[Interest] Packet arrival-time resolution? QUdpSocket

[Konstantin Tokarev]

01.12.2017, 13:28, "Konrad Rosenbaum" <konrad at silmor.de>:
> Hi,
>
> On Thu, November 30, 2017 16:49, Jason H wrote:
>>  I don't think you can say that Qt is not real-time. That's more of a
>>  kernel issue than a software library issue.
>
> On Thu, November 30, 2017 17:43, Konstantin Tokarev wrote:
>>  Actually, Qt is ported to real-time operating systems, such as QNX and
>>  Integrity
>
> This is a common misconception. Realtime is not some magic forest that
> will make everything realtime as soon as you enter it. All layers of hard-
> and software have to be realtime in order for the result to be considered
> realtime too. A simple use of new or malloc can destroy your realtime
> guarantees because memory allocation is not fully deterministic

I didn't state the opposite, just commented that on kernel side of things
there is enough flexibility already.

>
> Luckily we are talking about "soft" realtime with multimedia applications
> where the occasional deviation is not catastrophic and will likely not
> even be noticed by the users. Just make sure:
>
> a) all threads handling time and media are scheduled as realtime threads
> b) if at all possible make it ONE thread with all the critical stuff
> c) keep the code in that critical thread as short and fast as possible
> d) reduce dynamic object allocations during runtime to the absolute
> minimum - if possible allocate everything you need before you start
> e) try to work lock free (otherwise you'll have regular priority
> inversions; those are real fun!)
>
>   Konrad
>
> _______________________________________________
> Interest mailing list
> Interest at qt-project.org
> http://lists.qt-project.org/mailman/listinfo/interest

-- 
Regards,
Konstantin