All good points, but I think PC video has suffered different problems from what could be implemented over a wireless network:
#1 PC video is too difficult to setup and configure. There is little chance of getting your grandmother to get a video session up on her PC. Wireless devices will be preconfigured and nothing to plug-in. Airwaves go anywhere.
#2 PC video uses internet for data transmission. This lacks QOS and quality is very variable. Wireless networks will go over dedicated lines and have a minimum QOS or the call won't go through. You will have to pay for it, though.
#3 PC video is too big of a platform. How can you ensure quality when a user might be using a super shitty camera over a modem with a 486? Wireless devices will have a fixed platform so you can always be sure the other side can receive and decode the signal.
#4 PC video only allows people to share "bedroom" experiences. Wireless will let you share any aspect of your life.. take it to a wedding, a bar, etc. This makes the devices very marketable.
Why it may be a while, I think there will be an inevitable convergence of video cameras, digital camera, cell phones, and pdas into one device.
I think your numbers on gay males in SF are very low. This is the gay capitol of the world, I would guess that it's somewhere between 20-30% of the males here are gay. And it tends to be the attractive guys who are gay, but the ugly women who are dykes... so it's a good deal.
What's more is this town is so social that it's easy to meet tons of women. Other cities might have higher ratios, but that doesn't mean squat if you can't meet anyone. I have met a few cuties in Mexico City so I will agree with you there, but the communication barrier was too much. Some other countries really adore foriegners, no matter how you look. mostly non-english speaking countries though. Mexico City is cheap because the cost of living is subsidized by stealing parts from your car!:) I never felt very safe there, and I'm not one to worry.
The city with the highest single female/male ratio is New York (in the US).
Having lived in San Jose (the self-proclaimed 'Silicon Valley') and San Francisco, I can say they are totally different.
SF is compact with excellent public transport, a large number of people here are 20-30. Women are everywhere and it's common for me to have 2-3 different dates a week. Lots of gay guys makes the odds better for straight guys.:) Night life is incredible, dozens of very happening dance clubs, bars on every block. The big clubs stop at 5am or 7am, and clubs like the endup open at 7am so you can keep going for the whole weekend if you are immortal. Fashion is more up-scale. Young people here don't wear collars, striped shirts, sport coats, or ties (i.e. no NY style). Lots of leather skirts and 6" high heals for women. Both guys and gals wear pretty tight clothing so you better get in shape. The work scene is composed of a lot more smaller startups. You run into a lot of people. Cell phone coverage is excellent - I don't have a home or work phone and no one ever knows the difference. SF is more expensive than SJ for housing, but it's so worth it.
SJ is more spread out. Almost no public transport. Most people are 30+. Perfect weather. A good place to raise a family, but a shitty place to be 20 something and single. There are a few nightclubs, but not compared to SF. In SJ everything shuts down at 2am and the only place to go is home. SJ has a lot of big mega-corps where you aren't very likely to meet someone new. Despite being "Silicon Valley" I couldn't DSL or cable modem access where I lived. Cell phone coverage is sporadic. Parking isn't that big of a problem, because the city is so spread out. Women? Pretty hard to find. I was luckily to have 1 new date a month there. Fashion is more "business-like" during the day. I.e. ties and suits, and at night. Fairly conservative stuff compared to SF. Lot more porches, Z3s, etc. It's a yuppie town.
I current have At&T's digital cable. I don't have a clue what is going on behind the scenes, but from the macroblock artifacts on the screen it looks like it uses MPEG2 to transmit channels. Could I use this card to watch channels on a PC? If so, how does the cable company control what channels you can/cannot watch? i.e. We have to pay extra for all the movie channels. Are they broadcast in an encrypted form, or is there clear-text message in each channel telling each box what it can view? Can someone shed some light on this, or point to a FAQ?
---
The article mentions the system runs linux under it all. What is the point? An OS is supposed to provide:
multi-process management : don't need it in a game, simple threads work fine.
filesystem : How hard is it to read ISO9660?
virtual memory support : opps, can't have that without a HD. Even the X-box with a HD is not going to have VM.
device drivers : The system is supposed to be fixed. That's the beauty of console games, the hardware doesn't change. Why do you need loadable drivers?
so in conclusion, I think linux if used, will be truely bastardized to the point it is no longer linux and mearly an advertising bullet point.
This article is a total sham. First off, the PS2 doesn't store any of it's polygons in VRAM. VRAM is used solely for textures, frame buffers, and color lookup tables. There is no limit to the number of polys the PS2 can draw because they are drawn as they come in off the bus. There is a theoretical limit on how many it can draw in 1/60th of a second but this is limited by the bus speeds not by VRAM, and for the most part, not the GS.
The dreamcast, however, uses a different rendering technique that requires all polygon data be stored in VRAM because it does post-processing for sorting without using a z-buffer. The advantage here is that the dreamcast can render double-size and get free anti-aliasing. But this means the dreamcast has a hard limit that no amount of assembly can get you more polys drawn. There is a trick you can do on the PS2 to get anti-aliasing but you have to give up almost all of your VRAM. For this reason most PS2 games don't have anti-aliasing. The PS2 does provide hardware accelerated anti-aliased drawing, but for anti-aliased drawing you have to have pre-sorted polygons. And all PS2 games (I know of) use the Z-buffer for sorting so they disable anti-aliased drawing. In short, the anti-aliased drawing features on the PS2 are totally useless.
That is the big difference between PS2 and dreamcast. So the dreamcast actually looks better (anti-aliasing) but the PS2 is capable of drawing much much more poly data.
The problem is that game machines have drastically different capabilities and features. Emulation doesn't help for developing new games because they are expected to take full advantage of the hardware. I could see it for a puzzle game or a re-release of an old classic.
The Amiga 2k/Tao Elate system offers a better solution than emulation - which is run-time code translation and hardware abstraction to a certain degree... but they don't abstract a lot of things. Writing a high-end PS2 game requires learning 5-6 different assembly languages (EE-core, VU macro mode, VU micro mode, EE-multimedia instructions, VIF, GS). There really is no way to abstract these without diluting the power of the system.
You can write a game to the "lowest common denominator" but then you end up with crap. EA used to have this policy of "4 wide, and 4 deep" meaning every title would run on 4 systems and be translated into 4 languages. They gave up on this when they found out it was basically impossible given the demands of the customers and the time available to the developers.
My philosophy now is to pick a fixed target (i.e. console) and develop a game for it not worrying about portability or what might have to be done later. Then if it sells, bastardize the product and rip it to shreds to port it to other systems.
I wonder if this could lead to the first home solid printers. All the other systems require expensive parts & chemicals to create a model. Water is about the cheapest thing you can buy. A refregeration unit with 3 motors, a microcontroller, and some tubing - $400 or so for a home unit.
If nothing else, you could have fanastic looking ice scupltures for every meal...:) Heck, you could make ice-plates, forks, knives, and spoons. Not that they would be good for real use, but at least doing the dishes can be reduced to placing them in the sink and waiting a few hours. A lazy man's dream!
I just use google's "I'm feeling lucky" for everything... Then you don't have to remember a TLD and you don't end up going to a pron site when you meant to go to the real white house. I think there were unlimited TLDs then people would be forced to use a search engine for everything and names would have less value. Still there is so much marketing and demand out there for ".com" that it will likely always remain a tight space.
The title is "The 10 Most Important People of the Decade" Decade being 10 years. I think RMS's major contributions were made more than 10 years ago. Sure they carry forward into this decade, but if you use that logic why not put Turing or Shockley on there?
This will not work unless the card itself is "closed" and cannot be emulated. Think VMWare. It emulates at the hardware level, and it's not too difficult to get linux to save the audio output while VMWare runs. This is all digital and no driver or OS can ever stop this unless the hardware functionality is kept secret and cannot be emulated.
What if the whole system is running under VMWare or some other hardware emulator? Then the signed driver doesn't know the difference. Does that mean all sound cards with signed drivers are required to be "closed." This would mean they can't release technical specs on the card and the OS community couldn't make linux drivers.
It seems like a mute point for a while since I don't see record companies forgetting about Redbook audio as long as people are still buying CDs.
I would think the number of people willing to buy a license to use RSA is dwindling down to almost no one as the expiration date rapidly approaches. This way they get a little bit of free press and hopefully a positive spin when the patent does expire.
I had my article edited by two of the developers at the Tao Group, one of them being the lead programmer. I hope they know their own system better than whatever random people you are talking about.
One issue I haven't seen discussed is that the Elate/Amiga virtual machine always pretends like the machine is little endian (Intel-style). This is accomplished by using some trickery for byte and short pointers. On big endian systems, before a load instruction is executed (for 8 & 16 bit fetches) the pointer values are xor'ed with a mask to make it appear like the cpu is little endian.
Because of this, the result is a larger executable and slower execution times for big endian system (i.e. pretty much any non-intel system). I talked to Chris at Tao about this and he says most of the time is hidden in the pipeline. But I say you are always going to have negative effects. Your instruction cache will get filled faster because of more instructions and in the case where the pipeline doesn't stall on a load you can use that extra slot to do something else. My question is, how much effect does this have? I have yet to see anyone release perfomance numbers Java or otherwise for Amiga on a big endian system. Or maybe no one cares about non-intel anymore??
"Some eyeballa' scribbles " Accordin' t' da damn German Heise Newsticka' Linux
be da damn top serva' 4.de-domains bot' in terms o' I-P addresses an' domain
names hosted. Da survey carried out by da company Iku Netzwerklösungen
surveyed all o' mos' 2.5 million connected.de-domains distributed ova' 205.540
I-P addresses wit' da damn port scanna' nmap. 44 puh'cent o' I-P addresses
surveyed wuz hosted by Linux, 30 puh'cent by Windows. In terms o' domain names
hosted Linux has some even greata' laid, wit' ova' 1.1 million domaines.
Solaris follows wit' about 850,000 names, o' which 180,000 be hosted on plum
deuce Solaris boxes belongin' t' Germany's biggest webspace provida' Strato.
Windows follows in 3rd place wit' plum 10%. " Check out da damn fish if yo' ass
duzn't jive deutsch. Sheeeiit. "
I think C# is closer to Amiga/Elate than Java. Java was designed with security as a top priority. i.e. being able to run in a sandbox. This imposed a number of limitations that effect java's speed performance. C# allows you to write "unsafe" code if you like, whereas Java requires you to load some external C library. By unsafe, I mean doing unsafe typecast and accessing memory you probably shouldn't.
Also C# never runs under emulation or interpretation - it is always compiled to native code before it is run. This is similar to how Elate works.
For comparisons based on my limited knowledge see here:
I have messed with the Amiga2k SDK and also have been talking to the guys at the Tao Group (the real authors of Amiga2k). Here is an article I wrote comparing it with MS's.net platform. This article assumes you have some knowedge of both system so read the/. article first.
Re:A solution for me (In New Brunswick)
on
More Tivo Hacking
·
· Score: 1
How would you relate the resolution and quality of the ATI encoder? Do you have a general estimate of space per secound on the highest quality setting?
Slashdot Mirror
All good points, but I think PC video has suffered different problems from what could be implemented over a wireless network:
#1 PC video is too difficult to setup and configure. There is little chance of getting your grandmother to get a video session up on her PC. Wireless devices will be preconfigured and nothing to plug-in. Airwaves go anywhere.
#2 PC video uses internet for data transmission. This lacks QOS and quality is very variable. Wireless networks will go over dedicated lines and have a minimum QOS or the call won't go through. You will have to pay for it, though.
#3 PC video is too big of a platform. How can you ensure quality when a user might be using a super shitty camera over a modem with a 486? Wireless devices will have a fixed platform so you can always be sure the other side can receive and decode the signal.
#4 PC video only allows people to share "bedroom" experiences. Wireless will let you share any aspect of your life.. take it to a wedding, a bar, etc. This makes the devices very marketable.
Why it may be a while, I think there will be an inevitable convergence of video cameras, digital camera, cell phones, and pdas into one device.
I think your numbers on gay males in SF are very low. This is the gay capitol of the world, I would guess that it's somewhere between 20-30% of the males here are gay. And it tends to be the attractive guys who are gay, but the ugly women who are dykes... so it's a good deal.
:) I never felt very safe there, and I'm not one to worry.
What's more is this town is so social that it's easy to meet tons of women. Other cities might have higher ratios, but that doesn't mean squat if you can't meet anyone. I have met a few cuties in Mexico City so I will agree with you there, but the communication barrier was too much. Some other countries really adore foriegners, no matter how you look. mostly non-english speaking countries though. Mexico City is cheap because the cost of living is subsidized by stealing parts from your car!
The city with the highest single female/male ratio is New York (in the US).
--
Having lived in San Jose (the self-proclaimed 'Silicon Valley') and San Francisco, I can say they are totally different.
:) Night life is incredible, dozens of very happening dance clubs, bars on every block. The big clubs stop at 5am or 7am, and clubs like the endup open at 7am so you can keep going for the whole weekend if you are immortal. Fashion is more up-scale. Young people here don't wear collars, striped shirts, sport coats, or ties (i.e. no NY style). Lots of leather skirts and 6" high heals for women. Both guys and gals wear pretty tight clothing so you better get in shape. The work scene is composed of a lot more smaller startups. You run into a lot of people. Cell phone coverage is excellent - I don't have a home or work phone and no one ever knows the difference. SF is more expensive than SJ for housing, but it's so worth it.
SF is compact with excellent public transport, a large number of people here are 20-30. Women are everywhere and it's common for me to have 2-3 different dates a week. Lots of gay guys makes the odds better for straight guys.
SJ is more spread out. Almost no public transport. Most people are 30+. Perfect weather. A good place to raise a family, but a shitty place to be 20 something and single. There are a few nightclubs, but not compared to SF. In SJ everything shuts down at 2am and the only place to go is home. SJ has a lot of big mega-corps where you aren't very likely to meet someone new. Despite being "Silicon Valley" I couldn't DSL or cable modem access where I lived. Cell phone coverage is sporadic. Parking isn't that big of a problem, because the city is so spread out. Women? Pretty hard to find. I was luckily to have 1 new date a month there. Fashion is more "business-like" during the day. I.e. ties and suits, and at night. Fairly conservative stuff compared to SF. Lot more porches, Z3s, etc. It's a yuppie town.
---
Forced voiting is already a possibility with absentee ballots, right? How is internet voting going to change anything?
I current have At&T's digital cable. I don't have a clue what is going on behind the scenes, but from the macroblock artifacts on the screen it looks like it uses MPEG2 to transmit channels. Could I use this card to watch channels on a PC? If so, how does the cable company control what channels you can/cannot watch? i.e. We have to pay extra for all the movie channels. Are they broadcast in an encrypted form, or is there clear-text message in each channel telling each box what it can view? Can someone shed some light on this, or point to a FAQ?
---
...but many programs compiled for the Amiga/Tao platform used gcc. How do you think all the unix utilities exsist for amiga?
Also I have heard wind that metroworx is doing a port of their compiler to Tao.
Think most people want to program in anything called an assembly language? not I.
--
doh. It's one click away.
p arison.html
http://www.twingalaxies.com/downloads/indrema_com
Indrema vs Xbox vs PS2.
There is no mention of the PS2 in the article and there certainly is no "verus" comparisions with other systems.
The article mentions the system runs linux under it all. What is the point? An OS is supposed to provide:
multi-process management : don't need it in a game, simple threads work fine.
filesystem : How hard is it to read ISO9660?
virtual memory support : opps, can't have that without a HD. Even the X-box with a HD is not going to have VM.
device drivers : The system is supposed to be fixed. That's the beauty of console games, the hardware doesn't change. Why do you need loadable drivers?
so in conclusion, I think linux if used, will be truely bastardized to the point it is no longer linux and mearly an advertising bullet point.
--
This article is a total sham. First off, the PS2 doesn't store any of it's polygons in VRAM. VRAM is used solely for textures, frame buffers, and color lookup tables. There is no limit to the number of polys the PS2 can draw because they are drawn as they come in off the bus. There is a theoretical limit on how many it can draw in 1/60th of a second but this is limited by the bus speeds not by VRAM, and for the most part, not the GS.
The dreamcast, however, uses a different rendering technique that requires all polygon data be stored in VRAM because it does post-processing for sorting without using a z-buffer. The advantage here is that the dreamcast can render double-size and get free anti-aliasing. But this means the dreamcast has a hard limit that no amount of assembly can get you more polys drawn. There is a trick you can do on the PS2 to get anti-aliasing but you have to give up almost all of your VRAM. For this reason most PS2 games don't have anti-aliasing. The PS2 does provide hardware accelerated anti-aliased drawing, but for anti-aliased drawing you have to have pre-sorted polygons. And all PS2 games (I know of) use the Z-buffer for sorting so they disable anti-aliased drawing. In short, the anti-aliased drawing features on the PS2 are totally useless.
That is the big difference between PS2 and dreamcast. So the dreamcast actually looks better (anti-aliasing) but the PS2 is capable of drawing much much more poly data.
I was wondering if anyone can verify the hornet story? Is smh.com.au a reliable paper?
It sounds really fascinating! I wonder how much hornet juice would cost if they started selling it?
You mean like x86? Sure it's not ideal, but how much more standardized can you get?
The problem is that game machines have drastically different capabilities and features. Emulation doesn't help for developing new games because they are expected to take full advantage of the hardware. I could see it for a puzzle game or a re-release of an old classic.
The Amiga 2k/Tao Elate system offers a better solution than emulation - which is run-time code translation and hardware abstraction to a certain degree... but they don't abstract a lot of things. Writing a high-end PS2 game requires learning 5-6 different assembly languages (EE-core, VU macro mode, VU micro mode, EE-multimedia instructions, VIF, GS). There really is no way to abstract these without diluting the power of the system.
You can write a game to the "lowest common denominator" but then you end up with crap. EA used to have this policy of "4 wide, and 4 deep" meaning every title would run on 4 systems and be translated into 4 languages. They gave up on this when they found out it was basically impossible given the demands of the customers and the time available to the developers.
My philosophy now is to pick a fixed target (i.e. console) and develop a game for it not worrying about portability or what might have to be done later. Then if it sells, bastardize the product and rip it to shreds to port it to other systems.
I wonder if this could lead to the first home solid printers. All the other systems require expensive parts & chemicals to create a model. Water is about the cheapest thing you can buy. A refregeration unit with 3 motors, a microcontroller, and some tubing - $400 or so for a home unit.
:) Heck, you could make ice-plates, forks, knives, and spoons. Not that they would be good for real use, but at least doing the dishes can be reduced to placing them in the sink and waiting a few hours. A lazy man's dream!
If nothing else, you could have fanastic looking ice scupltures for every meal...
I just use google's "I'm feeling lucky" for everything... Then you don't have to remember a TLD and you don't end up going to a pron site when you meant to go to the real white house. I think there were unlimited TLDs then people would be forced to use a search engine for everything and names would have less value. Still there is so much marketing and demand out there for ".com" that it will likely always remain a tight space.
The title is "The 10 Most Important People of the Decade" Decade being 10 years. I think RMS's major contributions were made more than 10 years ago. Sure they carry forward into this decade, but if you use that logic why not put Turing or Shockley on there?
This will not work unless the card itself is "closed" and cannot be emulated. Think VMWare. It emulates at the hardware level, and it's not too difficult to get linux to save the audio output while VMWare runs. This is all digital and no driver or OS can ever stop this unless the hardware functionality is kept secret and cannot be emulated.
What if the whole system is running under VMWare or some other hardware emulator? Then the signed driver doesn't know the difference. Does that mean all sound cards with signed drivers are required to be "closed." This would mean they can't release technical specs on the card and the OS community couldn't make linux drivers.
It seems like a mute point for a while since I don't see record companies forgetting about Redbook audio as long as people are still buying CDs.
I would think the number of people willing to buy a license to use RSA is dwindling down to almost no one as the expiration date rapidly approaches. This way they get a little bit of free press and hopefully a positive spin when the patent does expire.
I had my article edited by two of the developers at the Tao Group, one of them being the lead programmer. I hope they know their own system better than whatever random people you are talking about.
One issue I haven't seen discussed is that the Elate/Amiga virtual machine always pretends like the machine is little endian (Intel-style). This is accomplished by using some trickery for byte and short pointers. On big endian systems, before a load instruction is executed (for 8 & 16 bit fetches) the pointer values are xor'ed with a mask to make it appear like the cpu is little endian.
Because of this, the result is a larger executable and slower execution times for big endian system (i.e. pretty much any non-intel system). I talked to Chris at Tao about this and he says most of the time is hidden in the pipeline. But I say you are always going to have negative effects. Your instruction cache will get filled faster because of more instructions and in the case where the pipeline doesn't stall on a load you can use that extra slot to do something else. My question is, how much effect does this have? I have yet to see anyone release perfomance numbers Java or otherwise for Amiga on a big endian system. Or maybe no one cares about non-intel anymore??
A have a little more info on this at my homepage.
--
Translation for us non-english speakers:
.de-domains bot' in terms o' I-P addresses an' domain
.de-domains distributed ova' 205.540
"Some eyeballa' scribbles " Accordin' t' da damn German Heise Newsticka' Linux
be da damn top serva' 4
names hosted. Da survey carried out by da company Iku Netzwerklösungen
surveyed all o' mos' 2.5 million connected
I-P addresses wit' da damn port scanna' nmap. 44 puh'cent o' I-P addresses
surveyed wuz hosted by Linux, 30 puh'cent by Windows. In terms o' domain names
hosted Linux has some even greata' laid, wit' ova' 1.1 million domaines.
Solaris follows wit' about 850,000 names, o' which 180,000 be hosted on plum
deuce Solaris boxes belongin' t' Germany's biggest webspace provida' Strato.
Windows follows in 3rd place wit' plum 10%. " Check out da damn fish if yo' ass
duzn't jive deutsch. Sheeeiit. "
I think C# is closer to Amiga/Elate than Java. Java was designed with security as a top priority. i.e. being able to run in a sandbox. This imposed a number of limitations that effect java's speed performance. C# allows you to write "unsafe" code if you like, whereas Java requires you to load some external C library. By unsafe, I mean doing unsafe typecast and accessing memory you probably shouldn't.
.NET
Also C# never runs under emulation or interpretation - it is always compiled to native code before it is run. This is similar to how Elate works.
For comparisons based on my limited knowledge see here:
Elate versus
I discuss the IL assembly that c# generates more than the language c# itself.
--
I have messed with the Amiga2k SDK and also have been talking to the guys at the Tao Group (the real authors of Amiga2k). Here is an article I wrote comparing it with MS's .net platform. This article assumes you have some knowedge of both system so read the /. article first.
http://jonathanclark.com/diary/amiga/
--
How would you relate the resolution and quality of the ATI encoder? Do you have a general estimate of space per secound on the highest quality setting?
Thanks,