Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

John Carmack on expert witnesses and 'non literal' copying

facebook.com

Post reply on HN
Print view View original

191–200 of 322 posts

Re: John Carmack on expert witnesses and 'non literal' copying

#191

Post by Ygg2 »

jonknee wrote: post #178
gregw2 wrote: post #162

It's kinda hard to square John's post-trial comments "I never tried to hide or wipe any evidence, and all of my data is accounted for, contrary to some stories being spread." with ZeniMax's post-trial comments at http://www.gameinformer.com/b/news/archive/2017/02/01/zenima... that " (vi) Carmack intentionally destroyed data on his computer after he got notice of this litigation and right after he researched on Google…

> Carmack intentionally destroyed data on his computer after he got notice of this litigation and right after he researched on Google how to wipe a hard drive Somehow I think John knows how to wipe a drive...

Every dev knows how to wipe a drive. Buy an identical drive. Throw old drive into the incinerator.

Re: John Carmack on expert witnesses and 'non literal' copying

#192

Post by jartelt »

dpark wrote: post #151
dragonwriter wrote: post #109

Earlier quoted context omitted.

> Jurors are not expected to actually learn anything about the law Since jury instructions both ask for a determination of whether a particular legal standard was met (phrased in legal terms), and explanation of what facts (in lay terms) the jury must find to say that that standard was met, I think your statement is not completely accurate.

> (phrased in legal terms) This was not my experience. My jury was given a simplified description of the law with only relevant sections discussed.

I had a similar experience. After jury selection, the trial seemed to start almost immediately and I was surprised that we were not given more background information or preparation before opening statements started. During the trial I felt like there was a lot of pressure to not ask any questions or ask for any clarifications. Even when the jury asked for a repeat of the definition of reasonable doubt, the request had to be talked about by the lawyers and the judge. Then during deliberation, we had to ask permission to view different pieces of evidence, which generally added friction to the process.

Re: John Carmack on expert witnesses and 'non literal' copying

#193

Post by ISL »

galacticpony wrote: post #184
ISL wrote: post #60

Earlier quoted context omitted.

I once espoused a similar opinion, but have since changed my mind. I'm a physicist, and once expressed to a judge I had met on the trail that I wished I could be judged by my academic peers. It was my belief that they would be more receptive to logical argument, understanding in their analysis, and more committed to fundamental principles. The judge was visibly surprised, but said little. What later changed my mind w…

"What later changed my mind was the realization that our society is a human one, and that humanity has a huge spectrum of experience. A group of physicists may not understand something about the world that others do. Furthermore, any select group of citizens will differ in belief somewhat from the interests of society at large." Why the hell would that realization change your mind? For society at large, that "huge sp…

As an example for the first point: Fewer-than-average professional scientists have known the desperation of poverty nor the devastation wrought by broken homes, disability, or crimes of passion. Scientists may be somewhat better prepared to find facts in a case, but may have diminished sensitivity to both the motivation for a crime and for the impact of a verdict.

An educated, reasoning, and principled jury pool should not be an exception. It should be the expectation. We should expect it of our society as a whole and our school system in particular.

Taking the time to share interesting and useful insights to people at a moment when they're interested in exchanging ideas can be very effective.

Re: John Carmack on expert witnesses and 'non literal' copying

#194

Post by simplicio »

ue_ wrote: post #25
addisonj wrote: post #10

I don't have the evidence so I can't make a judgement about whether Zenimax or Carmack is in the wrong here, but this does point out something strange in our court system and that is where a "jury of peers" which translates to normal people with little to no knowledge of the underlying subject matter. As our society becomes more specialized it seems a bit absurd to have people make judgements mostly based on how well…

This is a very common argument against juries, and I think it's one of the most powerful. Of course if you start to fill a jury with people competent in the area being judged, you run different risks, such as bias within the field. And it's still not clear that it would help with delivering justice. But if we take the idea of a jury to be about judging by one's peers, it should be patently clear (heh) that the jury o…

I think even if the field had a strong bias, even one I disagreed with, I'd still prefer it. One of the frustrating things about the US court system is how unpredictable it is. At some points in working with IP and corporate law, I've found myself just wanting to know what the rules are, so I can go ahead and follow them.

But way too often, I've had lawyers basically tell me that whether we should pursue something is basically a function of how much money we're willing to put into pursing it in court convolved with some vague probability function of how 12 random people will feel about how trustworthy our expert witness looks versus how the other guys.

I'd gladly trade a more "just" system for a more predictable one.

Re: John Carmack on expert witnesses and 'non literal' copying

#195

Post by r3xpi »

jonknee wrote: post #178
gregw2 wrote: post #162

It's kinda hard to square John's post-trial comments "I never tried to hide or wipe any evidence, and all of my data is accounted for, contrary to some stories being spread." with ZeniMax's post-trial comments at http://www.gameinformer.com/b/news/archive/2017/02/01/zenima... that " (vi) Carmack intentionally destroyed data on his computer after he got notice of this litigation and right after he researched on Google…

> Carmack intentionally destroyed data on his computer after he got notice of this litigation and right after he researched on Google how to wipe a hard drive Somehow I think John knows how to wipe a drive...

This is not a valid assumption. Shredding a drive to the extent that data recovery is not possible requires knowledge that, often times, revolves around the architecture of the drive in question (see: wear leveling).

An unrelated, yet illustrative, counter example is the seemingly reasonable, yet false, assumption that RMS performs his own installs of GNU/Linux.

Re: John Carmack on expert witnesses and 'non literal' copying

#196

Post by maxlybbert »

Retric wrote: post #77
geofft wrote: post #66

Earlier quoted context omitted.

Only because judges get to tell the jury what certain phrases in the contract mean, because judges are experts at understand contract complexity. We don't actually expect juries to learn contract law; the judge interprets the relevant parts of the contract for the jury, and tells the jury to decide whether it was broken.

Meaning what? Jury's can decide complex issues if there is a neutral expert on complex topics? If anything that sounds like a complaint about expert witnesses not juries.

A judge can appoint a special master for situations like this. For instance, Lawrence Lessig's role in the Microsoft antitrust trial. However, as far as I know, juries don't have the option of asking a special master's opinion, because there is a big risk juries would just defer to the special master instead of actually trying to judge the case.

Re: John Carmack on expert witnesses and 'non literal' copying

#197

Post by arca_vorago »

zardo wrote: post #175
gregw2 wrote: post #162

It's kinda hard to square John's post-trial comments "I never tried to hide or wipe any evidence, and all of my data is accounted for, contrary to some stories being spread." with ZeniMax's post-trial comments at http://www.gameinformer.com/b/news/archive/2017/02/01/zenima... that " (vi) Carmack intentionally destroyed data on his computer after he got notice of this litigation and right after he researched on Google…

I wouldn't trust the testimony of a paid expert witness. They always have strong convictions in favor of the side paying the bill, even if the evidence available clearly can't support such absolute conclusions. Even if they aren't just liars, the lawyers can just keep trying different 'experts' untill they get one who tells the story they want to hear. It's just cover for making an argument from authority.

I did an advanced forensics course a few years back that was geared towards preparing me for court cases for the company I was working for. I quickly realized how easily "expert" witnesses could and most likely were being abused in the system just as you say. Expert shop till you get what you want.

That's not even mentioning the epic level of incompetence I saw among others. I'm talking about "forensic experts" who didn't know any filesystem other than NTFS, and even then if the tool didn't do it for them they didn't see it... At least I know if I ever end up defending myself I have plenty of inside knowledge on how to undermine the prosecutors arguments.

Re: John Carmack on expert witnesses and 'non literal' copying

#198

Post by GlennS »

dragonwriter wrote: post #41
AnimalMuppet wrote: post #32

Earlier quoted context omitted.

If the "expert witness" is either lying (in court) or incompetent, shouldn't there be consequences to their professional reputation? Or do you think such things should be consequence-free?

There is already a consequence to lying as a witness, it's called perjury prosecution. It's unfortunately infrequently pursued when the perjury occurs in a civil case, but if you want to agitate for consequences, agitate for prosecutors willing to enforce the law. (Or for a civil cause of action for perjury, but that would have to be carefully crafted to prevent an infinite regress of lawsuits.)

What are the actual requirements for a perjury conviction in the USA? Does the witness have to be intentionally lying? Presumably you don't prosecute someone for being mistaken, misguided or over-confident in their own abilities?

It doesn't sound like Carmack was accusing the witness of lying, but rather they they believed a bit too much in their own pet theory about how to detect copying in code.

I don't usually like to comment on trials, but I think perhaps he has a point here. If there were a reliable method for detecting 'non-literal copying' of code, then the person who made it should be making a killing by selling refactoring tools.

Re: John Carmack on expert witnesses and 'non literal' copying

#199

Post by bisby »

That's a great analogy that I'm surprised I haven't heard before.

If I write a book called Gary Potter, about a boy who finds out he's a wizard and goes on (specific) adventures with his friends Jon and Germione... I'd be in trouble.

But a book about someone wishing for a better life > finding out they secretly had great powers > struggle to gain control/mastery of the powers > must use the powers to save the world from evil... that's just a story framework and applies to Harry Potter, Star Wars, Hercules(the Disney version at least). There's variations in why they want the better life (abuse, boredom, etc), how they find out, and what the exact evil is. but if you abstract far enough, they're all basically the same story.

Using exact code and changing variables like "main_character_name" is clearly infringement, but using a similar abstracted flow shouldn't be. ESPECIALLY in code. If your end goal is to put pixels on a screen. you run algorithms to determine pixels > add those pixels to a buffer > send the buffer to the screen. There's not many other ways to do it. There's only so many ways to implement some things. And if a way is good, independently coming up with the solution isn't copying. In fact, if I came up with a different solution, it would be sub-optimal. For some problems, the abstracted flow of the code is defined for you. "We're doing VR, we want X, Y, Z to happen"... There's not a whole lot of options on how to do that.

If Oculus only figured out the flow-charts based on copying zenimax's prototype, then they perhaps are copying. And thats the argument zenimax is shooting for. "this is a non-trivial, non-obvious solution they only got by copying us".

But the next question is: at what level of abstraction am i just changing variables at a large scale? If I wrote a book about a young boy living with abusive relatives because his parents died and he finds out he's a wizard and moves away to a magical wizard school, does that count as copyright infringement? There's a point where it is, and a point where it isn't.

I can never find the source, but I once saw a 256x256 picture of yoshi, that immediately below it had a 128x128 version of the same picture (stretched to 256x256). Below that was 64x64, etc until the 1x1 was just a green pixel. Along the side it said "at what point does it stop being copyright infringement? If you started at the top (the high res), you could get pretty far down and still go "oh yeah, this is still the same image". but if you started at the bottom, you could get much higher and go "this is some weird abstract art of random pixels"

Re: John Carmack on expert witnesses and 'non literal' copying

#200

Post by Strom »

LeicaLatte wrote: post #187

Carmack is a hero for me but he seems to come across as very naive here. I employ other people now and employees who argue this sort of stuff isn't copying are toxic to the culture and ethics of work. When you are paid a salary every month, I actually don't care if you are productive. But the least I expect you to be is loyal. And I am no monster for expecting that.

Loyalty in a free market is a pipe dream. Extremely few employers will stay loyal to their employees when times get rough, say when an employee gets very sick and can't work for several years. Thus no rational person should have any long term loyality to an employer if the market providers better alternatives.
Post reply on HN