Skip to Content.
Sympa Menu

ndt-dev - Re: [ndt-dev] Re: Comment on revision r1039 in ndt

Subject: NDT-DEV email list created

List archive

Re: [ndt-dev] Re: Comment on revision r1039 in ndt


Chronological Thread 
  • From: Richard Carlson <>
  • To:
  • Subject: Re: [ndt-dev] Re: Comment on revision r1039 in ndt
  • Date: Wed, 26 Mar 2014 18:41:26 -0400

Hi Sebastian;

I don't have the ability to test all the clients. The NDT package doesn't even have copies of all of them. The point is that M-Lab put servers up in the public domain and encouraged developers and users to take advantage of them. That means that some care and consideration needs to be taken when thinking about changes to the message content or wire protocol.

Keep up the good work.

Rich

On 03/26/2014 02:57 AM, Sebastian Kostuch wrote:
Hi Richard,
in fact the only change I have made on server side was appending
additional results at the end
of existing one. So there shouldn't be any problems on client side (as
reading old results is done
the same way it was before) unless some client (as java one) modifies
these results before being
interpreted (there was adding client and server IP addresses visible
from client side). Is there
possibility to ensure that remaining clients does not do the same?

Regards,
Sebastian Kostuch

On 25.03.2014 23:29, Richard Carlson wrote:
Remember these are NOT the only clients.

Off the top of my head there are:
Android applet
BitTorrent micro-torrent integrated client

Rich

On 03/25/2014 09:59 AM,

wrote:
Comment by
:

General Comment:
Yes, I have added additional condition on java client side and it works
correctly on both old and new server versions. C client does not use new
variables at all and flash one doesn't support middlebox test so there
are no steps required to make them compatible also.

For more information:
https://code.google.com/p/ndt/source/detail?r=1039







Archive powered by MHonArc 2.6.16.

Top of Page