Comments on: BitTorrent in the Browser http://ajaxian.com/archives/bittorrent-in-the-browser Cleaning up the web with Ajax Thu, 09 Feb 2012 06:55:33 +0000 hourly 1 http://wordpress.org/?v=3.2 By: kristianj http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270599 kristianj Tue, 20 Jan 2009 11:46:54 +0000 http://ajaxian.com/?p=5718#comment-270599 BitTorrent is a great technology, but it can be a bit hard to grasp for end users - especially that you must manually stop seeding the torrent. BitTorrent is a great technology, but it can be a bit hard to grasp for end users – especially that you must manually stop seeding the torrent.

]]> By: Brad Neuberg http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270593 Brad Neuberg Tue, 20 Jan 2009 05:32:25 +0000 http://ajaxian.com/?p=5718#comment-270593 This is an interesting idea; I like the use of Content Negotiation with the .torrent file returned. I thought that we would have seen BitTorrent integrated into browsers and the web years ago, and I've been surprised that its taken so long. I know that Firefox has had an open patch for many years around this. Good for Opera taking a leadership role here. This is an interesting idea; I like the use of Content Negotiation with the .torrent file returned. I thought that we would have seen BitTorrent integrated into browsers and the web years ago, and I’ve been surprised that its taken so long. I know that Firefox has had an open patch for many years around this. Good for Opera taking a leadership role here.

]]> By: raccettura http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270579 raccettura Mon, 19 Jan 2009 21:19:16 +0000 http://ajaxian.com/?p=5718#comment-270579 @dmw: You didn't read what I wrote.... 1. Accept Encoding is correct since the server would serve a .torrent (over HTTP) as the response should it find the browser can accept it. The DNS layer is a fallback <b>initiated by the browser</b> should the HTTP download server fail. This is by far the most common point of failure in a high traffic environment. Accept Encoding doesn't switch protocols as you claim it merely tells the server if it can offer a .torrent or not. 2. DNS would really hold only 1-2kb at most. It would hold the .torrent which generally run 1kb for essential info only. Megabytes is either a gross exaggeration or misunderstanding, I'll assume the ladder. As I mentioned this is not really that different than storing SPF data in a TXT or SPF record. Downloading the actual file is up to the torrent network. ISP's could cache the torrent itself if they wanted to keep it on their network and offer faster downloads to their customers. @dmw: You didn’t read what I wrote….

1. Accept Encoding is correct since the server would serve a .torrent (over HTTP) as the response should it find the browser can accept it. The DNS layer is a fallback initiated by the browser should the HTTP download server fail. This is by far the most common point of failure in a high traffic environment. Accept Encoding doesn’t switch protocols as you claim it merely tells the server if it can offer a .torrent or not.

2. DNS would really hold only 1-2kb at most. It would hold the .torrent which generally run 1kb for essential info only. Megabytes is either a gross exaggeration or misunderstanding, I’ll assume the ladder. As I mentioned this is not really that different than storing SPF data in a TXT or SPF record. Downloading the actual file is up to the torrent network. ISP’s could cache the torrent itself if they wanted to keep it on their network and offer faster downloads to their customers.

]]> By: Aimos http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270577 Aimos Mon, 19 Jan 2009 19:56:57 +0000 http://ajaxian.com/?p=5718#comment-270577 EVERY! - not only Opera - browser should support the torrent protocol. The protocol itself has advanced over the years and with webseeds, we have something which can benefit from both worlds: direct downloads and "load balancing" by p2p. . If every browser would have .torrent support you could set up servers which acts like p2p clients and akamai would be out of business in no time. All server-torrent-client would deliver the downloads via the protocol and each downloading user would be give something back because he is uploading also. . Opensource websites like sourceforge.net which might have terabytes of download each month and can only be done by sponsoring are websites I would completly redo with p2p-torrent downloads if every browser would support it out of the box. . I think at least in the OS community there is a strong give-something-back mind, so if you use a software and you give upload back for your favorite software and save a developer costs of 10GB it (would) feel right. . So! call for action: Firefox, Webkit, IE .. support the torrent protocol please! EVERY! – not only Opera – browser should support the torrent protocol. The protocol itself has advanced over the years and with webseeds, we have something which can benefit from both worlds: direct downloads and “load balancing” by p2p.
.
If every browser would have .torrent support you could set up servers which acts like p2p clients and akamai would be out of business in no time. All server-torrent-client would deliver the downloads via the protocol and each downloading user would be give something back because he is uploading also.
.
Opensource websites like sourceforge.net which might have terabytes of download each month and can only be done by sponsoring are websites I would completly redo with p2p-torrent downloads if every browser would support it out of the box.
.
I think at least in the OS community there is a strong give-something-back mind, so if you use a software and you give upload back for your favorite software and save a developer costs of 10GB it (would) feel right.
.
So! call for action: Firefox, Webkit, IE .. support the torrent protocol please!

]]> By: antini http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270569 antini Mon, 19 Jan 2009 16:53:05 +0000 http://ajaxian.com/?p=5718#comment-270569 For <a href="http://www.metalinker.org" rel="nofollow">metalink</a> (FTP/HTTP/P2P failover & other features), we’ve been using transparent content negotiation which is apparently not the correct thing to do in some people’s opinion, but works. I’ve been told the HTTP Link header is the correct thing for us, so I’m hoping it’d be good for torrents too. If anyone is interested in this, feel free to help us improve what we’re doing by dialoging and collaborating. For metalink (FTP/HTTP/P2P failover & other features), we’ve been using transparent content negotiation which is apparently not the correct thing to do in some people’s opinion, but works. I’ve been told the HTTP Link header is the correct thing for us, so I’m hoping it’d be good for torrents too.

If anyone is interested in this, feel free to help us improve what we’re doing by dialoging and collaborating.

]]> By: ajaxery http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270568 ajaxery Mon, 19 Jan 2009 16:49:07 +0000 http://ajaxian.com/?p=5718#comment-270568 i believe there are several firefox extensions that do this. right? i believe there are several firefox extensions that do this. right?

]]> By: DoubleAW http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270563 DoubleAW Mon, 19 Jan 2009 16:16:50 +0000 http://ajaxian.com/?p=5718#comment-270563 I agree with dmw about the implementation, but the idea itself could be very, very useful. Maybe if there was a better way to do it. I agree with dmw about the implementation, but the idea itself could be very, very useful.

Maybe if there was a better way to do it.

]]> By: dmw http://ajaxian.com/archives/bittorrent-in-the-browser/comment-page-1#comment-270554 dmw Mon, 19 Jan 2009 13:29:54 +0000 http://ajaxian.com/?p=5718#comment-270554 This is horrid. Not only is it an abuse of Accept-encoding (which indicates the values of "Content-encoding" response header which will be accepted), but it also goes on to suggest stuffing a file into DNS is a good idea. Let's reiterate: - Uses the Accept-encoding header to signal a protocol switch. - Suggests using the DNS to store files which could be megabytes in size and will only get bigger as time goes on. Compatibility-wise I guess it doesn't break anything, until lots of people start using it and ISPs find that 4gb is no longer sufficient RAM for their caching resolvers. I couldn't think of a worse solution to this problem. This is horrid. Not only is it an abuse of Accept-encoding (which indicates the values of “Content-encoding” response header which will be accepted), but it also goes on to suggest stuffing a file into DNS is a good idea.

Let’s reiterate:

– Uses the Accept-encoding header to signal a protocol switch.
– Suggests using the DNS to store files which could be megabytes in size and will only get bigger as time goes on.

Compatibility-wise I guess it doesn’t break anything, until lots of people start using it and ISPs find that 4gb is no longer sufficient RAM for their caching resolvers.

I couldn’t think of a worse solution to this problem.

]]>