[TuxSH]

To me this looks like a bad attempt at exposing dirty laundry in bad faith, which is not too surprising coming from him.

1. Commit 3ba50ec which Marcan is complaining about was pushed in 2008 and didn't just delete attrib specifically, but all (?) VCS comments indiscriminately. The file was barely touched since

2. "The authors of libogc didn't just steal proprietary Nintendo code (...) ignorance about the copyright implications of reverse engineering Nintendo binaries" ---> AFAIK it's software RE work, and nothing done in the console hacking scenes is truly cleanroom at all, and there's no point to it either as Nintendo can knock&talk and/or send strongly worded letters when they please, legality be damned

I don't know much about the Wii scene specifically, and libogc seems to be a mess in general, but what I do know is that libctru (3ds)/libnx (Switch) don't have that drama nor made the mistakes made in libogc

[delroth]

> AFAIK it's software RE work, and nothing done in the console hacking scenes is truly cleanroom at all

There's a wide gradient of how much effort people put into reverse engineering consoles in a legal way vs. just copying code straight from their decompiler and slapping an open source license on it. libogc is very much on the "didn't even try" side of that gradient, it's been known since pretty much forever, and even their documentation is straight up copied from Nintendo's SDKs for part of their libraries.

What's new here is discovering that even the parts people thought were developed "fresh" and not just straight-up asm2c'd from Nintendo are actually stolen from other open source projects in a way that tries to conceal the origin of the code.

Whether you'll find that "more morally reprehensible" or not will largely depend on your personal morals, but clearly for some people that seems to be the case...

[TuxSH]

Yes, libogc is a dumpster fire and the dkP org would be better served by rewriting a libogc replacement (w/ a different API) from scratch, quite honestly.

What I find odd is the timing, I highly suspect he learned about it many months ago.

> There's a wide gradient of how much effort people put into reverse engineering consoles in a legal way vs. just copying code straight from their decompiler and slapping an open source license on it.

Agreed (I replied the same in another comment)

[JoshTriplett]

> To me this looks like a bad attempt at exposing dirty laundry in bad faith, which is not too surprising coming from him.

It seems odd that you would complain about the messenger, here, since it seems you don't actually dispute the message.

> Commit 3ba50ec which Marcan is complaining about was pushed in 2008 and didn't just delete attrib specifically, but all (?) VCS comments indiscriminately.

So it's OK that they did something wrong because they did everything wrong?

> there's no point to it either as Nintendo can knock&talk and/or send strongly worded letters when they please, legality be damned

There's very much a point to it (when you're building an emulator or tooling, rather than e.g. romhacks where it's unavoidable), because if you carefully stay entirely above board, you can burn those strongly worded letters, make DMCA counter-notices, and otherwise rely on the fact that both emulation and reverse-engineering are legal.

[amiga386]

Hector just doesn't have enough drama going on in his life after resigning (twice!) from the LKML, eh?

Nonetheless, I respect him for calling out FOSS license violations, even if he's done it in the most drama-pilled way.

[amszmidt]

> 1. Commit 3ba50ec which Marcan is complaining about was pushed in 2008 and didn't just delete attrib specifically, but all (?) VCS comments indiscriminately. The file was barely touched since

FWIW -- I think this is the commit being talked about: https://github.com/m87h/libogc/commit/3ba50ecd4134ef37a0f18f...

Looks honestly totally innocent ...

[scripturial]

This by itself just looks to me like someone getting rid of the old CVS comments that are messing up the headers. Maybe the problem is that younger people don’t recognize what a cvs header is?

[amszmidt]

Yeah entirely possible.

The threading code does seem to be "similar" enough to warrant some investigation, but from the looks no license notice was "removed". Even the threading code in question has never had one, from the initial import from CVS/Subversion.

Making a mountain out of a molehill .... before the molehill is even there?

[sunaookami]

>I don't know much about the Wii scene

It shows. It's an open secret to everyone in the Wii scene that libogc is based on proprietary Nintendo code.

>but what I do know is that libctru (3ds)/libnx (Switch) don't have that drama nor made the mistakes made in libogc

Because WinterMute is not behind them.

[userbinator]

which Marcan is complaining about

"That prick again?" Not surprised at all. He's been trying to stir shit up for a long time, and best ignored as a troll.

[userbinator]

Yes, for those who don't know, this prick has even tried to take on Linus, with the expected result: https://news.ycombinator.com/item?id=42978765

(Read the dead comments there - very enlightening.)