192x8192 when they can for all the high-resolution art and texture atlasing (for batching). That is pretty much the cap for modern graphic cards I believe (been a while since I've checked, so even higher resolutions may be supported today).

But it is a huge waste. And again, the biggest concern is productivity from a development standpoint. If you "really" want to go that way, then you can, but I wouldn't ever recommend it.

I should've remembered that. The width is like 16000 though, so that would probably be on the upper limits. (could cut it in half, or just in 4 pieces, making it 4000*3000) But yeah, it seems like a bad idea to me in the end, even if I can't come up with a concrete reason against it.

Edit: I don't no why my browser keeps double posting...

So is he dividing it into separate maps with loading zones between?

Going to try having it as one big area and see how bad the latency is.

No... (at least I don't think so)

It is one map : he is just splitting that image into pieces and puttin'em down in the map editor...

Oh, so that you only have to load the 512x512 pictures that you're walking on instead of the whole map... I would like the see the difference in lag between those honestly, just out of curiousity.

The textures aren't going to influence the latency much, I think you've got the stuff confused.

1 map with 1 texture = uses a lot of your CPU/RAM
1 map with several 512s = uses less ram, about the same latency as above
4 maps with several 512s = least lag

That's exactly what I meant. I probably kept throwing lag around when I shouldn't have been.

humm that's huge lol