Re: #537: Remove segments (consensus call)

You're welcome to ask Twitter about it.

As I recall it, certain client implementations fail when chunks are not
done in a particular way, despite the fact that the spec says otherwise.
In my experience, supporting old hardware-based implementations requires
annoying hacks such as this when one really cares about compatibility.
-=R


On Tue, Jul 1, 2014 at 11:36 PM, Poul-Henning Kamp <phk@phk.freebsd.dk>
wrote:

> In message <CAP+FsNdwizDUMCunLGhftkfgbJfstvAG3Ux8qtpHuL=
> jHv5eAA@mail.gmail.com>, Roberto Peon writes:
>
> >For my part, I still see that END_SEGMENT or something similar provides
> for
> >interesting characteristics for HTTP2->1 gateways in allowing chunks to be
> >reproduced (and we know that there are issues there) as well as allowing
> >the coalescing/tunneling of protocols such as WS onto the connection,
> >making them more viable.
>
> What issues would those be and where do we know them to exist ?
>
> Are they hacks to be discouraged or architecture to be respected ?
>
> I have yet to hear about any issues with Varnish making totally random
> (seen form the applications point of view) decisions about chunk sizes.
>
> --
> Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
> phk@FreeBSD.ORG         | TCP/IP since RFC 956
> FreeBSD committer       | BSD since 4.3-tahoe
> Never attribute to malice what can adequately be explained by incompetence.
>