When you have a DVD, you're not dependent on the whim of a company. Consider shows like The Family Guy [fox.com] or Futurama [fox.com] where Fox never gave them a fair chance, then pulled the plug. They treated these shows like shit the first time; what possible reason do I have to believe that they're be treated any better "on demand?"
So... you can buy a dvd. And, what... watch it 35 times because you like that eposide?
What happens when some soulless bean counter decides that since I'm the only one who wants to watch Seriously Dude, Where's My Car? [imdb.com], they should just save the server space and dump it?
And you're going to get that now, at your local video store? Just try getting a copy of "Attack of the Killer Tomatoes". An online "video store" would need a small amount of very cheap disk space to store even tripe such as the above. Done right, you'd find VOD to have an incredible selection
Finally, why should I keep paying for the content through a subscription or a download fee each time? Compare the price of DVDs with rentals and pay-per-view -- if I think I might watch it three times in the rest of my life (or I might want to loan it to a friend) why not buy it outright for the extra ten bucks?
Why would you want to drive 20 minutes (round trip) to stand in a long line on Friday night, on the chance that they still have the movie you want in stock, only to make another 20 minute drive 2 days later to return the DVD?
VOD has serious advantages for the video renter. I see a market for VOD that's immense. I hate renting DVD/VHS, I hate standing in line, I hate the limited selection (the one you want to watch is always all rented out) and VOD answers all these nicely.
To do it, you'd have to...
1) Put together a central cluster of movie servers,
2) Put in the right number of caching movie servers at supporting ISPs,
3) Upgrade local Internet service to ~ 3-6 Mbit to support a reasonable, high quality connection. I see no reason to offer this speeds to anything but the caching movie servers mentioned in 2 above.
We've had the technology for some time. (Remember Film88 or it's predecessor, movie88.com?
I use vonage business VoIP services. I have calls drop and poor sound quality as is, now if I brought the unpredictability of Wi-Fi connections into play, it would only get worse.
I've been using Vonage for some time now, and am generally very happy with it.
Don't use Vonage if your 'net connection is running over ~ 50%. You'll start to get skips and stuttering in your connection.
But, for your small business with a few phone lines and an xDSL line, it's a godsend. I use it myself, and love the unlimited minutes for working with clients. The lag is comparable to a cell phone - sometimes a little awkward, but never a problem.
On my 1.5M/384K ADSL, I've only had a problem when downloading stuff at over ~100 KB/s or uploading over ~ 20 KB/s. I have all my "heavy lifting" stuff (backups, etc.) running via cron at night, so it's just not a problem.
I surf, stream MP3s, and talk on Vonage on the same line with no issues.
I think, though, with IPv6's QOS component, we might see VOIP jack up another notch.
The company with the most to gain from this (with its unique cash reserve - Microsoft) is the company most in opposition...
Yes, I said it. I'll say it again. Microsoft could gain *alot* from this movement.
With their resources, they are the ones that could easily afford a true source-code audit the likes of which the BSDs are only beginning to approach.
They could build an operating system that fully, completely, and truly matches the concept of "secure by default" and they have the resources, manpower, and ability to do so.
But, instead, they oppose it. Building a secure system is against corporate culture, so they won't do it.
It's been demonstrated that you can predict, to a high degree of likelyhood, when a military strike is about to happen by counting how many pizzas are delivered to the Pentagon.
It's somewhat like the before-mentioned leaky abstraction concept, but applied to information.
It's going to get alot worse long before it gets better. Those who believe that true privacy is possible in the future are delusional.
For a well though-out article on the subject, try reading this Wired article that
BT is an amazingly powerful bit of technology. To see it at work try torrentse.cx. Its main disadvantage?!?
It has to be handled thru a plugin. Imagine the savings if this HTML worked: <IMG SRC="/very_big_image.jpg.torrent">
Yeah, it works! (Red Hat 9 ISOs so soon were a miracle!)
But the Moz guys need to incorporate Torrent tech directly into the browser! That's serve as a huge wakeup call to IE, and we might see a new feature for the first time in NNN years...
They do not realize that they are taking food out of their future mouths.
What a load of crap.
Among my projects is a class I released to phpbuilder.com's "shared code library". It is a method for web servers to send emails through a remote MTA. (PHP's "mail()" function only works on *nix if you have sendmail installed)
By releasing this library into the public (under the LGPL) I've seen it grow and get better as others have used it, and occasionally, tweaked it to fit their needs.
Why people release software to the public is different for each person. It's really like asking: "Why do people drive on freeways?" or, "Why do people dig with shovels?".
Open source licensing is a tool. Different folks use that tool for different reasons. The point, however, is that we *have* this tool, and isn't it kinda neat?
For $40/mo, I get an IP phone that I can take and plug into any broadband network via DHCP. I get unlimited calls anywhere in the US and Canada, and no other fees.
The voice quality is good, and the price is excellent, and I can take the # anywhere I want to - just plug into broadband, it autoconfigures with DHCP, and in 10 seconds or less, I'm up!
The bells, with all their "X per minute on Wednesdays between 4 and 11PM" bull---t are ripe for a serious change in their business model.
Yeah, I've been on (what is now) SBC DSL for over 3 years. I've always consistently gotten ~ 1.5 Mb download, ~350(ish) Kb uploads.
I've had DNS, Email, Web, NTP, and SSH running on it basically the whole time, with NAT behind it. Actually, this is largely because at first, their DNS and EMail services were intermittent. (at best) I understand that they've gotten quite a bit better.
The only thing is that I have to lie a bit when calling tech support and pretend I'm running 'doze. But, more than once, I've been told that it's not a *problem* running *nix, (or Win 98) there's just no tech support.
Heck, SBC/Yahoo have this "Home Network Kit" they sell, that's essentially a DSL/Router with built-in PPPOE, NAT and DHCP!
Never a problem, never a complaint. And it works 24x7 with very few issues over the past 3 years. Hell, I've had a customer with a T1 have more issues!
I probably average ~ 25-50 GB (yes, BYTES) of transfer a month, most at night when performing off-site backups.
I've been working from the home for 3+ years. Usually good, sometimes bad. Biggest problem is having a place to meet clients - I have an "office" now that I go to 3 or so days/month to meet the occasional local client. (Most of my clients are out of the area - more than half of my income comes from people I've never met)
The rent at my office is cheap, but it looks very nice.
The bad is when my 5 kids get rowdy, the wife is running errands, and I'm in the middle of a nasty, nested/recursive loop and looking for an esoteric bug. (The kind where $var=$value or $var=&$value can make all the difference)
I know my neighbors, and I don't produce any noticable amount of traffic. My professional life exists via my DSL modem.
Do I like what I do? He11 yeah!
But working at home has its ups and downs - despite having 5 kids and a wife around here (and the incessant noise that accompany this many people) I sometimes wish I had other "co-workers". The closest I come are clients that I largely work with via my IP telephone.
Give it a shot - zoning (at least around here) is not an issue, and the licensing takes $50 and an afternoon. The real issue is getting paying customers. With that, the rest is easy!
From the article: If you're typing on a standard QWERTY keyboard, and most of us are, then your keyboard design is over 100 years old (135 years old, to be exact). Can you imagine using a hard drive that was designed a decade ago? Or a processor from two centuries past?
I call bulls--t!
The hard drive of today is a highly refined, miniaturized version of a design that goes back at least to the 1970s.
My ergonomic keyboard was definitely *NOT* designed 135 years ago. Mechanically, it has as much in common with a mechanical typewriter as a lawn mower has in common with a vibrator. (basically, nothing)
Reply or mod... Reply or mod... Augh spit! I'll bite...
the Linux community has to concede the desktop market to Microsoft and move on
PHP-GTK is a byproduct of a "Windows-like" toolkit (GTK) meeting a definitely server-based language. (PHP)
The result is quite impressive. I can use the same codebase for file i/o and communications on the Windows clients as on the Unix server, giving me guaranteed 100% compatability.
This is a natural for Web services and network-based software!
I welcome the improvements with Gnome 2.x, even though it A) will make the toolkit bigger (and thus, somehow, more "bloated") and B) for a while, less stable while the kinks are worked out.
If the gnome guys give up on GTK to become Samba developers, and Gimp becomes a thing of the past, I'd be sorely pissed. But, one of the blessings of open-source is that it's very unlikely to happen. Once open-source codebases become substantial enough, somebody inherits the base and carries on.
Witness the PostgreSQL team over the past few years. It's been turned over several times, and in the process, it's become a much better product! Lately, there's another "pay for service" company that's taken control of the site and codebase, we'll see what they do to it.
Nobody can argue that Linux is giving up on servers - it's just not. 2.4 not as stable as 2.2? What kind of hooey is that? 2.2 sucked at any load average above 2.0 - I've seen 2.4.x systems handle a load average of 35 and more while still being reasonably responsive.
That's not an improvement?
Microsoft has beat down the competition by commoditizing part of the computer industry - the hardware.
Linux takes this one step further, by commoditizing the software, too. Every new developer who picks up Linux is one more person who adds a bit to the overall formula, and that's one more area that Linux is steadily improving in.
If it didn't work this way, Linux wouldn't be the force it now is.
All this techno mumbo-jumbo is just an overly verbose representation of an old saw:
A man with a watch always knows what time it is. A man with two watches is never quite sure...
Re:May as well be the first to say it
on
AOL Sues Spammers
·
· Score: 1
And, might I add, the cost of SPAM filtering isn't cheap.
When you factor in TNEF expansion, antivirus, and Spam Asssassin, you can quickly take a server's normal load average of 0.3 up to 10 or more. (I've seen it!)
To protect a standard ISP customer, even "on the cheap" costs more than you'd think!
Sure, SpamAssassin on your *personal* system doesn't slow it down noticably, but what about when you are processing 10,000, 50,000, or 200,000 emails per day?
Let me assure you, the resources used aren't insignificant. Then you start talking about clustering, RAID, and multiple CPUs.
The essential problem with email is authentication. Solve this, and SPAM goes away.
I'd figure on using DNS-SEC to collapse DNS and SSL certificates into one mechanism (why would the server that tells you where to go - DNS - not then let you know if you really got there?) which would then allow for perfect back-tracing messages to the actual ISP customer.
This provides accountability, and this allows for direct recourse against would-be spammers, and suddenly, the criminal behavior that is SPAM becomes much more costly to the spammer.
Are you honestly saying that all broadband customers should restrict their email addresses to those assigned by their bandwidth providers?
No, that restriction is pretty lame. However, as a service provider, I'd be more than happy to accomodate you for a reasonable annual fee, assuming that you have a static IP.
Basically, the argument boils down to this:
1) You are using your DSL line as a home user, at which point "@verizon.com" is perfectly OK anyway, or
2) You are using your email addy for actual business, at which point the "reasonable annual fee" is insignificant compared to the actual cashflow of any viable business.
I've yet to see an actual answer to these questions:
1) Let's say I have a 500 MB file that gets slashdotted, and 10,000 people want to download it. Normally, that would be 500*10,000 MB or 5,000,000 MB (which is a !!@# of alot of bandwidth) of network capacity. Given an "ideal" scene, what would an expected bandwidth usage be if I ran a BitTorrent tracker and a see file? Could this conceivably be done on a T1, since the clients are (in theory) providing most of the upload bandwidth?
2) How much overhead does Bit Torrent add to connections that aren't all that busy? If 1 guy downloads that 500 MB file, how much more bandwidth would he use because of the BitTorrent protocol stuff?
I've yet to find a decent answer to these two questions anywhere.
Doesn't this sound an awful lot like having a brake pedal that's also the gas pedal in a car? (but with an ethernet port!)
I can just see it now...
"This here is a brake pedal, that also runs the gas! Want to speed up? Push that pedal! Want to slow down? Push that same pedal! Want to speed up or slow down REMOTELY, when you aren't even in the car?!? Just load VNC, and click on the 'PEDAL' button on your screen!"
OOOOH! aaaaahhhhhh!
Some ideas are just too stupid to take seriously. Anybody remember the bar code reader that was supposed to revolutionize reading magazines?
Ok, you don't want to start a flamewar, and you combine "inferior" and "Linux" in the same sentence without a strong negative somewhere in there??!?!?
Actually, you are right. Linux's thread support has traditionally been weak, but mainly because, compared to other *nixes, Linux creates and destroys *processes* so well.
AFAIK, on Linux, the difference between a process and a thread is much less distinct than on competetive OSs.
So, you could say (quite honestly) that threads on Linux is weak because Linux is otherwise so well done....
The basic problem isn't that Apache runs as "userX" or "userY" or even "root", it's that it ONLY runs as user "apache"!
If I have 100 clients using a web server, there's no way for me to protect their stuff from each other. NONE.
It doesn't matter what permissions I apply. I can run PHP in "safe" mode, and apply bandaids to the problem to mitigate this weakness, but it's still there.
Maybe make apache run under xinet.d. (Gee, there goes the "must run as root" problem!) Maybe just have a connection process that connects to an actual daemon for performance reasons.
But Apache should run as the user that owns the site being accessed!
If done right, you should be able to chroot user "client" and have the DocumentRoot be relative within the chrooted file system!
This is a feature of 2.x that is the *only* feature I'm looking forward to. And yet, for some reason, it's on the back burner. It's "unstable", or "in progress". In short, it still sucks.
So we continue to run in an inherently lame-brained environment with security leaks all over the place, with this "unpriveledged user" (typically "nobody") that has more permissions than any other user save root.
If the goal of the terrorists was to strike out against our liberties, to strike out against our freedoms, and our democratic way of life, they have won.
If the goal of the terrorists was to wage war on our way of life, to destroy the very fabric of our political and social structure, they have won.
If we want to fight the terrorists, if we want to win the "war on terrorism", we must cling closely to the umbrellas of freedom, privacy, due process, and democracy.
Until MS separates out the advertising a bit better and stops skewing the top links quite so much to suit its own opinion, people aren't going to use their search site.
Yeah, and the popups get real irritating after a while if you leave your browser on MSN.
Slashdot Mirror
When you have a DVD, you're not dependent on the whim of a company. Consider shows like The Family Guy [fox.com] or Futurama [fox.com] where Fox never gave them a fair chance, then pulled the plug. They treated these shows like shit the first time; what possible reason do I have to believe that they're be treated any better "on demand?"
So... you can buy a dvd. And, what... watch it 35 times because you like that eposide?
What happens when some soulless bean counter decides that since I'm the only one who wants to watch Seriously Dude, Where's My Car? [imdb.com], they should just save the server space and dump it?
And you're going to get that now, at your local video store? Just try getting a copy of "Attack of the Killer Tomatoes". An online "video store" would need a small amount of very cheap disk space to store even tripe such as the above. Done right, you'd find VOD to have an incredible selection
Finally, why should I keep paying for the content through a subscription or a download fee each time? Compare the price of DVDs with rentals and pay-per-view -- if I think I might watch it three times in the rest of my life (or I might want to loan it to a friend) why not buy it outright for the extra ten bucks?
Why would you want to drive 20 minutes (round trip) to stand in a long line on Friday night, on the chance that they still have the movie you want in stock, only to make another 20 minute drive 2 days later to return the DVD?
VOD has serious advantages for the video renter. I see a market for VOD that's immense. I hate renting DVD/VHS, I hate standing in line, I hate the limited selection (the one you want to watch is always all rented out) and VOD answers all these nicely.
To do it, you'd have to...
1) Put together a central cluster of movie servers,
2) Put in the right number of caching movie servers at supporting ISPs,
3) Upgrade local Internet service to ~ 3-6 Mbit to support a reasonable, high quality connection. I see no reason to offer this speeds to anything but the caching movie servers mentioned in 2 above.
We've had the technology for some time. (Remember Film88 or it's predecessor, movie88.com?
I use vonage business VoIP services. I have calls drop and poor sound quality as is, now if I brought the unpredictability of Wi-Fi connections into play, it would only get worse.
I've been using Vonage for some time now, and am generally very happy with it.
Don't use Vonage if your 'net connection is running over ~ 50%. You'll start to get skips and stuttering in your connection.
But, for your small business with a few phone lines and an xDSL line, it's a godsend. I use it myself, and love the unlimited minutes for working with clients. The lag is comparable to a cell phone - sometimes a little awkward, but never a problem.
On my 1.5M/384K ADSL, I've only had a problem when downloading stuff at over ~100 KB/s or uploading over ~ 20 KB/s. I have all my "heavy lifting" stuff (backups, etc.) running via cron at night, so it's just not a problem.
I surf, stream MP3s, and talk on Vonage on the same line with no issues.
I think, though, with IPv6's QOS component, we might see VOIP jack up another notch.
The company with the most to gain from this (with its unique cash reserve - Microsoft) is the company most in opposition...
Yes, I said it. I'll say it again. Microsoft could gain *alot* from this movement.
With their resources, they are the ones that could easily afford a true source-code audit the likes of which the BSDs are only beginning to approach.
They could build an operating system that fully, completely, and truly matches the concept of "secure by default" and they have the resources, manpower, and ability to do so.
But, instead, they oppose it. Building a secure system is against corporate culture, so they won't do it.
Thanks xBSD!
In two words: Information leaks.
It's been demonstrated that you can predict, to a high degree of likelyhood, when a military strike is about to happen by counting how many pizzas are delivered to the Pentagon.
It's somewhat like the before-mentioned leaky abstraction concept, but applied to information.
It's going to get alot worse long before it gets better. Those who believe that true privacy is possible in the future are delusional.
For a well though-out article on the subject, try reading this Wired article that
BT is an amazingly powerful bit of technology. To see it at work try torrentse.cx. Its main disadvantage?!?
It has to be handled thru a plugin. Imagine the savings if this HTML worked: <IMG SRC="/very_big_image.jpg.torrent">
Yeah, it works! (Red Hat 9 ISOs so soon were a miracle!)
But the Moz guys need to incorporate Torrent tech directly into the browser! That's serve as a huge wakeup call to IE, and we might see a new feature for the first time in NNN years...
-Ben
They do not realize that they are taking food out of their future mouths.
What a load of crap.
Among my projects is a class I released to phpbuilder.com's "shared code library". It is a method for web servers to send emails through a remote MTA. (PHP's "mail()" function only works on *nix if you have sendmail installed)
By releasing this library into the public (under the LGPL) I've seen it grow and get better as others have used it, and occasionally, tweaked it to fit their needs.
Why people release software to the public is different for each person. It's really like asking: "Why do people drive on freeways?" or, "Why do people dig with shovels?".
Open source licensing is a tool. Different folks use that tool for different reasons. The point, however, is that we *have* this tool, and isn't it kinda neat?
For $40/mo, I get an IP phone that I can take and plug into any broadband network via DHCP. I get unlimited calls anywhere in the US and Canada, and no other fees.
The voice quality is good, and the price is excellent, and I can take the # anywhere I want to - just plug into broadband, it autoconfigures with DHCP, and in 10 seconds or less, I'm up!
The bells, with all their "X per minute on Wednesdays between 4 and 11PM" bull---t are ripe for a serious change in their business model.
The Linux world in some sense is a lot like the Unix world. There is not much communality.
'Nuff said.
(Yet Another Useless Plug)
Yeah, I've been on (what is now) SBC DSL for over 3 years. I've always consistently gotten ~ 1.5 Mb download, ~350(ish) Kb uploads.
I've had DNS, Email, Web, NTP, and SSH running on it basically the whole time, with NAT behind it. Actually, this is largely because at first, their DNS and EMail services were intermittent. (at best) I understand that they've gotten quite a bit better.
The only thing is that I have to lie a bit when calling tech support and pretend I'm running 'doze. But, more than once, I've been told that it's not a *problem* running *nix, (or Win 98) there's just no tech support.
Heck, SBC/Yahoo have this "Home Network Kit" they sell, that's essentially a DSL/Router with built-in PPPOE, NAT and DHCP!
Never a problem, never a complaint. And it works 24x7 with very few issues over the past 3 years. Hell, I've had a customer with a T1 have more issues!
I probably average ~ 25-50 GB (yes, BYTES) of transfer a month, most at night when performing off-site backups.
-Ben
I've been working from the home for 3+ years. Usually good, sometimes bad. Biggest problem is having a place to meet clients - I have an "office" now that I go to 3 or so days/month to meet the occasional local client. (Most of my clients are out of the area - more than half of my income comes from people I've never met)
The rent at my office is cheap, but it looks very nice.
The bad is when my 5 kids get rowdy, the wife is running errands, and I'm in the middle of a nasty, nested/recursive loop and looking for an esoteric bug. (The kind where $var=$value or $var=&$value can make all the difference)
I know my neighbors, and I don't produce any noticable amount of traffic. My professional life exists via my DSL modem.
Do I like what I do? He11 yeah!
But working at home has its ups and downs - despite having 5 kids and a wife around here (and the incessant noise that accompany this many people) I sometimes wish I had other "co-workers". The closest I come are clients that I largely work with via my IP telephone.
Give it a shot - zoning (at least around here) is not an issue, and the licensing takes $50 and an afternoon. The real issue is getting paying customers. With that, the rest is easy!
If you're typing on a standard QWERTY keyboard, and most of us are, then your keyboard design is over 100 years old (135 years old, to be exact). Can you imagine using a hard drive that was designed a decade ago? Or a processor from two centuries past?
I call bulls--t!
The hard drive of today is a highly refined, miniaturized version of a design that goes back at least to the 1970s.
My ergonomic keyboard was definitely *NOT* designed 135 years ago. Mechanically, it has as much in common with a mechanical typewriter as a lawn mower has in common with a vibrator. (basically, nothing)
Reply or mod... Reply or mod... Augh spit! I'll bite...
the Linux community has to concede the desktop market to Microsoft and move on
PHP-GTK is a byproduct of a "Windows-like" toolkit (GTK) meeting a definitely server-based language. (PHP)
The result is quite impressive. I can use the same codebase for file i/o and communications on the Windows clients as on the Unix server, giving me guaranteed 100% compatability.
This is a natural for Web services and network-based software!
I welcome the improvements with Gnome 2.x, even though it A) will make the toolkit bigger (and thus, somehow, more "bloated") and B) for a while, less stable while the kinks are worked out.
If the gnome guys give up on GTK to become Samba developers, and Gimp becomes a thing of the past, I'd be sorely pissed. But, one of the blessings of open-source is that it's very unlikely to happen. Once open-source codebases become substantial enough, somebody inherits the base and carries on.
Witness the PostgreSQL team over the past few years. It's been turned over several times, and in the process, it's become a much better product! Lately, there's another "pay for service" company that's taken control of the site and codebase, we'll see what they do to it.
Nobody can argue that Linux is giving up on servers - it's just not. 2.4 not as stable as 2.2? What kind of hooey is that? 2.2 sucked at any load average above 2.0 - I've seen 2.4.x systems handle a load average of 35 and more while still being reasonably responsive.
That's not an improvement?
Microsoft has beat down the competition by commoditizing part of the computer industry - the hardware.
Linux takes this one step further, by commoditizing the software, too. Every new developer who picks up Linux is one more person who adds a bit to the overall formula, and that's one more area that Linux is steadily improving in.
If it didn't work this way, Linux wouldn't be the force it now is.
It'll only get stronger.
Viva la Linux!
All this techno mumbo-jumbo is just an overly verbose representation of an old saw:
A man with a watch always knows what time it is. A man with two watches is never quite sure...
All this techno mumbo-jumbo is just an overly verbose representation of an old saw:
A man with a watch always knows what time it is. A man with two watches is never quite sure...
And, might I add, the cost of SPAM filtering isn't cheap.
When you factor in TNEF expansion, antivirus, and Spam Asssassin, you can quickly take a server's normal load average of 0.3 up to 10 or more. (I've seen it!)
To protect a standard ISP customer, even "on the cheap" costs more than you'd think!
Sure, SpamAssassin on your *personal* system doesn't slow it down noticably, but what about when you are processing 10,000, 50,000, or 200,000 emails per day?
Let me assure you, the resources used aren't insignificant. Then you start talking about clustering, RAID, and multiple CPUs.
The essential problem with email is authentication. Solve this, and SPAM goes away.
I'd figure on using DNS-SEC to collapse DNS and SSL certificates into one mechanism (why would the server that tells you where to go - DNS - not then let you know if you really got there?) which would then allow for perfect back-tracing messages to the actual ISP customer.
This provides accountability, and this allows for direct recourse against would-be spammers, and suddenly, the criminal behavior that is SPAM becomes much more costly to the spammer.
Oh well. My own rants...
Are you honestly saying that all broadband customers should restrict their email addresses to those assigned by their bandwidth providers?
No, that restriction is pretty lame. However, as a service provider, I'd be more than happy to accomodate you for a reasonable annual fee, assuming that you have a static IP.
Basically, the argument boils down to this:
1) You are using your DSL line as a home user, at which point "@verizon.com" is perfectly OK anyway, or
2) You are using your email addy for actual business, at which point the "reasonable annual fee" is insignificant compared to the actual cashflow of any viable business.
-Ben
I've yet to see an actual answer to these questions:
1) Let's say I have a 500 MB file that gets slashdotted, and 10,000 people want to download it. Normally, that would be 500*10,000 MB or 5,000,000 MB (which is a !!@# of alot of bandwidth) of network capacity. Given an "ideal" scene, what would an expected bandwidth usage be if I ran a BitTorrent tracker and a see file? Could this conceivably be done on a T1, since the clients are (in theory) providing most of the upload bandwidth?
2) How much overhead does Bit Torrent add to connections that aren't all that busy? If 1 guy downloads that 500 MB file, how much more bandwidth would he use because of the BitTorrent protocol stuff?
I've yet to find a decent answer to these two questions anywhere.
-Ben
Doesn't this sound an awful lot like having a brake pedal that's also the gas pedal in a car? (but with an ethernet port!)
I can just see it now...
"This here is a brake pedal, that also runs the gas! Want to speed up? Push that pedal! Want to slow down? Push that same pedal! Want to speed up or slow down REMOTELY, when you aren't even in the car?!? Just load VNC, and click on the 'PEDAL' button on your screen!"
OOOOH! aaaaahhhhhh!
Some ideas are just too stupid to take seriously. Anybody remember the bar code reader that was supposed to revolutionize reading magazines?
I see a number of comments in this thread.
Anybody who says "it can't be done" is simply wrong. It can obviously be done, in a number of ways, with minimal repurcussions.
I'd almost give a left nut for something like this that actually worked.
So why hasn't it been done?
Ok, you don't want to start a flamewar, and you combine "inferior" and "Linux" in the same sentence without a strong negative somewhere in there??!?!?
Actually, you are right. Linux's thread support has traditionally been weak, but mainly because, compared to other *nixes, Linux creates and destroys *processes* so well.
AFAIK, on Linux, the difference between a process and a thread is much less distinct than on competetive OSs.
So, you could say (quite honestly) that threads on Linux is weak because Linux is otherwise so well done....
-Ben
The basic problem isn't that Apache runs as "userX" or "userY" or even "root", it's that it ONLY runs as user "apache"!
... logging, etc.
If I have 100 clients using a web server, there's no way for me to protect their stuff from each other. NONE.
It doesn't matter what permissions I apply. I can run PHP in "safe" mode, and apply bandaids to the problem to mitigate this weakness, but it's still there.
Maybe make apache run under xinet.d. (Gee, there goes the "must run as root" problem!) Maybe just have a connection process that connects to an actual daemon for performance reasons.
But Apache should run as the user that owns the site being accessed!
Imagine this in your httpd.conf:
<VirtualHost *>
ServerName www.clientsite.com
ServerAlias clientsite.com
DocumentRoot ~client/html
RunAsUser client
</VirtualHost>
If done right, you should be able to chroot user "client" and have the DocumentRoot be relative within the chrooted file system!
This is a feature of 2.x that is the *only* feature I'm looking forward to. And yet, for some reason, it's on the back burner. It's "unstable", or "in progress". In short, it still sucks.
So we continue to run in an inherently lame-brained environment with security leaks all over the place, with this "unpriveledged user" (typically "nobody") that has more permissions than any other user save root.
Ugh.
If the goal of the terrorists was to strike out against our liberties, to strike out against our freedoms, and our democratic way of life, they have won.
If the goal of the terrorists was to wage war on our way of life, to destroy the very fabric of our political and social structure, they have won.
If we want to fight the terrorists, if we want to win the "war on terrorism", we must cling closely to the umbrellas of freedom, privacy, due process, and democracy.
For, if we don't, terrorism has won.
Until MS separates out the advertising a bit better and stops skewing the top links quite so much to suit its own opinion, people aren't going to use their search site.
Yeah, and the popups get real irritating after a while if you leave your browser on MSN.
But, did you notice the ironic link (at the very top, no less) to
"Xbox Linux (software)"?
You administer a network of this size, and you haven't heard of anti-virus software?
Really, everybody knows Mcafee and Norton/Symantec, but F-prot blows these out of the water.
It's current.
It's fast (you can run on a P-200 and still have a usable computer!)
It's cheap. ($2/workstation, $300/server)
It runs on Windows, Linux, BSD, AIX, DOS, etc.
We use it on our Linux mail servers with excellent results as a free service to our clients.
-Ben
Yeah, but you've missed my point.
Why are "new" 100/1000 Mb networks setup with UTP cables, instead of Coax?
I understand quite well that Coax has a much "cleaner" signal than UTP, if only for the simple fact that it's shielded.
So why has UTP "won" over coax, and why is it you don't find 100 Mb NICs that use coax?
Is it price alone?