Hi Linda,

The key goal of the draft is to do a requirements analysis on the transport 
services needed for low-latency and real-time applications. The proposed 
abstract API is the minimal API to achieve this. The new API features, compared 
to the Sockets API, are those in Sections 3.4-3.6.

We have an experimental implementation of an API much like this, written as 
part of our TCP Hollywood work (https://csperkins.org/research/tcp-hollywood/). 
This is a research prototype, so there are no deployed applications using it. 

If you’re interested in the API work, the proposal in 
draft-trammell-post-sockets-00 is a more interesting, and radical, approach, 
and should also meet the requirements we outline.

Colin




> On 11 Nov 2016, at 10:01, Linda Dunbar <[email protected]> wrote:
> 
> Dear Authors, 
>  
> Your draft listed down several APIs. But which one is used by Applications 
> that need low latency?
>  
> Aren’t the “Socket Setup & Tear down” described in Section 3.1 the original 
> Berkley sockets APIs? If not, what are the differences?
>  
> How about the “getsocketopt” and “Setsockopt”?  and the APIs listed in 
> Section 3.3.?
>  
> Which APIs are used by Applications that need low Latency? 
>  
> Are there any Applications today using the APIs proposed in 
> draft-mcquistin-taps-low-latency-services?
>  
> Thank you very much, 
>  
> Linda Dunbar



-- 
Colin Perkins
https://csperkins.org/




_______________________________________________
Taps mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/taps

Reply via email to