Speed !== bandwidth, even with infinite bandwidth, a packet from here to the sun takes 8 minutes to arrive there, plus 8 minutes more for the ack packet to come back, 16 minutes during which the sending end believes the packets have been lost begins retransmitting again and again. IOW in TCP *sometimes* latency alone can totally ruin throughput, ack packets are the culprit, and TCP's ack mechanisms have been tweaked several times already for this reason. Perhaps Fujitsu has devised some clever hack, why not? Since when does big company necessarily === a bunch of incompetent idiots? -- Jorge.
On 31/01/2013, at 10:14, Mark Hahn wrote: > In order to be faster by 30X, the current tcp would have to only be using 3% > of the bandwidth. Kind of ridiculous. > > > On Wed, Jan 30, 2013 at 8:23 PM, Thomas Shinnick <tshin...@gmail.com> wrote: > To the extent that it sounds like a broad claim to solve the world's > problems, yes. It really is aimed towards congested networks. Note the > qualification "in a low-quality communications environment". Apparently > there's been such research ongoing for many years, and elsewhere. So, more > "ain't our company great" BS than marketing BS. That is, only the internal > people are going to be impressed...? > > > On Wednesday, January 30, 2013 9:51:12 PM UTC-6, Luke Arduini wrote: > Looks like marketing bs > > On Wednesday, January 30, 2013, Tristan Slominski wrote: > Came across this. Seems the Node.js community would care about it, if nothing > else, for situational awareness: > > http://www.fujitsu.com/global/news/pr/archives/month/2013/20130129-02.html > > The problem, of course, is that it's proprietary. > > Fujitsu Laboratories has succeeded at a software-only approach, developing: > 1) A new protocol that incorporates an efficient proprietarily developed > retransmission method based on user datagram protocol (UDP)(2), an optimized > way to deliver streaming media able to reduce latency resulting from data > retransmission when packet loss occurs; 2) Control technology that addresses > the problem of UDP transmissions consuming excess bandwidth by performing a > real-time measurement of available network bandwidth and securing an optimal > amount of communications bandwidth without overwhelming TCP's share of the > bandwidth; and 3) Technology that, by employing the new protocol, makes it > possible to easily speed up existing TCP applications without having to > modify them. > > -- > -- > Job Board: http://jobs.nodejs.org/ > Posting guidelines: > https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines > You received this message because you are subscribed to the Google > Groups "nodejs" group. > To post to this group, send email to nodejs@googlegroups.com > To unsubscribe from this group, send email to > nodejs+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/nodejs?hl=en?hl=en > > --- > You received this message because you are subscribed to the Google Groups > "nodejs" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to nodejs+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > > > -- > -- > Job Board: http://jobs.nodejs.org/ > Posting guidelines: > https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines > You received this message because you are subscribed to the Google > Groups "nodejs" group. > To post to this group, send email to nodejs@googlegroups.com > To unsubscribe from this group, send email to > nodejs+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/nodejs?hl=en?hl=en > > --- > You received this message because you are subscribed to the Google Groups > "nodejs" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to nodejs+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > > > > -- > -- > Job Board: http://jobs.nodejs.org/ > Posting guidelines: > https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines > You received this message because you are subscribed to the Google > Groups "nodejs" group. > To post to this group, send email to nodejs@googlegroups.com > To unsubscribe from this group, send email to > nodejs+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/nodejs?hl=en?hl=en > > --- > You received this message because you are subscribed to the Google Groups > "nodejs" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to nodejs+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > -- -- Job Board: http://jobs.nodejs.org/ Posting guidelines: https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines You received this message because you are subscribed to the Google Groups "nodejs" group. To post to this group, send email to nodejs@googlegroups.com To unsubscribe from this group, send email to nodejs+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/nodejs?hl=en?hl=en --- You received this message because you are subscribed to the Google Groups "nodejs" group. To unsubscribe from this group and stop receiving emails from it, send an email to nodejs+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
