Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
John Carmack Interview (nowgamer.com)
148 points by shawndumas on April 13, 2011 | hide | past | favorite | 19 comments


I had the great pleasure of meeting Mr. Carmack recently (here's the photo proof: http://jacoblear.com/carmack.jpg ) and was able to listen to him talk casually, as opposed to in an interview or speaking event. He doesn't turn his brilliance on and off, it's there all the time.

When I was a kid, I wanted to do 2 things, make video games and go to space. Here we have a man who is a legend in the games industry, and also started a company sending rockets into space. He's awesome.


This is why Carmack keeps shipping incredible products:

"I am more excited about what I’m doing and what I’m working on than ever before."

Great advice to those of us building companies, along the lines of the "do something you love and you'll never work a day in your life" quote banging around here recently


"I just turned 40 and I can be programming for another 40 years; there’s a lot yet to do!"

I love John Carmack.


"...At the last Quakecon I took a show of hands poll, and it was interesting to see how almost as many people there had an Android device as an iOS device. But when I asked how many peple had spent 20 bucks on a game in the Android store, there was a big difference. You’re just not making money in the Android space as you are in the iOS space."

Interesting


Apple are very good at getting people to hand over their credit card details before they even intend to make an app store purchase. Then, when it comes to buying something via an Apple store it doesn't require filling in any annoying forms to make the payment. On Android meanwhile, once you see something you want to buy you then have the tedious task of typing in numbers and for a trivial app many people will feel it's not worth the effort.


But...are there any $20+ games on the Android market? Are there any $20+ iPhone games, for that matter?

People will accept significantly higher prices for iPad games, because there's a different class of games that can be played on a tablet device. A phone's too small.

There's a bit of a catch-22 going on here in general, where Android users are spending less money on apps because there are fewer really good apps available. I'm one person who owns multiple devices. But I wind up spending lots of money on iOS because 1) there's plenty of good stuff, and 2) the app store is nicely designed to make me buy things.


The question he asked at Quakecon 2010 was actually "who has spent at least $20 on the Android app store?"[1], not just on a single game. Your catch-22 scenario still holds, though I think the quantity and quality of free apps in the Android app store has also raised the bar for pay apps.

[1] http://www.youtube.com/watch?v=1_SdC8LVODY&feature=playe...


I'm always happy to hear when developers optimize for refresh rate over graphical complexity, especially if competitive play will be involved.


Do you think 60Hz is the right rate? Just seems like such a huge jump over 30. Does 48 make more sense?


In competitive play it makes a big difference.

Each frame of a 60Hz game spans 16.66ms, while in a 30Hz game it's 33.33ms. Depending on how well the engine is coded, there will be at least one frame of latency between input and reactions, and maybe more. So we could be looking at the difference between 33 and 66 milliseconds of input latency, at least. This doesn't even consider latency from input to the computer, or from the computer to the display.

Now consider that we react to events within about 200-300ms. 66ms of latency on top is going slow things down considerably, and when games are poorly coded and dump a few additional frames on top - which could happen because of propagation through events and subsystems, network latency, or animation concerns - the control starts to feel "wrong" in a hard-to-identify way.


Digital Foundry has a series of interesting articles about measuring input lag in console games; I think this is the first: http://www.eurogamer.net/articles/digitalfoundry-lag-factor-...


Interesting article. I'm not much of a console gamer (not much of a gamer anymore in general) but those figures are horrifying for a PC gamer to look at! 100-150ms lag (not counting any network lag on-top) would be pretty hard to swallow for me, but I'm biased and have a bit of a purest view on these things. I guess the most important thing is that you have fun - in which case, if the game achieves this but lags a bit, who cares?


Agreed on it making a difference in competitive play. Though 200-300ms is a pretty slow response time (especially in a competitive environment - depends on the game I suppose). In most twitchy games (actually in most games) people's reactions would be in the 20-100ms range.

One of the interesting things that can happen if you v-sync a game that's capped to 60FPS on a screen which has a refresh rate that is not a multiple of 60 (e.g. 70 or 100, etc) is the addition of mouse lag to the game (there's an added delay between moving the mouse and actually seeing the action happen on the screen).


You can't control the actual refresh rate of the monitor or TV. The display refreshes at 60hz, you can draw every frame, or every other frame and have twice the time.

And yet, like he says, Rage looks better than most 30hz games.


Depends on the monitor. You're pretty much stuck when it comes to most current TVs and LCD/LED displays as well (60Hz, 70Hz, 75Hz sometimes 100Hz).

Before flat screens were affordable we were all using CRTs. A good CRT could push a 150Hz refresh rate. When Quake3 was still being played competitively (it /kind/ of still is with QuakeLive, but it's not the same anymore imo) we ran the game at a capped 120 FPS, set the monitor refresh to 120Hz and then turned on V-sync which equated to the smoothest and most responsive experience I've ever had in a game - it's pretty much indescribable and most people would argue that you wouldn't be able to tell the difference anyway, but even though solid 60 FPS on a v-synced 60Hz screen in most modern games comes pretty close, you can definitely feel the difference.

The worst is when you have some FPS lag with v-sync enabled - then it's a lot more noticeable than without v-sync.

I remember that during Quake4's (the 60FPS cap was hardcoded in Quake4, and somehow part of the engine, which in theory at least, made 120FPS/120Hz redundent) short run on the comeptitive scene, players complained that they wanted the cap lifted from 60FPS to 120FPS (or just any configurable number) so they could play on 120Hz monitors. id finally caved and released a patch which allowed this, and this made a huge difference. It definitely gave you an edge over the 60/60 players.


Thank you for pointing this out. People sometimes think I'm crazy when i talk about pushing really high fps. The human eye can detect objects that exist for 1/200th of a second, so pushing higher framerates will always make things smoother. I'm really surprised that all lcd monitors aren't running at 120hz or 240hz, obviously they can't be that hard to make if the tv manufactures are pushing them out.


You're right, I've argued with a few of my friends over this in the past. Also, something I should have mentioned before (and I should correct myself), the refresh rate on a LCD screen isn't quite the same thing as that on the CRT. So, even though a 120Hz LCD would refresh the image 120 times a second, for some reason it just doesn't feel as nice and fluid as on the CRT.

That v-sync mouse lag I mentioned has only ever happened to me when I'm using an LCD rather than CRT - which is the main reason why I wouldn't use v-sync when playing an FPS.


Depends on the game. For really twitchy FPSs, you'll want at least over 40. UT2k4 players (and others... just I had a lot more interactions with them) would bitch and moan about sub-60 fps.

The scary thing is that they weren't always bullshitting about that.

Also, as sister reply said, going at 60 means no screen tearing on -anything-. Don't need to vsync, or anything nasty. TVs, computer monitors, projectors, cellphones, whatever, one whole new frame per refresh. yum.


Out of interest, why 48 specifically?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: