Skip to Content.
Sympa Menu

ndt-dev - Re: [ndt-dev] Issue 162 in ndt: New NDT release

Subject: NDT-DEV email list created

List archive

Re: [ndt-dev] Issue 162 in ndt: New NDT release

Chronological Thread 
  • From: Aaron Brown <>
  • To: Jordan McCarthy <>
  • Cc: "" <>
  • Subject: Re: [ndt-dev] Issue 162 in ndt: New NDT release
  • Date: Fri, 7 Nov 2014 19:11:39 +0000
  • Accept-language: en-US

Hey Jordan,

In the (a) case, does it spit out an error telling people that there’s a
problem, or does it just go on as though everything is fine?


On Nov 7, 2014, at 1:36 PM, Jordan McCarthy

> Hash: SHA512
> I'm not sure if this should be handled separately, but the M-Lab team
> has a series of changes that we would like to add to the flash client
> to handle the fact that the Adobe flash runtime for Mac OS X and Linux
> is severely performance-limited.
> Here's the description of what we did:
> Our patch (attached) adds a customizable feature for dealing with
> performance-limited flash runtimes. In a nutshell, the patch allows
> anyone who hosts an NDT test to specify (via an HTML parameter)
> whether performance-limited flash runtimes should (a) be treated the
> same as non-limited ones, (b) only be allowed to run against a
> "fallback" server, or (c) not allowed to run the NDT test via flash at
> all. Measurement Lab will generally use option (c), because we are
> primarily interested in identifying and measuring bottlenecks in the
> performance of networks, rather than clients. However, we completely
> understand the argument that some researchers might actually want the
> speed limitations introduced by particular runtimes to be captured in
> experimental data, and our patch makes it very easy to switch between
> these two approaches, depending on experimenters' needs.
> The patch has been tested extensively on all major web browsers on
> Windows, Mac OS X, and Linux, and appears to work well in all
> environments except for Safari (where the flash client doesn't work at
> all).
> Thanks,
> Jordan
> Jordan McCarthy
> Open Technology Institute | New America Foundation
> Public Key: 0xC08D8042 | 4A61 3D39 4125 127D 65EA DDC2 BFBD A2E9 C08D 8042
> On 11/07/2014 11:00 AM,
> wrote:
>> Comment #3 on issue 162 by
>> :
>> New NDT release
>> All flash client related improvements have been fixed and changes
>> are available on FlashClientFixes branch. Please feel free to
>> review them and if all will look ok then I will merge them into
>> trunk in nearest time.
> Version: GnuPG v1.4.11 (GNU/Linux)
> gXgrrB+d/d/bb3/1Kn9EizsYzdaeRo7faLWNKeiO3ePjX8cfcAP8idBJko1E+COZ
> O4C1gkXkcVdeG3vN5JvF3gHcnRpC1pvfFVixJ3q9L6pWPVT4be3NtZDQ46xR9XxS
> iu3n3E2kP/C5vowm+coyuEZ9jVZvX74tyGxfc5vNDrJLE4Y9evXui9pJkh6vQ9nL
> CqkjCQMhuYadFZmiDdrZC5Fut8cSoirgq8prNY+odnoHKFOZFG1mL2O52/RWD/JH
> 1wFZxqJFo+lLgrTMhC/0n7Qg+k9fUW2tXTt7hVfKoEP/p/3F8Ml9aWQXEgAqWgA=
> =wgkG
> <ndt_flash_environment_detection.patch>

Archive powered by MHonArc 2.6.16.

Top of Page