NDT-DEV email list created

[Aaron Brown <>]

Hey Jordan,

Would it be possible to get a warning stuck in the results page as well?

Cheers,
Aaron

On Nov 7, 2014, at 2:25 PM, Jordan McCarthy 
<>
 wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA512
> 
> Hi Aaron,
>       Yes indeed: in all three cases, a red warning message is displayed
> describing the problem.  I forgot to mention that you can see examples
> of the three states here:
> 
> a. http://files.opentechinstitute.org/~mccarthy/flash-warn.html
> b. http://files.opentechinstitute.org/~mccarthy/flash-warn-limit.html
> c. http://files.opentechinstitute.org/~mccarthy/error.html
> 
> Regards,
> 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 02:11 PM, Aaron Brown wrote:
>> 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?
>> 
>> Cheers, Aaron
>> 
>> On Nov 7, 2014, at 1:36 PM, Jordan McCarthy
>> <>
>>  wrote:
>> 
>> 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 https://code.google.com/p/ndt/issues/detail?id=162
>>>>> 
>>>>> 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.
>>>>> 
>>> <ndt_flash_environment_detection.patch>
>> 
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.11 (GNU/Linux)
> 
> iQEcBAEBCgAGBQJUXRynAAoJEL+9ounAjYBCS2sH/2TVdbbNd8l4R86y7/at/Ax7
> 9BWRpkFSzzWmReHV0OPs2nAWhfqG89BgTFoqFu5BUNJjUu3dA5M7LSPm1/Xbmkp1
> vseIgPhXIQcJtnSn3b9cqYr3zRmvSvT2pJGqvP1PQg8QZLAeG47oqjjcftMok0Xu
> CF+DWD1/Y/ybGjAvXkTSuay7pkSvCdgtEOOzQfU7t2x1ewZspgBl3mzqniIFETHu
> IoQCrsfS3IKbgwiv8hGN8+IW59OgrXjGXPdX7Ig/Kit+WNYdBgay4dVMLgErV/u8
> 4A9e//THB5ihxObtD/C806aoWxlVojjDPKWJvr0vyNv5AzFgyfmBjr/8qa6Fwxc=
> =apLs
> -----END PGP SIGNATURE-----