NDT-DEV email list created

[Jason Zurawski <>]

On 8/5/11 1:45 PM, thus spake Tiziana Refice:
> Jason,
>
> Good point.
> I was not aware of the ndt-dev ml.
>
> Is there anybody in particular we should contact, to make sure the
> proposed change does not conflict with anybody else's work?

I would suggest using the mailing list:

https://lists.internet2.edu/sympa/subscribe/ndt-dev

Thanks;

-jason

> --Tiziana
>
> On Fri, Aug 5, 2011 at 10:43, Jason Zurawski 
> <
> <mailto:>>
>  wrote:
>
>     Colliding two worlds.
>
>     In the future mlab-ops should be sure to let ndt-dev know of changes
>     like this since there are a lot more hands in the ndt code these days.
>
>     -jason
>
>     On 8/2/11 3:41 PM, thus spake Tiziana Refice:
>
>         I briefly chatted with Thomas a few days ago and he said he
>         could do it
>         (with some help from you) by the end of this week.
>
>         @Thomas,
>         Is it still the case?
>
>         On Tue, Aug 2, 2011 at 11:44, Josh King 
> <
>         
> <mailto:>
>         
> <mailto:
>         
> <mailto:>>>
>  wrote:
>
>             Hi Tiziana,
>
>             I think this sounds like a great idea. These are pretty
>         simple changes
>             that I could probably make pretty easily, but as I'll be out
>         of town for
>             the first half of next week I'm not sure if I'll have time
>         to do it on
>             that timeframe, so if someone else is able to to they should
>         feel free
>             to take pass at it.
>
>             On 08/01/2011 01:43 PM, Tiziana Refice wrote:
>          > Hi all,
>          >
>          > To make it easier to analyze the data collected by NDT, I suggest
>             to modify
>          > the NDT client to log the application that runs a test.
>          > This way it will be possible to tell from the logs that a
>             specific test
>          > result was collected by a specific application (e.g., uTorrent,
>             M-Lab web
>          > site, Android app, FCC client, ...).
>          >
>          > In particular, I suggest to log such info in the meta file.
>          > This is the file that writes the meta file (
>          >
>         
> http://www.google.com/__codesearch#3CG7BnVocJs/trunk/__src/logging.c&q=log%20package:__http://ndt%5C.googlecode%5C.__com
>         
> <http://www.google.com/codesearch#3CG7BnVocJs/trunk/src/logging.c&q=log%20package:http://ndt%5C.googlecode%5C.com>
>         
> <http://www.google.com/__codesearch#3CG7BnVocJs/trunk/__src/logging.c&q=log%20package:__http://ndt%5C.googlecode%5C.__com
>         
> <http://www.google.com/codesearch#3CG7BnVocJs/trunk/src/logging.c&q=log%20package:http://ndt%5C.googlecode%5C.com>>),
>          > function writeMeta().
>          > For example, we could add an extra line after line 553, something
>             like
>          > fprintf(fp, "client application: %s\n", meta.application);
>          > where meta.application could be initialized from some
>         property file.
>          >
>          > What do you think?
>          >
>          > The goal is to make the change by the beginning of next week.
>          >
>          > In addition, I wonder if it would make sense to also log the
>             version of the
>          > NDT client.
>          >
>          > Cheers,
>          > --Tiziana
>          >
>
>             --
>             Josh King
>             Technologist
>             Open Technology Initiative
>             New America Foundation