retransmission

And the issue is with delayed retransmission when you can actually afford to send the packets again preemptively.

If the problem is ephemeral, it will be resolved on the retransmission.

Are you talking about something at a lower level, perhaps the TCP level mechanics for retransmission of dropped packets?

TCP's minimum retransmission timeout is currently 200 milliseconds, which would make even a single dropped packet catastrophic for latency).

Dish has carriage agreements and pays retransmission fee's to the networks they carry via this service.

Even dynamic content can benefit from caching in the case of multicast (e.g., teleconferencing) or packet retransmission after a packet loss.

TP4, where the 4 stands for class 4, certainly had reordering, retransmission and congestion controll.

All long range transmissions use forward error correcting codes which allow error detection and data recovery without retransmission (to a certain degree).

This avoids ambiguity about which packets have been received and avoids dreaded retransmission timeouts.

This makes the upper layer connection stall very quickly and every retransmission just adds to the problem - an internal meltdown effect.

To the point about retransmission: in UDP cases, it's often not worth bothering about.

The trouble with this is that it's still possible to identify the retransmission.

I would still agree that you should have a bit in the TCP stack that says "use preemptive retransmission instead of waiting so damn long".

Except as otherwise expressly permitted under copyright law, no copying, redistribution, retransmission, publication or commercial exploitation of downloaded material will be permitted without the express permission of CNN and the copyright owner." The other sites you liked have similar clauses.