is John Carmack's opinion still as "relevant" in the industry?
Read Masters of DOOM: How two guys created an empire and transformed pop culture and you'll have a good grip on what Carmack has done over the years. As for now? Well, he is certainly relevant - of course. But so is any other professional with the proper credentials working in this field - whether it's Joe at the bar or the graphic god John. So, really, it doesn't matter where the technique comes from - just learn and be grateful so many are willing to share (which, btw, was always Johns strongest point - his ethics).
None of which is really relevant. Clean and modular does not imply slower.
super inteliigent mathmatician he may be but i wouldnt ask him to work with a team of web developers. His code is a total mess. EPIC wrote clean modular slower maintainable code.
if he didn't spearhead his own company with his enthusiasm he would have been that guy who just hacks shit together which nobody else can understand.[/quote]Precisely. When you are ahead of the curve, nobody can understand you.
Once you can write clean, modular, picture perfect code you are working on commodity, might as well outsource it or license it from third party. Hence my original comment. Unless you are pushing the envelope by directly profiting from it, use commodity stack. It's simply too expensive to develop new tech for single use. That, or use a one-off implementation which solves the immediate need and move on.
EPIC commoditized the market. So unless you have some completely different venue, trying to compete on engine is absurd.
6 months down the line the code breaks because carmack got another job.[/quote]
Are you sure?
Doom source ports
If it's so well engineered, how easy would it be to port Unreal engine on every single OS and device in existence. These are not rewrites or emulations, they are port of original source code.
Besides, do you have Unreal source code to compare? Unreal engine has, over the years, suffered from some fairly horrible and almost inadmissible flaws in many high-profile titles. The original (pre-3) also didn't age well due to so many hacks it broke on plenty of hardware. Same reason - at the time UE was still pushing the limits. It no longer needs to do that today, having effective monopoly over industry.
You can't really hold up Unreal as a marvel of great engineering -- I've been in several interviews where I was asked to spot the errors in ~20-50 lines of code, and after the exercise was finished, that code was revealed to be none other than code taken directly out of UE3. Once, I was told in advance that the code in question came from "a popular middleware game engine" and I actually missed naming the error -- not because I didn't spot it, but because I did and thought to myself "Red Herring! Professional code couldn't possibly contain such novice mistake." I guess that was my own naivety though 
Not to pick on Epic specifically. I've worked as a technical writer (API documentation) as well as in a capacity where I reviewed source code for many AAA retail and Xbox Live Arcade titles, including games which used UE3, so I've been privy to quite a bit of large-scale "professional" code. I can say that, in my experience, the code *almost always* sucks as far as readability is concerned, and that no one is immune to general block-headedness. Therefore I've been lead to the conclusion that the majority of programmers are hired for their willingness to suffer long hours and get more than their share of work done using whatever means necessary, rather than their ability to produce good design and good code, even under deadline pressure. Obviously this must work out, since not many studios are crushed under the weight of their own shitty source code, but I wish they would strive for better -- after all, even if the average consumer will never see, much less understand, that code, it turns out that I, in my own little psuedo-hell, *will* have to
Of all the code I've seen, only Remedy (Alan Wake, Max Payne) get a reprieve. Their code is a beautiful thing to behold. I'd work for that team any day.
Not to pick on Epic specifically. I've worked as a technical writer (API documentation) as well as in a capacity where I reviewed source code for many AAA retail and Xbox Live Arcade titles, including games which used UE3, so I've been privy to quite a bit of large-scale "professional" code. I can say that, in my experience, the code *almost always* sucks as far as readability is concerned, and that no one is immune to general block-headedness. Therefore I've been lead to the conclusion that the majority of programmers are hired for their willingness to suffer long hours and get more than their share of work done using whatever means necessary, rather than their ability to produce good design and good code, even under deadline pressure. Obviously this must work out, since not many studios are crushed under the weight of their own shitty source code, but I wish they would strive for better -- after all, even if the average consumer will never see, much less understand, that code, it turns out that I, in my own little psuedo-hell, *will* have to
Of all the code I've seen, only Remedy (Alan Wake, Max Payne) get a reprieve. Their code is a beautiful thing to behold. I'd work for that team any day.
he would have been that guy who just hacks shit together which nobody else can understand.
Considering the number of independent competitors the iD engines have spawned, I whole-heartedly disagree.
http://en.wikipedia.org/wiki/Quake_engine#Games_using_the_Quake_engine
http://en.wikipedia.org/wiki/Quake_II_engine#Games_using_the_Quake_II_engine
http://en.wikipedia.org/wiki/Id_Tech_3#Uses_of_the_engine
http://en.wikipedia.org/wiki/Id_Tech_4#Games_using_id_Tech_4
http://en.wikipedia.org/wiki/Id_Tech_5#Games_using_id_Tech_5
They may be losing ground these days, but back in the day there were a lot of brand-new studios that took enormous financial and emotional risks with a code that "nobody else can understand".
Praise CryEngine and Unreal, but don't act like Carmack was a hack.
The thing about John is that he is not sitting in his office, wistfully hoping that game engine technology comes back around to ray casting. He is on the curve at worst and ahead of the curve at best. At all times. If a company comes out of left field with a game engine technology that he hasn't yet conceived of, he, in very short order, not only becomes an expert on the tech, but improves on it.
When he suggests that game engines should go back to 13h mode and ray casting, then I'll devalue his opinion.
When he suggests that game engines should go back to 13h mode and ray casting, then I'll devalue his opinion.
The thing about John is that he is not sitting in his office, wistfully hoping that game engine technology comes back around to ray casting. He is on the curve at worst and ahead of the curve at best. At all times. If a company comes out of left field with a game engine technology that he hasn't yet conceived of, he, in very short order, not only becomes an expert on the tech, but improves on it.
When he suggests that game engines should go back to 13h mode and ray casting, then I'll devalue his opinion.
I wouldn't necessarily agree with that. Now, I'm not denouncing anything Carmack has done, but part of his mystique can be attributed to the fact that he happened to be at the right place at the right time. There probably are quite a few people just as competent as he is, but weren't around in the field at the time in the early 1990's to come up with a 3D Engine. In today's technology spectrum, it becomes harder to prove your worth, mainly due to fewer opportunities.
As for relevance, obviously the answer is no. Of course his opinion is still relevant, and more relevant than most people, but not to the extent it was a decade ago. Time moves on and there are other competitors. If you ask people to name an awesome looking upcoming game,games using the Rage engine may or may not come up. 15 years ago though, there would have been no debate. It would have been Quake.
When he suggests that game engines should go back to 13h mode and ray casting, then I'll devalue his opinion.
until it turns out he is actually right somehow D:
[quote name='RivieraKid' timestamp='1300133123' post='4785728']
he would have been that guy who just hacks shit together which nobody else can understand.
Considering the number of independent competitors the iD engines have spawned, I whole-heartedly disagree.
http://en.wikipedia....he_Quake_engine
http://en.wikipedia....Quake_II_engine
http://en.wikipedia....s_of_the_engine
http://en.wikipedia....using_id_Tech_4
http://en.wikipedia....using_id_Tech_5
They may be losing ground these days, but back in the day there were a lot of brand-new studios that took enormous financial and emotional risks with a code that "nobody else can understand".
Praise CryEngine and Unreal, but don't act like Carmack was a hack.
[/quote]
most of those games are just swapping out models / textures / weapons and adding a few behaviours.
The most successful games (HalfLife + COD) turn the code inside out. They probably had no choice.
Daikatana / Deus ex -> Id tech / Unreal tech. Says it all really.
UnrealScript is highly readable. The underlying crap is abstracted away so it doesn't matter.
The potential problem I see with Carmack is that because his skill is around hardware and specifics of what hardware can do, he can very quickly get left behind an end up being someone who is an expert in X, but X isn't how things work anymore. Like if you take an expert C programmer and make them use Java, you end up with C-in-Java. I don't know if he's reached the age he has become inflexible, hopefully not, but it's a risk for anyone specialising in such a fast-moving area.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
