Re: Differences between iQUIC and gQUIC

[Ian Swett <[e-mail]>]

I agree with all of Matt's points.

And there are a lot of differences, more than I could enumerate.

Ian

On Mon, Oct 18, 2021 at 2:55 PM Matt Joras <[e-mail]> wrote:

> Hi Eric,
>
> In general I would advise not assuming the findings for studies
> focused on Google QUIC will apply to IETF QUIC. While many findings
> may apply, they are different protocols with different designs, and
> any research done on Google QUIC most likely focused on a single
> implementation (the Chromium implementation). Even within a single
> standard (IETF QUIC), we expect there are going to be different
> experimental results depending on implementations studied [1]. As such
> I would say right now it's best to rely on research conducted on IETF
> QUIC, rather than trying to transform results previously obtained for
> Google QUIC.
>
> All the best,
> Matt Joras
>
> [1]
> https://www.researchgate.net/publication/342783300_Same_Standards_Different_Decisions_A_Study_of_QUIC_and_HTTP3_Implementation_Diversity
>
> On Mon, Oct 18, 2021 at 2:43 PM Feltgen, Eric
> <[e-mail]> wrote:
> >
> > Dear QUIC working group,
> >
> > I am a CS student at RWTH Aachen University currently researching papers
> on the performance of QUIC in the context of a seminar. In the past years,
> there have been many papers on this topic testing many different versions
> of Google's QUIC. Over the years, many changes have been made to the
> protocol and the implementation at Google.
> >
> > In order to understand which results are comparable to the standardized
> version of QUIC (RFC 9000), it would be helpful to which changes have been
> made in the last years of the Google version and how the last QUIC versions
> by Google compare to the QUIC standard. In my research so far, I found out
> that forward error correction has been removed from the Google version a
> couple of years ago. So my question is whether there are any breaking
> changes that have been made in the development of the IETF standard which
> could result in different behaviour.
> >
> > Thank you in advance for any response! It would help a lot to understand
> the development stop of the QUIC standard within the last years.
> >
> > Best regards,
> >
> > Eric Feltgen
> >
>
>