exactly.

I do think that the source code of source was leaked. Its just that valve has to themselves give the code of source. FLOSS community would really do it. Trust me.

Well for the FLOSS community to do anything Valve would not just need to release the coede but also put it under a free software license, which isn't the case even for the game code released here.

Modders in general won't care though.

Well in a sense you are right that it is lazy, in that it would be quite a lot of work to disentangle those parts.

So they could probably do it, but it would be costly, while just releasing the TF2 code is basically free.

Don't bother with subtle disentanglement. Just take an axe to it and release a broken mess. It will be fixed and running within a month.

It depends on how the code is structured. It might be that the licensed parts are sprinkled throughout the functions, or that they would have to reorganize things to not reveal too much about a specific NDA'd technology (like a console SDK)... The resulting code might end up being just as useless as no code at all, and I assume it would be very hard to get the management to allocate resources for that.

Perhaps, but a multiplatform codebase won't call the console SDK directly from hundreds of random functions. And even if the developers were that crazy those calls need to be excluded when building for other platforms so can also be stripped form the source code in an automated fashion.

> Perhaps, but a multiplatform codebase won't call the console SDK directly from hundreds of random functions.

You’d think so, but some developers have a very adversarial relationship with abstractions, such as abstracting away platform-specific code behind a common interface. It happens more often than you’d think.

> And even if the developers were that crazy those calls need to be excluded when building for other platforms so can also be stripped form the source code in an automated fashion.

So they’re going to leave people code that either doesn’t compile or, if it does, doesn’t work? And what’s stripped out may have very important technical considerations that might only be known based on which function calls were yanked out.

> So they’re going to leave people code that either doesn’t compile or, if it does, doesn’t work?

I think that in general, potential open source developers would not care at all about that. Some code (that doesn't yet work) is better than no code. As someone else mentioned, a motivated open source group could probably easily fill in all the missing gaps within a month.

There is always the risk that some of the code (non 3rd party) may look a lot similar to code of a previous employer of the person that committed it

It's all legal risk with little commercial upside