halflife

Reverse engineering existing game binaries seems kind of silly when a lot of the most important AAA knowledge is already available to the general public. https://github.com/ValveSoftware/halflife/blob/master/pm_sha... The LLMs have presumably already consumed information like this as part of their training sets. Converting between Hammer and Unity scale is a fairly trivial linear operation. You can dump the BSPs to recover geometry and rapidly accelerate map development using timings that are already known to work. Attacking the raw binary directly is certainly impressive, but it's totally unnecessary.

https://github.com/ValveSoftware/halflife on the other hand is the full Half-Life (GoldSrc) engine, but it's not quite an open license.

https://github.com/ValveSoftware/halflife is the HL1 engine, is it not?

They bought the HL1 rights back from Sierra in the early 2000s. The real problem is that the code is not in a distributable state and nobody at Valve feels like working on putting it into a distributable state ( https://github.com/ValveSoftware/halflife/issues/1712#issuec... ): > A while back Valve [had] a partner perforce server that had depots of the source files for both gold source and source that were shared with development partners and some mod teams. This server had a major meltdown and those depots were lost. At the time there was no requests and no activity around gold source development. Resources to rebuild the depots did not exist and still don't so that code is just not available. Once a year someone talks about maybe pulling it together to open source it but once again there are not resources to do the actual work need to p

There is an open Half-Life 1 SDK on Valve's GitHub [1], not sure if it's missing something regarding the engine. [1] https://github.com/ValveSoftware/halflife

Funnily enough, on Half-Life 1 engine-based games (i.e. the engine that came before HL2 - on which Team Fortress 2 runs; such as Counter-Strike 1.6), a different allocator problem exists -- glibc's malloc() just decides to fail miserably[0] on some setups. [0] https://github.com/ValveSoftware/halflife/issues/3158

The original Half Life, including goldsrc, is source-available: https://github.com/ValveSoftware/halflife

Even Steam itself is "just enough." Off the top of my head, the Steam Linux client: * Ignores flags (so `steam --help` for instance just starts Steam the same way `steam` does) * Ignores signals (including SIGTERM, I believe) * Screws up my tiling window manager sometimes, so it probably ignores EWMH I personally think Valve needs to just separate the core Steam functionality into some backend and allow users to write their own frontends. =========== And now for an unrelated "Valve/Linux" rant while I'm here: Half-Life: Opposing Force has been unbeatable on Linux since its release. The final boss's death animation plays too quickly and the trigger to change to the next map never occurs. There's been an issue open on the ValveSoftware/halflife GitHub repo about this since 2013. And the real kicker? Someone (not a Valve employee) actually w

Interestingly, Half-Life (released in 1998) uses a slightly more complex table-based generator: https://github.com/ValveSoftware/halflife/blob/master/dlls/u... Note that this is only used for certain parts of the game. For others it uses a closed source non-table based generator.

I think it's here, as part of the HalfLife source: https://github.com/ValveSoftware/halflife

Look at the number of issues! https://github.com/ValveSoftware/halflife/issues