The airwaves have too little bandwidth available
on
Laptop Pentium IIIs
·
· Score: 2
Unfortunately, the airwaves just don't have enough capacity for everyone to a link directly back home.
A better solution would be short-range base units scattered 200 feet apart in all civilized areas (perhaps on top of all light posts) that have a fibre connection to the internet. Thus lower-power radios can be used, and since there would be only at max 200 notebooks in a 200 foot radius to one of these stations, there would be plenty of bandwidth for all! Then you would just securely VNC into your home computer for all the juicy processing tasks you need to do...
The problem with this is that Java, at best, will only run about 40% the speed of native code. A better idea would be to release the source code to this layout engine, or better yet to have it as some sort of COM component that can be wrapped around by any browser. Because the source code is available, it can be ported to any platform fairly easily.
It would be cool to have 3 or 4 different layout engines out there, and being able to plug whichever one you want into your favorite browser.
If everyone would just follow the standards right to the letter, then it wouldn't matter which layout engine you used...they'd all look the same.
That may be the case, however, if the user is required to have a pentium iii 550 just for a dumb terminal to operate at full speed, why not just run the applications on the client!?
I have huge doubts that this will ever happen. Try running VNC over a LAN to see how slow sending just screen updates over the internet can be. Just imagine the size of servers that will be required for this. Why would I want to pay $50 a month for reasonable bandwidth when I can just add a $7 Cyrix processor to my 'dumb terminal'. Terminal-style computing may have worked back in the old days, but today's users (read non-geeks) demand a hoggish UI that will consume bandwidth like we've never even seen before.
Software companies all seem to have this unattainable vision of charging ever user by the minute for the use of their product. I cannot see how this could ever take off. It made sense back in the old days when memory and cpus were extremely expensive, but with the advent of powerful microcomputers a full blown workstation can be had for only a few more bucks than a dumb terminal.
This sort of 'revolution' will have about the same impact on our lives as the huge 'revolution' of Java did 2 years ago.
It really depends on the song. Some mp3s can be encoded at 128 kbps and are virtually indistinguishable from the original wav. Others, with an extremely wide range of frequencies (ie extremely rich electric guitar distortion with constant crashing cymbals) can sound terrible all the way up to 256 kbps. I have yet to find an song that when encoded at 256 kbps sounds difference.
To hear the mp3 'distorting' of the original sound, you have to have decent speakers/headphones, and you have to know what you are listening for. I find that what gives the truest sound are my $200 Sony headphones (with a frequency response from 20 Hz to 20 000 Hz) and, depending on the sound, you can hear a sloshy sound in the extreme high end, kind of like what you get from a gurbled up cassette tape. It sounds vaguely electronic.
Myth #1: MP3s have a reduced dynamic range compared to CD tracks: This is false, the dynamic range of an mp3 is high, just as much as the original recording.
Myth #2: MP3s are missing the high end Unless encoded by Xing's codec (which sucks, never use it), an mp3 will have virtually identical frequency response all the way into the high end. However, when encoded at lower bitrates, an mp3 will mutate the high end around by adding noise at approximately the same frequencies, thus saving the space where the original sound was. This is a distortion to the original high end, but there is still a high end there.
For most shovelware modern pop music 128kbps is adequate, but when encoding classic recordings, 192kbps and even 256kbps is recommended to guarantee the waveform has not become too distorted from it's original shape. It is similiar to a JPEG image--it looks gross when the compression is turned way up, but when compressed at the right ratio, is indistiguishable from the original image.
Sure, it's cool that Apple is releasing their source for the Darwin Streaming Server, but how useful is this anyway? It should be non-trivial to write a streaming server, espcially since UDP is virtually useless as a delivery protocol these days (losing packets and such). Is there any fully GPLed RTSP servers?
Slashdot Mirror
A better solution would be short-range base units scattered 200 feet apart in all civilized areas (perhaps on top of all light posts) that have a fibre connection to the internet. Thus lower-power radios can be used, and since there would be only at max 200 notebooks in a 200 foot radius to one of these stations, there would be plenty of bandwidth for all! Then you would just securely VNC into your home computer for all the juicy processing tasks you need to do...
It would be cool to have 3 or 4 different layout engines out there, and being able to plug whichever one you want into your favorite browser.
If everyone would just follow the standards right to the letter, then it wouldn't matter which layout engine you used...they'd all look the same.
That may be the case, however, if the user is required to have a pentium iii 550 just for a dumb terminal to operate at full speed, why not just run the applications on the client!?
Software companies all seem to have this unattainable vision of charging ever user by the minute for the use of their product. I cannot see how this could ever take off. It made sense back in the old days when memory and cpus were extremely expensive, but with the advent of powerful microcomputers a full blown workstation can be had for only a few more bucks than a dumb terminal.
This sort of 'revolution' will have about the same impact on our lives as the huge 'revolution' of Java did 2 years ago.
To hear the mp3 'distorting' of the original sound, you have to have decent speakers/headphones, and you have to know what you are listening for. I find that what gives the truest sound are my $200 Sony headphones (with a frequency response from 20 Hz to 20 000 Hz) and, depending on the sound, you can hear a sloshy sound in the extreme high end, kind of like what you get from a gurbled up cassette tape. It sounds vaguely electronic.
Myth #1: MP3s have a reduced dynamic range compared to CD tracks:
This is false, the dynamic range of an mp3 is high, just as much as the original recording.
Myth #2: MP3s are missing the high end
Unless encoded by Xing's codec (which sucks, never use it), an mp3 will have virtually identical frequency response all the way into the high end. However, when encoded at lower bitrates, an mp3 will mutate the high end around by adding noise at approximately the same frequencies, thus saving the space where the original sound was. This is a distortion to the original high end, but there is still a high end there.
For most shovelware modern pop music 128kbps is adequate, but when encoding classic recordings, 192kbps and even 256kbps is recommended to guarantee the waveform has not become too distorted from it's original shape. It is similiar to a JPEG image--it looks gross when the compression is turned way up, but when compressed at the right ratio, is indistiguishable from the original image.
Sure, it's cool that Apple is releasing their source for the Darwin Streaming Server, but how useful is this anyway? It should be non-trivial to write a streaming server, espcially since UDP is virtually useless as a delivery protocol these days (losing packets and such). Is there any fully GPLed RTSP servers?