using an OSD serialization rather than auto forces serialization code to be placed in OpenSim.Framework
this makes the media texture raw data staging post in PrimitiveBaseShape redundant, now we just directly call the code in PrimitiveBaseShape.MediaList itself
THIS WILL BREAK ANY EXISTING MEDIA TEXTURE SERIALIZATIONS. If you're testing this, please start with new databases.
This makes media textures serialized in the same way, which is probably better in the long term.
As far as I can determine, r3287 + r3294 patch was the previous update to the OpenMetaverse libraries
This change just adds r3378 to overcome problems storing media textures with certain permission combinations in inventory
This is a limited change in order to isolate moap from any other possible libomv update issues
An update to the forthcoming libomv 0.8.3 is expected in due course to replace this
This commit also deletes OpenMetaverse.Utilities.* as it's unused (on the advice of jhurliman).
This is better modularity. It also allows MoapModule to be replaced with some other media module that may behave completely differently in the future.
Remaining non-modularity:
PrimitiveBaseShape needs explicit Media and MediaRaw fields. MediaRaw is required in order to shuttle the pre-serialization data back and forth from the database layer.
The database also needs to know about MediaRaw though not about Media.
IMO, it would be extremely nice to remove these hard codings but this is a bridge too far at the present time.
THIS WILL BREAK EXISTING MEDIA TEXTURE PERSISTENCE. Please delete your existing sqlite databases if you are experimenting with this branch.
This wrapper will make it easier to maintain compatibility if the media texture data evolves.
This will also make it easier to store non-sl media texture data.
Media flags is cleared via a direct TextureEntry update from the client. If the clearing leaves no media textures on the prim, then a CAP ObjectMediaUpdate is not received. If there are still media textures present then one is received.
This change fixes drag-and-drop on Windows (and Mac?) clients. It may also fix problems with clearing and then subsequently setting new media textures.
scripts using these functions should now compile but I don't know how well the methods themselves work yet
llSetPrimMedia(), at least, appears to have problems when a current url is set for a face that doesn't yet have a texture
at the moment, only checking for the exact name prefix is implemented
for some reason, whitelists are not persisting
this commit also fixes a very recent problem where setting any media texture parameters after the initial configuration would not work
Other avatars can now see the webpages that you're navigating to.
The requirement for an initial prim click before the texture displayed has gone away.
Flash (e.g. YouTube) appears to work fine.
Still not persisting any media data so this all disappears on server restart
This allows the media texture to persist in memory - logging in and out will redisplay it (after a click) though navigation will be lost
Next need to implement media uri on prim and delegate more incoming llsd parsing to libomv
Justin Clark-Casey (justincc)