Same as game development. When I started in '94, we had brand new top spec 90MHz Pentium based PCs. When designing the next set of games, we were told to aim for a recommended spec of 200MHz. 200! Seemed like pure science fiction to me! I found it hard to believe this level was going to be reached in about two years, but sure enough, it was.
I'll concede, though, that games and OSs are completely different kettles of fish. Whereas you'd expect a game to use every ounce of power the machine had, the OS better be doing something special to require those specs. If it hasn't got a fully functional voice interface, then I'm going to be disappointed!
Errrrrr.. no. The technology is getting more and more complex. Sure, it's easier to write a previous generation game on current generation hardware, than previous generation hardware. Using all the bells and whistles the new tech provides certainly isn't easy.
However, you're spot on about content, though. It seems that we're heading towards project teams with a handful of programmers and a army of artists, animators, modellers and musicians.
Researchers at Georgia Tech are concluding their two-year distributed analysis of network usage, concluding that most bottlenecks were, in fact, caused by NETI@Home traffic.
I was having a discussion about Noah's ark and the flood and all that with a Jehovah's Witness who came to the door one day. His explanation was that prior to the flood, the Earth was surrounded by clouds and that every morning, a mist would descend to water the land. Then came the great flood, when God purged the land of the evil people and animals (but not ducks or fish, apparently), where it rained and rained and rained and rained and indeed the whole Earth was covered with water. That's why you find fossils of sea creatures up mountains, you see. Then, vast underground caverns opened up, the water drained in and formed the pattern of land and seas that we are familiar with today.
Rrrrrrright...
It's a more logical argument than I'd heard before - I don't know if it's recorded anywhere, or this guy concocted it to support his own beliefs. It did kinda leave me on my doorstep waving my finger and opening and closing my mouth unable to find any words for a moment.
Thomas Warfield, the "another developer" mentioned above, notes the distribution of the games through third party portals and the reliance on this as a failure point. He goes on in another post to mention online shareware/software distributors such as Digital River, eSellerate and SWREG. I'd be interested in hearing anyone's experiences of these services, especially involved with generating income from smaller games (ie. 1 or 2 person projects). Are there hidden costs that impose a break-even sales limit, and so on?
When I first bought Dark Age of Camelot, it ran fine on my laptop, but it sat back and laughed at my voodoo3-clad desktop PC. I tried playing it, but at a rate of one frame every second or so in some of the more built up areas, it must have been pretty entertaining for other players watching me staggering into walls, and taking several attempts to open a door and walk through it.
SWG is even more severe. It refuses to run at all on any machine whose graphics card does not support hardware transform and lighting. This ruled my laptop out completely. New pc buyers beware, as I believe that the integrated Intel graphics chipsets sold in many a new budget pc will be found lacking...
I, too, have found disks recorded at a slower rate to be more reliable than faster burns. Maybe it's down to the individual burner?
I was writing/testing software reading data off a CD. The burner allowed me 1x, 2x and 4x record speeds, but we often found disks recorded at 4x suffered from read errors. There was always pressure from management to record test disks at 4x for speed, while I wanted nothing more than 2x for reliability. I hated spending hours tracing through for spurious errors only to find they're caused by duff disks.
A few years back, a CV landed on my desk, belonging to a guy who was applying for a programming position. On it, he mentioned running as an interest, and added that since his high school physical education teacher told him the secret behind breathing correctly while running, he can now run pretty much any distance without getting tired. Unfortunately, he didn't quite make the grade for the programming job, but I must admit I was tempted to get him in for interview just so I could ask what the secret was.
Since when does marketing decide that a product is finished or not?
All the flipping time.
And if you're really lucky, they'll tell you.
Take this from someone who found release candidate n had been approved, duplicated and shipped to retailers, while he was still up all hours submitting candidate n+2 for testing.
Don't go to the self checkout, though. I spent five minutes at HEB the other night arguing with the machine over whether or not I'd put the damned yoghurt in the fecking bag already.
if (dist > 1.0)
{
CheckTrajectoryAndMakeChanges();
}
else
{// Too late for any more course changes
printf("And now, the end is near;\n");
printf("And so I face the final curtain.\n");
printf("My friend, I'll say it clear,\n");
printf("I'll state my case, of which I'm certain.\n");
printf("I've lived a life that's full.\n");
printf("I've traveled each and ev'ry highway;\n");
printf("And more, much more than this,\n");
printf("I did it my way");
I had a similar experience.. I was asked to finish off the PlayStation1 version of a game when the previous programmer left. The programmer who left seemed like a really good guy, and pretty clued up about what he was doing. When I finally got my hands on his code, I found that vast chunks were useless #ifdef'd out junk, and many sections were just empty functions stubs with// TODO comments. To make matters worse, I'd quizzed him about memory consumption, because with only 2Mb on the console, we needed to be careful. He said it was fine, the mesh compression algorithm was great and he had plenty of spare memory. It wasn't until he'd gone, I found he was running the devkit in 8Mb debug mode, with only about 3Mb left free. Pretty much had to start that project from scratch.
Insult+injury : He'd crammed in some ridiculously high polygon meshes, arranged them in a game-like form and taken screenshots. Management knew it was a complete fabrication at the time, but were expecting the final title to look like that. Still, despite a rewrite, one or two of these old "screenshots" remained on the box and in the instruction manual.
Re:Warning: Coffee contains DHMO
on
Death by Coffee?
·
· Score: 1
All y'all whining about moderation need to just relax.
The example shows the development costs as $600k, plus licencing fees, which leads me to believe this is quick+easy conversion, not a brand new original title.
Take a small to medium sized dev team - 8 programmers, 10 artists, 1 designer, 1 project manager, 1 audio guy. Give them 18 months to take a game from concept to finished product. Even with a modest average salary of $35k, you're looking at over $1.1m for wages alone. Add in equipment costs, software licences, non-dev staff, QA staff, office rental, insurance, utilities, and so on, and you're looking at much higher figures for development costs. At this point, you're going to be looking at much higher than 90k sales required to return a profit.
A method to reduce the incidences of buffer overflow exploits would be change to an ascending stack, so local data/buffers aren't immediately proceeded by return addresses. This won't stop buffers further down the stack being overflowed for exploits, but I'll bet a significant number of exploits are due to local buffers.. overflowing these won't break anything until it runs into other code or data in memory. You might cause a crash, but you're unlikely to be able to execute your own malware.
Alternatively, split the stack into two distinct stacks : one for data, one for return addresses. Ok, it's not going to solve all cases - I could have a local function pointer on the data stack that I call, which a buffer overflow could alter. Again, I'll bet the majority of overflow exploits are due to alteration of the return address, which this will get around with minimal impact to run time performance.
Just rambling thoughts.. please tear these theories to shreds!
I was kind of hoping that this means they will, too. I've tried switching to Linux twice - the last time I found that it did a pish poor job of printing docs on my HP 720c deskjet. That rendered it pretty much useless for my needs at the time, so Linux was rarely booted. Surely, if they're going to distribute Linux PCs, they'd have to distribute appropriate drivers for their peripherals?
I'm not quite sure what sort of tools you mean. There are plenty of middleware companies producing audio libraries, video codecs, physics simulations, 3d graphics engines, networking libraries and so on. Main game loops are pretty simplistic at top level, and any further in soon becomes extremely specific to whatever game you're making.
I've never fully understood the rules for this. How many generations are you allowed to go? Am I just English? English-English? Or Viking-English? Or African-English, if you trace back the origins of mankind?
Does it work on birthplace or citizenship? If I became a US citizen, would I then be English-American, or still just English?
Oh bugger.. I just rambled right off topic. I'm off to mod myself down.
As a games developer, we got a Japanese import of the game when it was released, for urm, research purposes. Unfortunately, none of us could read Japanese, so the manual and on-screen instructions were muddled through by trial and error.
The voice control was a bit disappointing. The microphone sat on a short stalk, plugged into a memory card slot on the controller, if I recall. Our Seaman wasn't progressing much, even with plenty of vocal encouragement - "Hello little fishy! How are you today? Who's my fishy-wishy?". It then occurred to us that being a Japanese import, it probably only understood Japanese. Cue many awful impressions of kung-fu film inspired Japanese speech. That didn't work, either, but it provided an afternoon's entertainment.
Slashdot Mirror
Same as game development. When I started in '94, we had brand new top spec 90MHz Pentium based PCs. When designing the next set of games, we were told to aim for a recommended spec of 200MHz. 200! Seemed like pure science fiction to me! I found it hard to believe this level was going to be reached in about two years, but sure enough, it was.
I'll concede, though, that games and OSs are completely different kettles of fish. Whereas you'd expect a game to use every ounce of power the machine had, the OS better be doing something special to require those specs. If it hasn't got a fully functional voice interface, then I'm going to be disappointed!
Damn you - I have coffee all over my keyboard now...
Errrrrr.. no. The technology is getting more and more complex. Sure, it's easier to write a previous generation game on current generation hardware, than previous generation hardware. Using all the bells and whistles the new tech provides certainly isn't easy.
However, you're spot on about content, though. It seems that we're heading towards project teams with a handful of programmers and a army of artists, animators, modellers and musicians.
Researchers at Georgia Tech are concluding their two-year distributed analysis of network usage, concluding that most bottlenecks were, in fact, caused by NETI@Home traffic.
I was having a discussion about Noah's ark and the flood and all that with a Jehovah's Witness who came to the door one day. His explanation was that prior to the flood, the Earth was surrounded by clouds and that every morning, a mist would descend to water the land. Then came the great flood, when God purged the land of the evil people and animals (but not ducks or fish, apparently), where it rained and rained and rained and rained and indeed the whole Earth was covered with water. That's why you find fossils of sea creatures up mountains, you see. Then, vast underground caverns opened up, the water drained in and formed the pattern of land and seas that we are familiar with today.
Rrrrrrright...
It's a more logical argument than I'd heard before - I don't know if it's recorded anywhere, or this guy concocted it to support his own beliefs. It did kinda leave me on my doorstep waving my finger and opening and closing my mouth unable to find any words for a moment.
Thomas Warfield, the "another developer" mentioned above, notes the distribution of the games through third party portals and the reliance on this as a failure point. He goes on in another post to mention online shareware/software distributors such as Digital River, eSellerate and SWREG. I'd be interested in hearing anyone's experiences of these services, especially involved with generating income from smaller games (ie. 1 or 2 person projects). Are there hidden costs that impose a break-even sales limit, and so on?
When I first bought Dark Age of Camelot, it ran fine on my laptop, but it sat back and laughed at my voodoo3-clad desktop PC. I tried playing it, but at a rate of one frame every second or so in some of the more built up areas, it must have been pretty entertaining for other players watching me staggering into walls, and taking several attempts to open a door and walk through it.
SWG is even more severe. It refuses to run at all on any machine whose graphics card does not support hardware transform and lighting. This ruled my laptop out completely. New pc buyers beware, as I believe that the integrated Intel graphics chipsets sold in many a new budget pc will be found lacking...
I, too, have found disks recorded at a slower rate to be more reliable than faster burns. Maybe it's down to the individual burner?
I was writing/testing software reading data off a CD. The burner allowed me 1x, 2x and 4x record speeds, but we often found disks recorded at 4x suffered from read errors. There was always pressure from management to record test disks at 4x for speed, while I wanted nothing more than 2x for reliability. I hated spending hours tracing through for spurious errors only to find they're caused by duff disks.
3.8774 Libraries of Congress per second.
What? They don't store porn? I don't believe you.
I'm no runner, trust me on that one!
A few years back, a CV landed on my desk, belonging to a guy who was applying for a programming position. On it, he mentioned running as an interest, and added that since his high school physical education teacher told him the secret behind breathing correctly while running, he can now run pretty much any distance without getting tired. Unfortunately, he didn't quite make the grade for the programming job, but I must admit I was tempted to get him in for interview just so I could ask what the secret was.
Was it just BS or is there really a secret?
Since when does marketing decide that a product is finished or not?
All the flipping time.
And if you're really lucky, they'll tell you.
Take this from someone who found release candidate n had been approved, duplicated and shipped to retailers, while he was still up all hours submitting candidate n+2 for testing.
Wha? Meteorologicos exterior! Sminky pinky...
Don't go to the self checkout, though. I spent five minutes at HEB the other night arguing with the machine over whether or not I'd put the damned yoghurt in the fecking bag already.
I had a similar experience.. I was asked to finish off the PlayStation1 version of a game when the previous programmer left. The programmer who left seemed like a really good guy, and pretty clued up about what he was doing. When I finally got my hands on his code, I found that vast chunks were useless #ifdef'd out junk, and many sections were just empty functions stubs with // TODO comments. To make matters worse, I'd quizzed him about memory consumption, because with only 2Mb on the console, we needed to be careful. He said it was fine, the mesh compression algorithm was great and he had plenty of spare memory. It wasn't until he'd gone, I found he was running the devkit in 8Mb debug mode, with only about 3Mb left free. Pretty much had to start that project from scratch.
Insult+injury : He'd crammed in some ridiculously high polygon meshes, arranged them in a game-like form and taken screenshots. Management knew it was a complete fabrication at the time, but were expecting the final title to look like that. Still, despite a rewrite, one or two of these old "screenshots" remained on the box and in the instruction manual.
All y'all whining about moderation need to just relax.
Maybe they've had too much coffee this morning.
The example shows the development costs as $600k, plus licencing fees, which leads me to believe this is quick+easy conversion, not a brand new original title.
Take a small to medium sized dev team - 8 programmers, 10 artists, 1 designer, 1 project manager, 1 audio guy. Give them 18 months to take a game from concept to finished product. Even with a modest average salary of $35k, you're looking at over $1.1m for wages alone. Add in equipment costs, software licences, non-dev staff, QA staff, office rental, insurance, utilities, and so on, and you're looking at much higher figures for development costs. At this point, you're going to be looking at much higher than 90k sales required to return a profit.
A method to reduce the incidences of buffer overflow exploits would be change to an ascending stack, so local data/buffers aren't immediately proceeded by return addresses. This won't stop buffers further down the stack being overflowed for exploits, but I'll bet a significant number of exploits are due to local buffers.. overflowing these won't break anything until it runs into other code or data in memory. You might cause a crash, but you're unlikely to be able to execute your own malware.
Alternatively, split the stack into two distinct stacks : one for data, one for return addresses. Ok, it's not going to solve all cases - I could have a local function pointer on the data stack that I call, which a buffer overflow could alter. Again, I'll bet the majority of overflow exploits are due to alteration of the return address, which this will get around with minimal impact to run time performance.
Just rambling thoughts.. please tear these theories to shreds!
I was kind of hoping that this means they will, too. I've tried switching to Linux twice - the last time I found that it did a pish poor job of printing docs on my HP 720c deskjet. That rendered it pretty much useless for my needs at the time, so Linux was rarely booted. Surely, if they're going to distribute Linux PCs, they'd have to distribute appropriate drivers for their peripherals?
How about M, B, or W?
Don't go too over the top with those, though, else expect a letter from BMW's legal dept.
I'm not quite sure what sort of tools you mean. There are plenty of middleware companies producing audio libraries, video codecs, physics simulations, 3d graphics engines, networking libraries and so on. Main game loops are pretty simplistic at top level, and any further in soon becomes extremely specific to whatever game you're making.
Yeah, but, do you remember those happy days when it wasn't?
I've never fully understood the rules for this. How many generations are you allowed to go? Am I just English? English-English? Or Viking-English? Or African-English, if you trace back the origins of mankind?
Does it work on birthplace or citizenship? If I became a US citizen, would I then be English-American, or still just English?
Oh bugger.. I just rambled right off topic. I'm off to mod myself down.
Good Lord! She's even given them internet access!
As a games developer, we got a Japanese import of the game when it was released, for urm, research purposes. Unfortunately, none of us could read Japanese, so the manual and on-screen instructions were muddled through by trial and error.
The voice control was a bit disappointing. The microphone sat on a short stalk, plugged into a memory card slot on the controller, if I recall. Our Seaman wasn't progressing much, even with plenty of vocal encouragement - "Hello little fishy! How are you today? Who's my fishy-wishy?". It then occurred to us that being a Japanese import, it probably only understood Japanese. Cue many awful impressions of kung-fu film inspired Japanese speech. That didn't work, either, but it provided an afternoon's entertainment.