I question how desirable that even is. So like some rich kid buys a golden suit and can wear it in a bunch of games...I mean, sure, ok? Not really groundbreaking stuff. Would you buy a game because it lets you do that? Assuming the total cost is the same, I'd just as soon buy different skins in each of my games...assuming I care about skins at all.

And that's setting aside the really significant implementation challenges.

As a player I think I would find it less than desirable to be playing Darkest Dungeon and view Halo content next to cartoon Fallguys. The NFT owner probably wants more than their NFT transformed into bytes and used as a random seed to determine the grain on their wooden shield. They almost certainly want a direct map to their resource.

The developer still needs to implement how this skin or texture is applied in their specific game, so there's still net additional work per skin or texture. You don't just get it for free as soon as someone else implements it in one game.

I'd be curious if even that would be so simple. There could be competing blockchains or types of NFTs. You might also get into the situation where instead of every company building the software to verify ownership they just pay specialized companies.In that situation Web3 technology isn't really all that different than a company building a virtual shop API. A company like Valve could probably do this better than Opensea could and without all the extra Web3 technology.

I agree, and would go even further - is this even desirable to the creators of the game? Sounds a lot like shooting themselves into the foot. They profit off having closed ecosystems and kids having to buy skins instead of already owning them from somewhere else

(This is of course assuming some theoretical process where the NFT has a bunch of data attached that makes it Just Work, instead of it requiring Capcom to sink a bunch of development time into making assets for characters that can only be unlocked by having paid money to another company entirely.)

Nevermind the technical and practical concerns. What does it even mean to apply a Fortnite skin to a battleship or a racecar?

I am with you, but I am a bit more cynical after watching my 6-10 year old nephews get into gaming. All they care about are the skins, it's insane to me but they really like it. Knowing that makes me dislike this angle more because I feel like their entire market vertical will be 6-10 years olds with their parents money. Not to mention the whole, creating addicts out of kids in the name of digital consumerism.

There's this idea of the "metaverse" which I think is informed entirely from fiction and hype. It's very detached from the actual incentives at play.

Is this even viable? Can a skin from say an Unreal engine game be copied over to and used in a Unity engine game?

With difficulty, yes. If you have the asset in the creator tools, rather than just inside the game as a game user, it's straightforward. There's some standardization in this area, and it's getting better. There are now open formats. Look up "GLTF" and "USD" formats. Until about two years ago, most content was interchanged in proprietary formats, but we're now past that. The Khronos Group, which standardized Vulkan, is behind this standardization effort.

Currently, you can usually move rigid objects from one platform to another, but articulated skeletons and scripted objects need manual rework to port.

"Ripping" models from games is a thing. There are tools.

NFTs do not enable asset portability. They are part of a DRM system. The goal of the DRM is to prevent more asset portability than someone wants. This only works if the receiving system pays attention to the DRM. There's no technical obstacle right now to finding the JPEG for a Bored Ape Yacht Club and printing it as a poster for your wall. But you won't be able to mint it as a new NFT and sell it on OpenSea, because OpenSea checks.

We had fbx for decades, only difference is that it wasn't open source. Even better USD is an art format, nearly every concept it supports is for art direction, comparison of various designs, hiding of details that may get in the way of some modellers work. It isn't dynamic, it doesn't even have the concept of level of detail, so it will render a 10 million poly model even if all you see is a three pixel wide spot on the screen.

The technical problem is probably the smaller issue. The larger issue is that the publishers of the respective games have to allow this type of cross-usage, and they do not have an incentive to make their games interoperable with other games.

Even in business software, where there are practical incentives to interop, compatibility at the level of asset sharing remains very rare.