Burvey, black seams? besides testing this on a good windows7 setup i'm testing this on a brand new nexus7 wich has a very high dpi, and i don't see a single seam, possible to do a screenshot? too see how it relates to the original.

anyone else are seeying seems on adjusted version?

The biggest issue i have concerning this, is when i preview my game with my optimal settings (tested on chrome,firefox,ie) i have no seams, and i mean none, not less, i now also tested it with high contrast tiles,if it wasn't the case i simple wouldn't have bothered to raise this issue

then when i export, suddenly there are, so its an export problem

this to me, means that the technology is perfectly fine to display tiles with floating positions and no seams, but on export the texture atlas needs tweaking, wich i done and is possible to recreate the preview quality

adjusted texture atlas

https://dl.dropboxusercontent.com/u/61666915/noseamhighcontrast/index.html

normal exported result (seams are really bad,because of upscaling)

https://dl.dropboxusercontent.com/u/61666915/seamhighcontrast/index.html

hmm, i was hoping you would at least look further into the issue then just stating, use letterbox integer and pixel rounding, if adjustments in the texture generation can improve or remove seaming, its worth checking out imho.

letterbox integer scale is much harder to get right on multiple devices, its way less dynamic, point scaling results in jittery motion, so anyone who wants to do smooth scrolling or upscaling gets a less then optimal result, so its not good in all circumstances.

there are two options to remove seaming, padding is for texturebleeding,extruding is for the opacity problem (gaps)

have you checked the texture options of texturpacker? If it wouldn't have any effect, the options wouldn't exist

some reference links:

http://talkbinary.com/cocos2d-2/how-to-fix-strange-lines-in-cocos2d-by-using-extrude-in-zwoptex/

http://www.matim-dev.com/texture-packer---get-rid-of-sprite-artefacts.html

http://forum.starling-framework.org/topic/tile-anti-alias-pixels

tilemap with same settings as post above

https://dl.dropboxusercontent.com/u/61666915/tilemaptest/index.html

if you scale browser up and down, you can see that the second line of tiles has bleeding of the toprow (green line), so i think tilemap should improve with padding and extrude

tgeorgemihai

i just checked the tilemap export, and it works differently than sprite and tilebackground, on export it compacts all the tiles, no padding

but i think internally it works still the same and probably has the same issue, so its something ashley should look into

—

i see what you mean now about tilemap, it has pretty bad seams (even on preview)

on another note, i just tested the tilebackground, and where i know in the beginning of c2 had the same seam problem, now works perfect without adjusting

letterbox scale / fullscreen high q / high dpi / linear / pixel rounding off

https://dl.dropboxusercontent.com/u/61666915/tiledbg/index.html

the seems are the flickering black lines between the tiles,(background is showing true)

i agree this is not the best demo because some of the tiles are not oriented wright but if you compare the two demolinks, you should notice it pretty well.

—

i didn't overlap or make the tiles bigger in the editor, its the same export, only on the second i made changes in the export texture-atlas (extend the last tilepixel border 1pixel where the 1px gap is on export)

this doesn't make the sprites bigger or changes the position(ingame)

i think this fix can be used for tilemaps, sprites used as tiles and tiling bg for use with letterbox scale and linear sampling

i also got a description here from texturepacker program

Border/shape padding: The space between the border of the sprite sheet and the sprites themseves. If you see artifacts from neighboring sprites in your game, you may want to increase the values here to add more of a separation between the edges/sprites.

Extrude: This repeats the pixels around the borders of sprites. This is the opposite of adding padding ? if you see transparent ?gaps? around the edges of your sprites, you may wish to set this to a higher value.

are you doing on collision with family > set family fx brightness ?

this should work for picking the wright one.

Ashley

i was aware of the seaming problem when using something other then point sampling, but i dislike the jittery motion it gives,and it gives the impression of bad performance

so i went thru with linear sampling and to my suprise i had no real seaming while running previews on pc and mobile,that is until you finally export.. bad seaming..

while utterly sad and broken <img src="smileys/smiley17.gif" border="0" align="middle">, i found it odd that seams only showed on export

so i searched, and its like we know a problem of float position and bleeding of the texture, to prevent that, on export there is a 1 pixel border, this fixes the textures bleeding in each other

but here's the thing, while it doesnt bleed in another texture, the 1px transparent border still bleeds into the texture, and thats the problem..

the fix: extrude 1px border from the texture outwards

this is a real approach and its also used in texturepacker

demo project with seams

https://dl.dropboxusercontent.com/u/61666915/seams/index.html

same exact project with borders extruded (in photoshop) / no seams

https://dl.dropboxusercontent.com/u/61666915/noseams/index.html

my suggestion would be, to be able to flag a sprite as tiling and then generate this textures with a 1px extrude, only question is, where the texture borders meet, do you extrude each 1px what means you need a 2px gap, but i guess thats the only way to do it.

anyway i'm happy that i can continue without seams, yey!

just watched a good intro movie to webrtc, alot of answers, think its gonna be good

found how you get the <element> name (not the value)

just myxml.StringValue("name()")

you can do: on collision, set sprite.var to sprite.y

subcondition: sprite pick instance with highest var,

i couldn't make it work with pick by comparison,

oh man, great, it works, thank you !

would you know if its possible to get the actual text instead of the value

ex. > name="Playerstatus" you get "name"

i want to make something that works for multiple xml structures, so that would be nice