From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-io0-x229.google.com (mail-io0-x229.google.com [IPv6:2607:f8b0:4001:c06::229]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority G2" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 97FF921FAFD for ; Fri, 14 Aug 2015 07:29:26 -0700 (PDT) Received: by iodb91 with SMTP id b91so86004848iod.1 for ; Fri, 14 Aug 2015 07:29:25 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=DCx3VtwclpPceIm+wBZS5hOFmEEOsk1vvbDSmUx56SE=; b=tUX7mkFlQdAOWA2zKFxnBvKSlc9tG69Ln0VGpQriTDsMYbDrPDwsApBdlanrPuMoaf icr52LDMoGXjXs6KKIRfqZV1L71rtKu6QK4Qle1+kL15ZumQEpgbKVG9I99UFIC9FClc 1+qD5GbldrQRq4Ix6ESk8wQFrlas8u1ruPQkdFo1/0qQBWp0soPRfqU46y5cQIlCa21W Ir2kzBBLeOFD4t+oWSN9DfCmKYFm8AdOTmn/7fOaybv2sA128VgHmm7qkMJ+QHeElHjL BWwXM0DK+53LBQL0qrxCW/neBGlF3vwyz7xJwJMBqfBRfFy+8y5IgddXl6W4/dSh43QS KHCg== MIME-Version: 1.0 X-Received: by 10.107.10.9 with SMTP id u9mr10799571ioi.118.1439562565150; Fri, 14 Aug 2015 07:29:25 -0700 (PDT) Sender: justinbeech@gmail.com Received: by 10.50.107.42 with HTTP; Fri, 14 Aug 2015 07:29:25 -0700 (PDT) In-Reply-To: <55CDA6A2.605@oldelvet.org.uk> References: <55CDA6A2.605@oldelvet.org.uk> Date: Sat, 15 Aug 2015 00:29:25 +1000 X-Google-Sender-Auth: TcI6GMw8ZqvcmMIXa9JHglIncYc Message-ID: From: jb To: Richard Mortimer Content-Type: multipart/alternative; boundary=001a113ee47e53f057051d464716 Cc: bloat Subject: Re: [Bloat] teh tech guy has got bufferbloat bad on his link X-BeenThere: bloat@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: General list for discussing Bufferbloat List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 14 Aug 2015 14:29:49 -0000 --001a113ee47e53f057051d464716 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Yes speedtest.net license a white label version where the ISP can apply the branding for example Xfinity/Comcast test is just speedtest.net in a new dress. As you mentioned ISPs like it because they don't have to pay for as much bandwidth and locating it in their noc can avoid "testing" a lot of other things. Things both in their control, and out of their control. Also externally hosted tests did themselves no favours by being single stream so were quickly crippled by only running as fast as the TCP send/receive maximum window sizes let them for the latency, but speedtest was (on the quiet!) loads of parallel streams. I've learned that customers want to believe the fastest result they can possibly get. In fact a super successful speed test would ask the user how fast their line is supposed to be, then it would display some nifty graphics for a while and then display that speed back, plus 3% extra :) On Fri, Aug 14, 2015 at 6:28 PM, Richard Mortimer wrote: > > > On 14/08/2015 08:30, jb wrote: > > hey, Leo is a good guy :) > > > > If someone was to explain to him bufferbloat he'd probably find a way t= o > > include something on it in a future show. > > Leo did mention bufferbloat towards the end of that segment around 1:51 > when he was explaining the dslr results screen. But it would have been > nice if that had been mentioned to the caller direct. > > I have also heard Leo and Steve talk about bufferbloat on Security Now > occasionally. A quick search shows that they devoted a whole episode to > it in March 2012 > https://www.grc.com/sn/sn-345.htm > https://www.grc.com/sn/sn-359.htm > > Just one thing to note about the ISPs preference for speedtest.net. My > ISP here in the UK use what is I think a co-located (rebadged) > speedtest.net server in their data centres. I think that doing it that > way means they are testing from their core network to me and can > discount any wider internet issues outside of their control. They also > get to see the results direct from the server so they do not need to > "trust" my version of the results. > > I had a problem a few months back and they told me to test there instead > of at dslr. However their license key for the speedtest server had > expired and it gave me a nice error screen revealing a lot of the setup > behind the curtains that revealed it was powered by speedtest. > > Regards > > Richard > > > > > > On Thu, Aug 13, 2015 at 2:36 AM, Dave Taht > > wrote: > > > > but missed talking about it. Peaking at 600ms of delay at 100mbit, > > shown at about 1:45. > > > > see vlog at: > > > > https://www.twit.tv/shows/the-tech-guy/episodes/1204?autostart=3Dfa= lse > > > > and spends 5k a month on his internet access. > > > > -- > > Dave T=C3=A4ht > > worldwide bufferbloat report: > > http://www.dslreports.com/speedtest/results/bufferbloat > > And: > > What will it take to vastly improve wifi for everyone? > > https://plus.google.com/u/0/explore/makewififast > > _______________________________________________ > > Bloat mailing list > > Bloat@lists.bufferbloat.net > > https://lists.bufferbloat.net/listinfo/bloat > > > > > > > > > > _______________________________________________ > > Bloat mailing list > > Bloat@lists.bufferbloat.net > > https://lists.bufferbloat.net/listinfo/bloat > > > --001a113ee47e53f057051d464716 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Yes speedtest.net lic= ense a white label version where the ISP can
apply the branding for exa= mple Xfinity/Comcast test is just speedtes= t.net
in a new dress.

As you mentioned ISP= s like it because they don't have to pay for as much
bandwidt= h and locating it in their noc can avoid "testing" a lot
of other things. Things both in their control, and out of their control.<= /div>

Also externally hosted tests did themselves no fav= ours by
being single stream so were quickly crippled by only runn= ing as fast
as the TCP send/receive maximum window sizes let them= for the
latency, but speedtest was (on the quiet!) loads of para= llel streams.
I've learned that customers want to believe the= fastest result they can
possibly get.

I= n fact a super successful speed test would ask the user how
fast = their line is supposed to be, then it would display some nifty
gr= aphics for a while and then display that speed back, plus 3% extra :)
=


On Fri, Aug 14, 2015 at 6:28 PM, Richard Mortimer <richm@old= elvet.org.uk> wrote:


On 14/08/2015 08:30, jb wrote:
> hey, Leo is a good guy :)
>
> If someone was to explain to him bufferbloat he'd probably find a = way to
> include something on it in a future show.

Leo did mention bufferbloat towards the end of that segment around 1= :51
when he was explaining the dslr results screen. But it would have been
nice if that had been mentioned to the caller direct.

I have also heard Leo and Steve talk about bufferbloat on Security Now
occasionally. A quick search shows that they devoted a whole episode to
it in March 2012
https://www.grc.com/sn/sn-345.htm
https://www.grc.com/sn/sn-359.htm

Just one thing to note about the ISPs preference for speedtest.net. My
ISP here in the UK use what is I think a co-located (rebadged)
speed= test.net server in their data centres. I think that doing it that
way means they are testing from their core network to me and can
discount any wider internet issues outside of their control. They also
get to see the results direct from the server so they do not need to
"trust" my version of the results.

I had a problem a few months back and they told me to test there instead of at dslr. However their license key for the speedtest server had
expired and it gave me a nice error screen revealing a lot of the setup
behind the curtains that revealed it was powered by speedtest.

Regards

Richard


>
> On Thu, Aug 13, 2015 at 2:36 AM, Dave Taht <dave.taht@gmail.com
> <mailto:dave.taht@gmail.com>> wrote:
>
>=C2=A0 =C2=A0 =C2=A0but missed talking about it. Peaking at 600ms of de= lay at 100mbit,
>=C2=A0 =C2=A0 =C2=A0shown at about 1:45.
>
>=C2=A0 =C2=A0 =C2=A0see vlog at:
>
>=C2=A0 =C2=A0 =C2=A0https:= //www.twit.tv/shows/the-tech-guy/episodes/1204?autostart=3Dfalse
>
>=C2=A0 =C2=A0 =C2=A0and spends 5k a month on his internet access.
>
>=C2=A0 =C2=A0 =C2=A0--
>=C2=A0 =C2=A0 =C2=A0Dave T=C3=A4ht
>=C2=A0 =C2=A0 =C2=A0worldwide bufferbloat report:
>=C2=A0 =C2=A0 =C2=A0http://www.dslreports= .com/speedtest/results/bufferbloat
>=C2=A0 =C2=A0 =C2=A0And:
>=C2=A0 =C2=A0 =C2=A0What will it take to vastly improve wifi for everyo= ne?
>=C2=A0 =C2=A0 =C2=A0https://plus.google.com/u/0/= explore/makewififast
>=C2=A0 =C2=A0 =C2=A0_______________________________________________
>=C2=A0 =C2=A0 =C2=A0Bloat mailing list
>=C2=A0 =C2=A0 =C2=A0Bloat@lists.bufferbloat.net <mailto:Bloat@lists.bufferbloat.net>
>=C2=A0 =C2=A0 =C2=A0https://lists.bufferbloat.net/li= stinfo/bloat
>
>
>
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat= .net
> https://lists.bufferbloat.net/listinfo/bloat >

--001a113ee47e53f057051d464716--