Re: [rmcat] delay recovery phase

[John Leslie <[e-mail]>]

Stefan Holmer <[e-mail]> wrote:
> 
> In https://tools.ietf.org/html/draft-alvestrand-rmcat-congestion-02#page-10
> we go into a HOLD state where the sender should keep its bitrate constant
> while waiting for buffers to drain.

   The first time I read this, I thought you were saying to hold the rate
constant regardless of what was happening to the delay.

   I think it would be good to explicitly state something to the effect
that rmcat will have a target delay, and will not increase sending rate
until that target is achieved.

   (Of course, that target may change if it turns out to not be
achievable...)

> We don't try to estimate the time needed to drain the buffers, but
> instead look at the delay change estimate m(i); if it's less than zero
> buffers are likely draining.

   There must be a better way of saying this...

   We can only track the delay (including rate-of-change): IMHO it's
better not to talk in terms of buffers draining, merely delay.

   (Isn't that the actual goal of rmcat -- to achieve low-enough delay?)

--
John Leslie <[e-mail]>