You bring up a good point because I wonder if updates count... correct me if I am wrong but updates require virtually (or exactly) the same size download as the original correct? Updates + variations = skewed statistics perhaps?
(unless you are in the very small formally-proven-systems-written-in-Ada market, or something of that sort) v1.0 of snazzynewprotocol is a bit of a clusterfuck, and is available in only a single implementation, also highly dubious, while the old standbys have been polished considerably and have a number of implementations available...
Careful that we do not open Pandora's box here... (You know exactly what I am talking about, heh)
But on another note your exactly right. This article seems to talk about how protocols "evolved" but this is just as useful as painting a picture of the internet: Time and time again I will see models looking at a picture of the internet "all at once", but without knowing what and why with each individual link, protocol, implementation, etc... this is a complete waste of time.
As you have said in so many words above, what these "researchers" did is a complete waste of time. Maybe they need to do some research on peering 101?
Disclaimer: I have not read the actual paper just a poorly written article linked by Slashdot. Perhaps there is much more to their work; if so, I do apologize.
See the following process comparable to a maddening wikipedia link chain with no surprises:
(1) click on semi-relevant article with a title that seems like it may help to resolve said issue. (2) find that this article has to be written for the likes of your grandmother (completely retarded user) or some weird error code (that uses like 8 digits these days) and is completely irrelevant. (3) Look to the bottom and see See Also: (4)Go to Step (1)
True story. In fact I bet this is really easy for content service providers because they only need to cache perhaps around 100 different help articles that can co-link and co-create a tangled mess from hell.
with what seems like a recent increase in spam and general crap here I'm in a bit of a bit mood with/. and when that happens I tend to mod down rather than up.*
Ah I haven't been here too long myself but I can entirely relate.
modded posters get replies telling them to have a nice day
I usually end with that or cheers... for everyone. Although it is true people are a bit less insulting (in general) to moderators.
I'm not going to stop modding such things down but in future it'll be offtopic instead; I don't think that does anything to karma. Does it?
His comment, like many, is funny. Fortunately funny does not award karma but just makes/. users a tiny bit happier inside (not smarter or more well informed).
Modding funny does not affect karma and some people can choose to filter it if they want to read intelligence instead of a stream of jokes... this is what other people have told me in the past so take it with a grain of salt.
Personally I believe if you have the will to reply to my comment and to moderate good, you sir are ahead of 95%+ of the users here.
Yes. Just moderate then join back in the discussion on the next article. Posting comments is borderline taunting in some cases even if that was not your intention.
SuricouRaven wrote: A well-resourced CDN provides far better and more consistant performance than P2P, and does so while placing less load on the networks. It's big downside is the very considerable cost - the hardware, rack rental, and negociating deals with ISPs. They arn't even doing to talk to any company that doesn't have a few million dollars to bolster their reputation. So the CDN remains the prefered distribution method of the well-financed company seeking to deliver the most reliable service at a premium price (netflix) while P2P remains the perfered distribution method of those who need good-enough delivery but can't afford to spend millions of dollars (Pirates, independent games developers, linux distros, non-profit media). There are a few exceptions like Blizzard who use p2p as a way to save a few bucks, but that's mostly how it goes. P2P offers 'good enough' for free, while CDN offers 'excellent' at a hefty price.
CDN = Internet features like Netflix, Amazon.com, etc we know today
Without a CDN structure in the web today, the speeds we see would only be a pipe dream.
SuricouRaven wrote: It's big downside is the very considerable cost - the hardware, rack rental, and negotiating deals with ISPs.
Costs are falling as demand is rising. Only good news here (for the CDN companies at a start). Why does it always have to be the ISPs... you know?
The comment by Kjella above addresses the seemingly moral boundary that exists with these costs:
Kjella wrote: Yeah. Sadly the Internet didn't evolve a generic, open CDN-ish system. Something not unlike a HTTP proxy, except you store binary blocks by hash. If some other guy from my ISP has downloaded the same torrent, I'd just grab it from my ISPs "CDN" server instead. Really just a HDD with a LRU cache, new stuff is added and the least used falls off.
Depending on how complicated you want it you could have a hierarchy of them, like first try locally, regionally, nationally etc. so stuff would only get pulled long-distance once. The MAFIAA would of course jump all over it but you could store encrypted pieces - wouldn't do them much good unless they had the access key. It'd probably save them bandwidth and money while we get faster content - a win-win. [interesting]
Instead you have to make deals with the big CDNs with again have to make deals with the big ISPs and everybody wants $$$ all the way. [Nothing new] Ah well the good thing is that P2P drives fiber everywhere - I heard that in my home town of 150k people there's now 2400 cable gates that need asphalt. Soon the "last mile" problem is a thing of the past. Well, except places like the US but I don't care about those:) [cable quantity and how that cable is laid are two different issues!]
However, we do have to remember on what existing backbone the internet was implemented: cable TV and internet still share the same physical connection for the majority of the country and thus the existing architectures were best equipped to deliver content to the users not from one user to another. People will always be greedy (see $$$ reference) but I say let the investors pay for the content delivery and do not regulate the rest...server farms are not cheap!
(Yes, there is much i missed and breezed over here SuricouRaven, et al...take it as a general observation and not as concrete fact please)
The future** is bright! _____ **patent pending.
^^Above comment does not argue with your guys ideas, just rambles really...
What I don't understand is why Netflix doesn't go to a BitTorrent style P2P swarm type streaming.
An ISP can use your traditional tv-cable easily to send you stuff, however uploading is rather difficult in many implementations. P2P/Bit-torrent-style distribution relies on uploading from the end-user. Instead, using a content delivery network through Level 3 communications, Netflix is able to almost have the "common" content "pre-delivered" to a more nearby location. This is good for the cable companies and users like me that still want a functioning internet when the nation logs onto Netflix in the evenings... This recent Ars Technica article explains some of this upload limitations and I found it to be a rather enjoyable read. Perhaps when the internet is ready and moved beyond the cable era, uploading will not be as much of a concern. Disclaimer: This is not to say there isn't room for P2P like implementations or various improvements in current algorithms and models...just your traditional P2P / Bittorrent distribution might not be the best implementation (sadly) here. Also I am no expert, I just stayed at a Holiday Inn Express...
AC stated: The result, almost everyone enrolled in the university is paying as much as demanded.
Take this in consideration...with the following (which you have) again:
From the sounds of it, there are a few different groups, and they each receive their own price. There's the people who can afford it, the people who can't but merit learning this, and the people in between. This is fine, and is nowhere near perfect price discrimination.
You nailed it.
and you're legally required to hand this information over,
Some facts are legally required, some not. (Or you can not provide them and mysteriously have your application rejected - or sent back with an incomplete ribbon.) Ironically, however, financial aid based on income of household (I have heard this so take it with a grain of salt)......accounts for a very small portion of financial aid.
I have been blessed in life but I also worked my a** off for some scholarships. Result: I paid in the bottom 10% (knowing a dead of admissions at another school provided me this insight.)
Lastly,
I'd be surprised if you don't have a case under your countries competition laws.
There should be a case for a lot of things these days in my opinion, esp IP Law, copyright, and other message-board-hot-topics... you get the point. (we all strive for a better world on this stuff, but reality is a b****)
Haha no-one does, but the fun part is to see whether a Windows based virus that uses a java vulnerability can still execute on a linux installation. That is all the "lightly-humored" comment of mine was targeted towards...
NO I do not run Java under Wine let alone a non-free distro... gnu all the way.
Slashdot Mirror
You bring up a good point because I wonder if updates count... correct me if I am wrong but updates require virtually (or exactly) the same size download as the original correct? Updates + variations = skewed statistics perhaps?
cheers
We changed our mailing address to our neighbors who did the complaining for us :)
(unless you are in the very small formally-proven-systems-written-in-Ada market, or something of that sort) v1.0 of snazzynewprotocol is a bit of a clusterfuck, and is available in only a single implementation, also highly dubious, while the old standbys have been polished considerably and have a number of implementations available...
Careful that we do not open Pandora's box here... (You know exactly what I am talking about, heh)
But on another note your exactly right. This article seems to talk about how protocols "evolved" but this is just as useful as painting a picture of the internet:
Time and time again I will see models looking at a picture of the internet "all at once", but without knowing what and why with each individual link, protocol, implementation, etc... this is a complete waste of time.
As you have said in so many words above, what these "researchers" did is a complete waste of time. Maybe they need to do some research on peering 101?
Disclaimer: I have not read the actual paper just a poorly written article linked by Slashdot. Perhaps there is much more to their work; if so, I do apologize.
I wonder if this rise in popularity can be attributed to the Chrome ads on Google's homepage we've seen in the past...
The article did not provide much analysis but rather a "news report" style...oh well.
mod parent up.
The shape of the netbook has changed, what we knew may be gone. Great point.
Exactly and I still love my linux book.
Touchpad - starting at least ~500.00
Netbook - ~250.00 (ish)
They are a different market (sorta). Personally I like buttons.
See the following process comparable to a maddening wikipedia link chain with no surprises:
(1) click on semi-relevant article with a title that seems like it may help to resolve said issue.
(2) find that this article has to be written for the likes of your grandmother (completely retarded user) or some weird error code (that uses like 8 digits these days) and is completely irrelevant.
(3) Look to the bottom and see See Also:
(4) Go to Step (1)
True story. In fact I bet this is really easy for content service providers because they only need to cache perhaps around 100 different help articles that can co-link and co-create a tangled mess from hell.
(You could say I am a Microsoft fan.)
What exactly gives a Bitcoin its value?
All the commercials by Goldline (mostly on FoxNews) of course.
with what seems like a recent increase in spam and general crap here I'm in a bit of a bit mood with /. and when that happens I tend to mod down rather than up.*
Ah I haven't been here too long myself but I can entirely relate.
modded posters get replies telling them to have a nice day
I usually end with that or cheers... for everyone. Although it is true people are a bit less insulting (in general) to moderators.
I'm not going to stop modding such things down but in future it'll be offtopic instead; I don't think that does anything to karma. Does it?
His comment, like many, is funny. Fortunately funny does not award karma but just makes /. users a tiny bit happier inside (not smarter or more well informed).
Modding funny does not affect karma and some people can choose to filter it if they want to read intelligence instead of a stream of jokes... this is what other people have told me in the past so take it with a grain of salt.
Personally I believe if you have the will to reply to my comment and to moderate good, you sir are ahead of 95%+ of the users here.
cheers and good night.
Yes. Just moderate then join back in the discussion on the next article. Posting comments is borderline taunting in some cases even if that was not your intention.
Have a nice day.
Sir, I do not have a Facebook account.
Although pointing a finger at the angry mob covered in shit surrounding me also may not be wise haha
SuricouRaven wrote:
A well-resourced CDN provides far better and more consistant performance than P2P, and does so while placing less load on the networks. It's big downside is the very considerable cost - the hardware, rack rental, and negociating deals with ISPs. They arn't even doing to talk to any company that doesn't have a few million dollars to bolster their reputation. So the CDN remains the prefered distribution method of the well-financed company seeking to deliver the most reliable service at a premium price (netflix) while P2P remains the perfered distribution method of those who need good-enough delivery but can't afford to spend millions of dollars (Pirates, independent games developers, linux distros, non-profit media). There are a few exceptions like Blizzard who use p2p as a way to save a few bucks, but that's mostly how it goes. P2P offers 'good enough' for free, while CDN offers 'excellent' at a hefty price.
Agree [one hundred gazillion percent (mod 100)] + 100
CDN = Internet features like Netflix, Amazon.com, etc we know today
Without a CDN structure in the web today, the speeds we see would only be a pipe dream.
SuricouRaven wrote:
It's big downside is the very considerable cost - the hardware, rack rental, and negotiating deals with ISPs.
Costs are falling as demand is rising. Only good news here (for the CDN companies at a start). Why does it always have to be the ISPs... you know?
The comment by Kjella above addresses the seemingly moral boundary that exists with these costs:
Kjella wrote:
Yeah. Sadly the Internet didn't evolve a generic, open CDN-ish system. Something not unlike a HTTP proxy, except you store binary blocks by hash. If some other guy from my ISP has downloaded the same torrent, I'd just grab it from my ISPs "CDN" server instead. Really just a HDD with a LRU cache, new stuff is added and the least used falls off.
Depending on how complicated you want it you could have a hierarchy of them, like first try locally, regionally, nationally etc. so stuff would only get pulled long-distance once. The MAFIAA would of course jump all over it but you could store encrypted pieces - wouldn't do them much good unless they had the access key. It'd probably save them bandwidth and money while we get faster content - a win-win. [interesting]
Instead you have to make deals with the big CDNs with again have to make deals with the big ISPs and everybody wants $$$ all the way. [Nothing new] Ah well the good thing is that P2P drives fiber everywhere - I heard that in my home town of 150k people there's now 2400 cable gates that need asphalt. Soon the "last mile" problem is a thing of the past. Well, except places like the US but I don't care about those :) [cable quantity and how that cable is laid are two different issues!]
However, we do have to remember on what existing backbone the internet was implemented: cable TV and internet still share the same physical connection for the majority of the country and thus the existing architectures were best equipped to deliver content to the users not from one user to another. People will always be greedy (see $$$ reference) but I say let the investors pay for the content delivery and do not regulate the rest...server farms are not cheap!
(Yes, there is much i missed and breezed over here SuricouRaven, et al...take it as a general observation and not as concrete fact please)
The future** is bright! _____ **patent pending.
^^Above comment does not argue with your guys ideas, just rambles really...
Good. When shit hits the fan now I can point my finger at one entity. Make my life easier!
What I don't understand is why Netflix doesn't go to a BitTorrent style P2P swarm type streaming.
An ISP can use your traditional tv-cable easily to send you stuff, however uploading is rather difficult in many implementations. P2P/Bit-torrent-style distribution relies on uploading from the end-user. Instead, using a content delivery network through Level 3 communications, Netflix is able to almost have the "common" content "pre-delivered" to a more nearby location. This is good for the cable companies and users like me that still want a functioning internet when the nation logs onto Netflix in the evenings...
This recent Ars Technica article explains some of this upload limitations and I found it to be a rather enjoyable read. Perhaps when the internet is ready and moved beyond the cable era, uploading will not be as much of a concern.
Disclaimer: This is not to say there isn't room for P2P like implementations or various improvements in current algorithms and models...just your traditional P2P / Bittorrent distribution might not be the best implementation (sadly) here. Also I am no expert, I just stayed at a Holiday Inn Express...
cheers
See also: recent video in a recent Slashdot article.
forgot to add (politely), I agree with what you say!
Should have taken the blue pill.
example
Damn fruit.
Thanks for that link though... I'm a CS student currently "in training"
(not a mistake i need to make in an interview...and knowledge i find useful)
More research and "reading up" is warranted on my part, also additional links are welcome.
Thanks man. (Yes I am a Slashdotter who admits when they are wrong, very rare)
Crap it is getting late.
Valid point sir as javascript may use as much memory as it deems necessary (and well the operating system allows)... and other factors,etc.
Good point sir! - (I am going to sleep now!)
I didn't post AC above... but i do read replies :)
AC stated: The result, almost everyone enrolled in the university is paying as much as demanded.
Take this in consideration...with the following (which you have) again:
From the sounds of it, there are a few different groups, and they each receive their own price. There's the people who can afford it, the people who can't but merit learning this, and the people in between. This is fine, and is nowhere near perfect price discrimination.
You nailed it.
and you're legally required to hand this information over,
Some facts are legally required, some not. (Or you can not provide them and mysteriously have your application rejected - or sent back with an incomplete ribbon.) Ironically, however, financial aid based on income of household (I have heard this so take it with a grain of salt)... ...accounts for a very small portion of financial aid.
I have been blessed in life but I also worked my a** off for some scholarships. Result: I paid in the bottom 10% (knowing a dead of admissions at another school provided me this insight.)
Lastly,
I'd be surprised if you don't have a case under your countries competition laws.
There should be a case for a lot of things these days in my opinion, esp IP Law, copyright, and other message-board-hot-topics... you get the point. (we all strive for a better world on this stuff, but reality is a b****)
cheers!
Wow I think that's the first time I've ever asked that question.
It is common practice at Slashdot to place stories and whatnot in the wrong section. Lemme go get my tin foil hat and ill show ya!
cheers!
Hmmm... Maybe that explains why your (obviously much more powerful) laptop gets the same virtual processor speed as my puny netbook!
guys, guys... it's Java remember.
Just increase the applet memory.
(Correct me if I'm wrong though: I'm curious too!)
So sad it should come to this...we tried to warn you all but oh dear!
Ah this translator seems much more useful than I previously had thought at first glance... maybe this didn't belong in idle after all?
Haha no-one does, but the fun part is to see whether a Windows based virus that uses a java vulnerability can still execute on a linux installation. That is all the "lightly-humored" comment of mine was targeted towards...
NO I do not run Java under Wine let alone a non-free distro... gnu all the way.
have you tried WINE?
Seriously, who would buy something like this?
I cannot think of a better gag gift for that someone who is addicted these gizmos.
Imagine the families' faces when that individual opens the gift to find a nose extension just as he/she had set down their iPhone... brilliant.