NDT-DEV email list created

[Aaron Brown <>]

Hey Jakub,

We’ll need to make sure that any changes like this are backwards compatible 
with existing clients. As long as it doesn’t affect older versions of the 
applet, the command-line client or things like uTorrent that have NDT tests 
built in, I have no strong opinions on the matter.

Cheers,
Aaron

On Feb 19, 2014, at 9:24 AM, Jakub Sławiński 
<>
 wrote:

> 
> Hi,
> 
> in my opinion we should keep only 9988 and 9977 statuses. 9988 should be
> send only when the server is too busy to handle new clients. In all
> other cases the server should send 9977.
> 
> The specification of the NDT Protocol should be changed accordingly.
> 
> Please notice, that Aaron and/or Tiziana should confirm the above proposal.
> 
> 
> Regards,
>  Jakub.
> 
> On 02/13/2014 02:46 PM, Sebastian Malecki wrote:
>> According to thread https://code.google.com/p/ndt/issues/detail?id=102, I
>> would like to discuss about which statuses should be send by server.
>> 
>> 1. "9999 - Server Busy: Please wait 60 seconds for previous test to 
>> finish" -
>> should it be? 
>> At first I will try explain why client should still handle "9999" message.
>> It's because of version compatibility. I look at review history and to the
>> rev.340 server sent "9999", so in my opinion it should still be handle to 
>> be
>> compatibility with previous server version. 
>> I'm not sure if it should be send from anywhere by server now.
>> 
>> 2. "9988" or "9977"?
>> For now "9988" means "Server Busy: Too many clients waiting in server 
>> queue. 
>> Please try again later". The comparison with max client allowed is at 
>> https://code.google.com/p/ndt/source/browse/trunk/src/web100srv.c#2291 
>> and it's good.
>> 
>> At present version i propose to change some "9988" to "9977" - (Server 
>> fault):
>> https://code.google.com/p/ndt/source/browse/trunk/src/web100srv.c#2030
>> https://code.google.com/p/ndt/source/browse/trunk/src/web100srv.c#2049
>> https://code.google.com/p/ndt/source/browse/trunk/src/web100srv.c#2083
>> 
>> 3. SRV_QUEUE messages in previous review.
>> In other hand, at rev.340 we can see that someone (rcarlson501) try to add
>> some other flag: 9444, 9666 but in rev.341 cancelled it and changed 
>> everything
>> for 9988, what is for now. Maybe someone remember that times and could 
>> explain
>> if it was intentional action.
>> 
>> Rev.340:
>> https://code.google.com/p/ndt/source/detail?r=340&path=/trunk/src/web100srv.c
>> Rev.341: 
>> https://code.google.com/p/ndt/source/detail?r=341&path=/trunk/src/web100srv.c
>> 
>> Wiki about SRV_QUEUE messages is poor:
>> https://code.google.com/p/ndt/wiki/NDTProtocol#Queuing_clients
>> 
>> Do you have any suggestion, ideas or sth? Should we change anything?
>> 
>> Hope, you understand
>> Sebastian
>> 
>