It's a cool idea - I like it. Next I'd just need more time to work on mapmaking :D

Cool. Sounds good.

Working on this right about now.(probably won't be finished soon)

Yay! It works!

Currently, its a fairly involved process, as I don't entirely feel like doing the rest...
also, it only works with objects for now.

To change a tileset, you open the map editor, select the layer that you want to switch, then hit the new button. It will bring up a list of options for which tileset to switch to
("merlin obejects" is actually merlin4objects; the font that is for the menu items currently has no numbers)
Once it is done, you will find yourself with a strange looking map. Close the map editor, and open your map in notepad(or any other non-fancy writing tool). Change the map layer definition to reflect the changes. Reopen the map editor, you're done!

In the future, I'll try and get that all done in the editor itself...
Also, if anyone wants to write definitions for the other tilesets, I would be more than happy to have them done for me! (its a matter of giving each and every tile a name, and making a list of where in each tileset that tile appears,(by name) it's a lot of mind numbing work)

Ugh, I hate that kind of boring copypasta work... Well, who doesnt :/


Thanks for doing the base for it though..

Cool.

I wrote up one for merOpenActive before writing all the code(forgetting that I could test with the objects...)

It took at least three quarters of an hour.

Doing the rest of the code, will probably take longer than documenting the other tilesets as I will need to find out what the most important place to change what the current tileset used is (there are a lot of references to tilesets around in the map editor, and I have a feeling that I know which one is the important one, but am not sure where to find it, nor how I will get to it through code) finding out how to update the pictures shown probably won't be much work from then on.

Ultimately, we should probably aim to have all the tiles in a big list so that each one gets a unique id. Then we would need conversions between the tilesets and the big list. But once a tileset could be converted to the big list, it could be converted to any other tileset.

That sounds like a pretty functional plan for the future, but I think I'll compile said list last. It will consist of one extremely long gif or png, and have a correspondingly long key, to account for every possible tile in all the different keysets.

On this kinda note, looking at the map editor, i noticed a reference to a foreground passive tileset. Does this actually exist? If so, pending that its not really slow, it might be a nice sign repository.(instead of having them in background active)

How's this: the excess tiles from the grand list is added to the end of each tileset(that apply), so you could change the tilesets without having to worry about losing data when switching. Furthermore, there could be targeted tilesets, which would only have a couple parts active(such as a tileset of just signs) and put them down, and then switch back to continue working on the rest of the map.(the things not in the current tile set would probably disappear though, causing some challenges)

That sounds like it could work.

Let me just get off my chest exactly what I was imagining so you can see where I was coming from.

Ok, so basically we replace all the individual tilesets with a list of tiles stored in the database on this site.

TMB members can then submit tiles which are then stored in the db.

The online version of the Map Editor shows a scrolling list of all the tiles submitted by TMB members, and maps are made up from this list.

In order to make things more navigable, the tiles also have category information with them to make things easier to find.

Because they are in the db, you can sort them by newness, rating, number of times used, search for a key word etc.

To publish an offline version of a given map, a script would save only the tiles used into a tiles folder. The site would deliver an .exe along with all the other assets (scripts, gfx, music) required to make this level run.

So that kind of the ultimate idea of where I'm heading but obviously you need something that's going to work now.

Actually, I've been talking here about individual tiles - but maybe we should have tile groups. Like what you were saying about having parts of a tileset active. But let's take it one step further and have a little tile set for say the golden palace, another for the dead woods etc etc.

Cool! That's quite a final product!

For a UI of that, I'd imagine that you'd have your map on the left hand side, a working tileset in the middle, and the bulk/searchable tile list on the right.

For making a map, you would be able to grab tiles from either your working tileset or the full list. The working tileset would be something of a favorites list, where you could put grass, or trees to get back to them quickly. To make the working tileset, you would find the appropriate tiles in the list, and drag'n'drop them into slots on the tileset.

For the list, I could see it being wider than one tile, and so, when you filter for say, the cave tiles, a small block of cave related tiles would show up. You could then select a bunch of them(like getting stuff from tilesets in the current editor) and move the whole group over into your working tileset.

that sounds great! also if the user adds tiles from the big list directly from the list, the map editor could add them to the working tilset.

i think it's important that the map would refrence the tiles unique id (from the database) not it's position in the working tileset.

indeed.

I was thinking that the tileset would actually be a second map, and selecting from in would be the same as using the grabber tool.

That's how it works at the moment, I think.

Tileset switching now works nicely, though I've not yet compiled all the lists yet. Thus, it will still only work with the objects layer.

Cool, good work!