With or without gzip, 12.5mbit is easy and cheap. A 2.4ghz Celeron with a 20mbit unmetered Cogent connection goes for $239 US/mth at ServerMatrix. For these big sites complaining about bandwidth, $239 per month is peanuts.
I consider netcode to be the method of which the server and client communicate game states, and how that data is used. Unlagged makes several changes to the way that the client deals with that data, such as implementing much better client-side player movement prediction, as well as server-side latency correction.
I think you should understand what exactly moding entails. It is not an addition, it is a modification, of whatever sections of the engine source the company has chosen to make available. In the case of Quake 3, while the low-level network handling code isn't provided, as I understand it the modder has control over what goes over that link.
HL1 and HL2 have excellent netcode, probably the best available on the market. If you are going to claim otherwise you're going to have to give a reason. HL1/HL2 work well in low bandwidth situations, is very adjustable based on available bandwidth server and client side, has good movement prediction, features latency correction, works well even with packetloss, and so on. It is generally quite reliable. And it is most assuredly better than Quake 3, especially from the server side. Unlagged makes a great deal of progress, however, to improving Quake 3's playability under worse than normal conditions. Despite the main feature of the mod being the latency correction, the new movement prediction probably has a greater impact.
UT2003/2004 has "ok" netcode. On a low latency low-loss link, it is rather decent, though the lack of latency correction found in many modern games is a downside. However UT2003/4 don't handle packetloss or high latency very well.
BF1 was horrible. I have very limited experience with BF2, so I can't really comment on it, but from what I've seen it has made at least some small improvements. How small I can't tell without further play.
As Carmack has said himself, the Quake 3 source was supposed to be released as soon as DOOM3 was released, however Carmack underestimated the life of the Q3 engine. When D3 shipped, there were still new Q3-based games in development, and it wouldn't have been fair to the licencees.
Obviously the last of the Q3-based games have now shipped.
The later iterations of the Quake 3 engine had a much improved server browser. It has yet to be seen what version/branch of Quake 3's source will be released.
Regardless, the server browser is such a minor part of the engine, and it's a truely trivial addition to make.
Only on models, not the entire game world. And last time I tried to enable them, shadows were super buggy, madly flashing in and out of existence in parts, and distorted.
Why bother. Make a mod for Enemy Territory. It's based on the Quake 3 engine (though more advanced of course), supports mods, and it's free to download. Why bother messing around with building a full game with a bare engine when you can borrow all of ET's assets?
MP3 decoding is a patented process, and especially for projects with a smaller music budget, many developers can't afford to license it.
Console manufacturers require hefty licence fees just to develop games for a platform, so MP3 licencing fees are probably not a big deal. CPU power really isn't a big deal. The official licencing fee for MP3 is 75 cents per unit, meaning per shipped copy of a game, or US $50k one-time. As I mentioned, peanuts compared to the platform licencing fees and publisher fees.
The Gamecube's 485mhz PowerPC processor can handle such an operation with ease. After all, we've been decoding MP3 since the days of the 486, and that was before we had any fancy instruction set extensions;)
Riiight. You can stick any card you want in, but that doesn't mean you'll be able to use it. As I understand it, there are special mac versions of soundcards and videocards that were required.
Besides, I'm not the one you should be argueing to. I'm responding to all the people who are claiming that OSX-x86 will never work properly on whiteboxes due to the lack of drivers. If there is in fact a driver shortage, I think people will step up to the plate and write their own.
At least the WarioWare series (currently two GBA titles, one GameCube title, and one Nintendo DS title) uses MIDI because it needs to vary the tempo of each individual song from 140 BPM to over 300 BPM. And not many GBA games use compressed audio, even though the software decoder is available and permissively licensed.
Again, handhelds exist in a different marketspace than consoles. For one thing they are generally 1 to 2 (5 to 10) years behind consoles in capabilities, so the current crop of handhelds are equivalent to the SNES (GBA) or N64 (DS). I do know that PSP games can and do use compressed audio where possible (See Lumines as an example). I'm sure others don't. The DS certainly has the power to decode compressed audio, but not the storage space. The carts only store up to 128MB.
You overestimate how much processing power is required to decode simple compressed audio. Vorbis is a bit more demanding, but there are lots of other solutions in use. The PSP supports MP3 and ATRAC3plus, for example, and probably uses one of the two for games.
What you say may be true of the Xbox and PlayStation 2 but not of the GameCube and PSP. GameCube discs have about 1.5 GB of space; PSP UMDs have 1.8 GB. A quarter of that for music will get you roughly 400 MB.
PC games have been doing pre-recorded audio with a heck of a lot less than 400MB for years. Early games used hybrid CDs to store uncompressed audio on the disc that was played during gameplay. An example of this is Warcraft 2. Both the GameCube and PSP have more than enough power to decode MP3 or a similar format. MP3 decoding is virtually free on the modern CPUs in the GameCube (PowerPC derivitive IIRC) and the PSP. You want to put 400MB aside for audio, fine. That's enough for 7 hours of music, which is more than virtually all games on the market. I can't name one with that much composed music, that's for sure.
Now the consoles' built-in wavetable synths use some form of ADPCM (4 bits per mono sample with predictive filtering) as a native waveform encoding, giving only 2.5 hours at 44100 Hz stereo (352 kbps).
I think you are confusing your terms. Wavetable synthesis is a method for producing music from notational data such as MIDI. Most modern devices use wavetable synthesis for their synthesized audio.
Compare this to the typical play time of an RPG. Modern RPGs use prerecorded compressed audio, full stop. All of Square Enix's RPGs use compressed audio, mostly by noted composer Nobuo Uematsu. Interestingly, Uematsu has been composing Square's music since the 4-channel synth (Squarewave I think) to the wavetable synth on the SNES and PSX to the later use of compressed audio.
Or were you planning on using Vorbis decoded in software, which eats a lot of CPU time compared to ADPCM decoded in hardware? Even then you get only three times that. No, Vorbis is overkill. MP3 will do, or even simpler compressed formats. 3x that is more than enough, since you grossly overstate how much music games have by at least a factor of 10.
What is the budget of a Square Enix RPG? The budget is irrelevant, the music budget is what is important and it is nowhere close to 2 million dollars. Square enix games don't have anywhere close to 42 hours of music. Final Fantasy 7 has something like 2 or 3 hours of music, and even the latest Final Fantasy X has a similar amount. We're talking about less than 200MB of music for a Squaresoft game that is shipped on a DVD.
You seem to be thinking as if it were still 1990. We've progressed a bit in the past 15 years.
You do realize that pretty much every game made in the last 5 years has featured compressed or CD audio?
If the game ships on DVD, it has roughly 9.4GB to work with. Put a quarter of the DVD towards compressed audio, that is 42 hours of compressed audio. At the cost listed in the article, let's say $800 per minute, that much music would cost over 2 million dollars.
So as you can see, storage space is NOT a limitation. As much music as you can produce for a game, you can store. I can't speak for the DS or PSP, handhelds may still use enduser wavetable synth. But for PC and console, they are all prerecorded now and have been for years.
I believe there will be a small to medium community porting linux drivers to OSX to allow people to use unsupported hardware, both in their official Apple boxes and their whiteboxes.
And more people than just John Carmack made DooM 3. However, like Carmack, Tim Sweeney is "the engine guy" who does the graphical engine. I'm not certain if he truely does the entire graphics engine himself, though it is certainly possible.
You misunderstand. I refer to the composer purchasing a synthesizer and samples, and using that to produce a score. The audio in the games is simply compressed audio, or a soundtrack for a TV show.
The cost savings I was referring to was the cost of the composer purchasing a professional synthesizer and professional samples, as opposed to repeatedly hiring an orchestra to record the performances. The cost of high-end hardware may well be more than the cost of the orchestra for a game, but you can use the hardware to produce each piece whereas you have to hire the orchestra each time.
I was not referring to end-user synthesizers. They are not up to the level of quality of which I described.
I've found that the higher quality synthesizers with the high-end samples are virtually indistinguishable from a real orchestral performance, as far as soundtracks go at least. The cost of entry for quality hardware and samples might be high, but you can re-use it for no additional cost, unlike a real orchestra.
I understand that this is why several television composers use their own synths instead of a real orchestra. I believe the music for StarGate is done on synth rather than real orchestras.
Is there a difference between live and good synth? Yes. Is it big enough to matter in most circumstances where there is a lower budget? No, the average person can't tell (and doesn't care about) the difference.
Don't get me wrong, I love a good orchestral rendition of a videogame soundtrack.
I'd rather have seen them earmark the money for a spam-fighting foundation.
Think about it, create a mostly financially independant child company or foundation, or even division, and give them the $7 million. That entity then uses the $7 million to sue other spammers. It then gets more money from other spammers, and uses that to fight additional spammers.
This way, Microsoft could have a spam-fighting operation going costing them nothing, since it would make it's own money to continue operating. They'd also get a lot of good press about it.
Also there is something poetic about using spammers' money to sue other spammers.
Problem is if a user pays for software they have a reasonable expectation of support for that software. Not tech support, but general support. Software updates are considered support, for example.
There are adapters that let you use mice and keyboards to control any game by emulating a regular controller. They also let you use PS/2 (the connector type) keyboards and mice with the PS2 (the console).
Very unfortunate if they did (try to charge $200 for a keyboard).
I am considering buying an xbox 360. The decision hinges on the ability to use it with my unused 19" CRT monitor. It's in excellent condition, but a laptop means I don't use it anymore.
The 360 is supposed to have VGA out, so that's that.
But of course, I'm a PC gamer. I was hoping that I'd be able to use a mouse and keyboard to control FPS games on the 360. After all, the 360 has USB ports, so it's theoretically possible if there are standard keyboard/mouse drivers.
But if devices need to be licensed to work, I can forget about this, because the mouse and keyboard are obviously not licensed.
And so the situation you describe could very well happen, although we might end up in an even more rediculous situation; we made need an adapter to hook up a keyboard and mouse to the 360, like we do for the current xbox and PS2. This time around, though, the adapters and the keyboards plugged into them will have the same type of connector. So stupid!
I meant the pennies as parts of a packet (Individual bits even) and the random number of pennies as a whole packet.
The same thing happens in the wired as wireless world; two people transmitting at the same time causes an overlapping signal like two people talking at the same time, and nobody can understand what is being said by either party.
Wireless networks do incorporate several techniques to avoid or even prevent collisions, but that just stops them from happening; you still can't have more than one transmission at a time.
Is the premise the same in the air as it is on a wire? Do signals on the same frequency really interfere with each other that much?
Yes.
Think about it like this, you and your friend are standing in front of a bowl. You have a bunch of pennies. You randomly let some fall out of your hand into the bowl. Your friend writes down how many pennies are in the bowl, 5. That is how many pennies you dropped. You clean out the bowl.
Next your friend does the same thing. He drops 7. You write that down and take them out.
That is sort of how wireless works.
Now, imagine if you were both to both drop the random number of pennies into the bowl at the same time. There are now 12 pennies in the bowl. Who dropped how many pennies? It is impossible to know! All you know is there are 12 pennies, not who put which ones in.
That is called a collision, and when it happens in the wireless world, you would clear out the pennies, wait a random amount of time, then dump more pennies back in hoping your buddy waited a different amount of time than you. Usually this works, if not, try again.
So, you see, when you broadcast a wireless signal, you are putting that signal into the air (Well, not really the air, but you understand). Once it is out there it is free to mix and mingle with any other signal out there. You must simply hope that while you are sending your signal (packet), nobody else interrupts you, or else a collision will occur.
Slashdot Mirror
With or without gzip, 12.5mbit is easy and cheap. A 2.4ghz Celeron with a 20mbit unmetered Cogent connection goes for $239 US/mth at ServerMatrix. For these big sites complaining about bandwidth, $239 per month is peanuts.
I consider netcode to be the method of which the server and client communicate game states, and how that data is used. Unlagged makes several changes to the way that the client deals with that data, such as implementing much better client-side player movement prediction, as well as server-side latency correction.
I think you should understand what exactly moding entails. It is not an addition, it is a modification, of whatever sections of the engine source the company has chosen to make available. In the case of Quake 3, while the low-level network handling code isn't provided, as I understand it the modder has control over what goes over that link.
HL1 and HL2 have excellent netcode, probably the best available on the market. If you are going to claim otherwise you're going to have to give a reason. HL1/HL2 work well in low bandwidth situations, is very adjustable based on available bandwidth server and client side, has good movement prediction, features latency correction, works well even with packetloss, and so on. It is generally quite reliable. And it is most assuredly better than Quake 3, especially from the server side. Unlagged makes a great deal of progress, however, to improving Quake 3's playability under worse than normal conditions. Despite the main feature of the mod being the latency correction, the new movement prediction probably has a greater impact.
UT2003/2004 has "ok" netcode. On a low latency low-loss link, it is rather decent, though the lack of latency correction found in many modern games is a downside. However UT2003/4 don't handle packetloss or high latency very well.
BF1 was horrible. I have very limited experience with BF2, so I can't really comment on it, but from what I've seen it has made at least some small improvements. How small I can't tell without further play.
Q3's netcode is pretty shoddy until you add in Unlagged. Then it's netcode performs on par with good modern netcode.
As Carmack has said himself, the Quake 3 source was supposed to be released as soon as DOOM3 was released, however Carmack underestimated the life of the Q3 engine. When D3 shipped, there were still new Q3-based games in development, and it wouldn't have been fair to the licencees.
Obviously the last of the Q3-based games have now shipped.
The later iterations of the Quake 3 engine had a much improved server browser. It has yet to be seen what version/branch of Quake 3's source will be released.
Regardless, the server browser is such a minor part of the engine, and it's a truely trivial addition to make.
Only on models, not the entire game world. And last time I tried to enable them, shadows were super buggy, madly flashing in and out of existence in parts, and distorted.
Why bother. Make a mod for Enemy Territory. It's based on the Quake 3 engine (though more advanced of course), supports mods, and it's free to download. Why bother messing around with building a full game with a bare engine when you can borrow all of ET's assets?
MP3 decoding is a patented process, and especially for projects with a smaller music budget, many developers can't afford to license it.
;)
Console manufacturers require hefty licence fees just to develop games for a platform, so MP3 licencing fees are probably not a big deal. CPU power really isn't a big deal. The official licencing fee for MP3 is 75 cents per unit, meaning per shipped copy of a game, or US $50k one-time. As I mentioned, peanuts compared to the platform licencing fees and publisher fees.
The Gamecube's 485mhz PowerPC processor can handle such an operation with ease. After all, we've been decoding MP3 since the days of the 486, and that was before we had any fancy instruction set extensions
Riiight. You can stick any card you want in, but that doesn't mean you'll be able to use it. As I understand it, there are special mac versions of soundcards and videocards that were required.
Besides, I'm not the one you should be argueing to. I'm responding to all the people who are claiming that OSX-x86 will never work properly on whiteboxes due to the lack of drivers. If there is in fact a driver shortage, I think people will step up to the plate and write their own.
At least the WarioWare series (currently two GBA titles, one GameCube title, and one Nintendo DS title) uses MIDI because it needs to vary the tempo of each individual song from 140 BPM to over 300 BPM. And not many GBA games use compressed audio, even though the software decoder is available and permissively licensed.
Again, handhelds exist in a different marketspace than consoles. For one thing they are generally 1 to 2 (5 to 10) years behind consoles in capabilities, so the current crop of handhelds are equivalent to the SNES (GBA) or N64 (DS). I do know that PSP games can and do use compressed audio where possible (See Lumines as an example). I'm sure others don't. The DS certainly has the power to decode compressed audio, but not the storage space. The carts only store up to 128MB.
You overestimate how much processing power is required to decode simple compressed audio. Vorbis is a bit more demanding, but there are lots of other solutions in use. The PSP supports MP3 and ATRAC3plus, for example, and probably uses one of the two for games.
What you say may be true of the Xbox and PlayStation 2 but not of the GameCube and PSP. GameCube discs have about 1.5 GB of space; PSP UMDs have 1.8 GB. A quarter of that for music will get you roughly 400 MB.
PC games have been doing pre-recorded audio with a heck of a lot less than 400MB for years. Early games used hybrid CDs to store uncompressed audio on the disc that was played during gameplay. An example of this is Warcraft 2. Both the GameCube and PSP have more than enough power to decode MP3 or a similar format. MP3 decoding is virtually free on the modern CPUs in the GameCube (PowerPC derivitive IIRC) and the PSP. You want to put 400MB aside for audio, fine. That's enough for 7 hours of music, which is more than virtually all games on the market. I can't name one with that much composed music, that's for sure.
Now the consoles' built-in wavetable synths use some form of ADPCM (4 bits per mono sample with predictive filtering) as a native waveform encoding, giving only 2.5 hours at 44100 Hz stereo (352 kbps).
I think you are confusing your terms. Wavetable synthesis is a method for producing music from notational data such as MIDI. Most modern devices use wavetable synthesis for their synthesized audio.
Compare this to the typical play time of an RPG. Modern RPGs use prerecorded compressed audio, full stop. All of Square Enix's RPGs use compressed audio, mostly by noted composer Nobuo Uematsu. Interestingly, Uematsu has been composing Square's music since the 4-channel synth (Squarewave I think) to the wavetable synth on the SNES and PSX to the later use of compressed audio.
Or were you planning on using Vorbis decoded in software, which eats a lot of CPU time compared to ADPCM decoded in hardware? Even then you get only three times that. No, Vorbis is overkill. MP3 will do, or even simpler compressed formats. 3x that is more than enough, since you grossly overstate how much music games have by at least a factor of 10.
What is the budget of a Square Enix RPG? The budget is irrelevant, the music budget is what is important and it is nowhere close to 2 million dollars. Square enix games don't have anywhere close to 42 hours of music. Final Fantasy 7 has something like 2 or 3 hours of music, and even the latest Final Fantasy X has a similar amount. We're talking about less than 200MB of music for a Squaresoft game that is shipped on a DVD.
You seem to be thinking as if it were still 1990. We've progressed a bit in the past 15 years.
You do realize that pretty much every game made in the last 5 years has featured compressed or CD audio?
If the game ships on DVD, it has roughly 9.4GB to work with. Put a quarter of the DVD towards compressed audio, that is 42 hours of compressed audio. At the cost listed in the article, let's say $800 per minute, that much music would cost over 2 million dollars.
So as you can see, storage space is NOT a limitation. As much music as you can produce for a game, you can store. I can't speak for the DS or PSP, handhelds may still use enduser wavetable synth. But for PC and console, they are all prerecorded now and have been for years.
I believe there will be a small to medium community porting linux drivers to OSX to allow people to use unsupported hardware, both in their official Apple boxes and their whiteboxes.
And more people than just John Carmack made DooM 3. However, like Carmack, Tim Sweeney is "the engine guy" who does the graphical engine. I'm not certain if he truely does the entire graphics engine himself, though it is certainly possible.
You misunderstand. I refer to the composer purchasing a synthesizer and samples, and using that to produce a score. The audio in the games is simply compressed audio, or a soundtrack for a TV show.
The cost savings I was referring to was the cost of the composer purchasing a professional synthesizer and professional samples, as opposed to repeatedly hiring an orchestra to record the performances. The cost of high-end hardware may well be more than the cost of the orchestra for a game, but you can use the hardware to produce each piece whereas you have to hire the orchestra each time.
I was not referring to end-user synthesizers. They are not up to the level of quality of which I described.
And UE3 is done by Tim Sweeny fulltime. 3 developers working in their spare time should be roughly equal to one full time developer :)
From what I saw their engine is similar to Unreal Engine 3, but a lot less polished. It doesn't look nearly as impressive as UE3.
It will take more than specular highlighting on a normal mapped model to dethrone UE3, since UE3 already does it, and better.
I've found that the higher quality synthesizers with the high-end samples are virtually indistinguishable from a real orchestral performance, as far as soundtracks go at least. The cost of entry for quality hardware and samples might be high, but you can re-use it for no additional cost, unlike a real orchestra.
I understand that this is why several television composers use their own synths instead of a real orchestra. I believe the music for StarGate is done on synth rather than real orchestras.
Is there a difference between live and good synth? Yes. Is it big enough to matter in most circumstances where there is a lower budget? No, the average person can't tell (and doesn't care about) the difference.
Don't get me wrong, I love a good orchestral rendition of a videogame soundtrack.
I'd rather have seen them earmark the money for a spam-fighting foundation.
Think about it, create a mostly financially independant child company or foundation, or even division, and give them the $7 million. That entity then uses the $7 million to sue other spammers. It then gets more money from other spammers, and uses that to fight additional spammers.
This way, Microsoft could have a spam-fighting operation going costing them nothing, since it would make it's own money to continue operating. They'd also get a lot of good press about it.
Also there is something poetic about using spammers' money to sue other spammers.
Problem is if a user pays for software they have a reasonable expectation of support for that software. Not tech support, but general support. Software updates are considered support, for example.
No driver support, no support in games.
There are adapters that let you use mice and keyboards to control any game by emulating a regular controller. They also let you use PS/2 (the connector type) keyboards and mice with the PS2 (the console).
So we're back to the original problem, Microsoft will only allow their own super expensive keyboards to work with the 360.
Very unfortunate if they did (try to charge $200 for a keyboard).
I am considering buying an xbox 360. The decision hinges on the ability to use it with my unused 19" CRT monitor. It's in excellent condition, but a laptop means I don't use it anymore.
The 360 is supposed to have VGA out, so that's that.
But of course, I'm a PC gamer. I was hoping that I'd be able to use a mouse and keyboard to control FPS games on the 360. After all, the 360 has USB ports, so it's theoretically possible if there are standard keyboard/mouse drivers.
But if devices need to be licensed to work, I can forget about this, because the mouse and keyboard are obviously not licensed.
And so the situation you describe could very well happen, although we might end up in an even more rediculous situation; we made need an adapter to hook up a keyboard and mouse to the 360, like we do for the current xbox and PS2. This time around, though, the adapters and the keyboards plugged into them will have the same type of connector. So stupid!
I meant the pennies as parts of a packet (Individual bits even) and the random number of pennies as a whole packet.
The same thing happens in the wired as wireless world; two people transmitting at the same time causes an overlapping signal like two people talking at the same time, and nobody can understand what is being said by either party.
Wireless networks do incorporate several techniques to avoid or even prevent collisions, but that just stops them from happening; you still can't have more than one transmission at a time.
Is the premise the same in the air as it is on a wire? Do signals on the same frequency really interfere with each other that much?
Yes.
Think about it like this, you and your friend are standing in front of a bowl. You have a bunch of pennies. You randomly let some fall out of your hand into the bowl. Your friend writes down how many pennies are in the bowl, 5. That is how many pennies you dropped. You clean out the bowl.
Next your friend does the same thing. He drops 7. You write that down and take them out.
That is sort of how wireless works.
Now, imagine if you were both to both drop the random number of pennies into the bowl at the same time. There are now 12 pennies in the bowl. Who dropped how many pennies? It is impossible to know! All you know is there are 12 pennies, not who put which ones in.
That is called a collision, and when it happens in the wireless world, you would clear out the pennies, wait a random amount of time, then dump more pennies back in hoping your buddy waited a different amount of time than you. Usually this works, if not, try again.
So, you see, when you broadcast a wireless signal, you are putting that signal into the air (Well, not really the air, but you understand). Once it is out there it is free to mix and mingle with any other signal out there. You must simply hope that while you are sending your signal (packet), nobody else interrupts you, or else a collision will occur.
I was referring to the Google Maps images, which google hosts themselves.
But in your example, Google still hosts the thumbnails.