unicorn team mailing list archive

[Johan Jansson <jjan@xxxxxxxxxx>]

On Tue, Jun 28, 2011 at 7:55 PM, Niclas Jansson <njansson@xxxxxxxxxx> wrote:
> Johan Jansson <jjan@xxxxxxxxxx> writes:
>
>> How about just appending to the file, and then cleaning up the overlap
>> after the simulation has completed? We could write a simple Python
>> tool which cuts away all lines with times which are not strictly
>> increasing.
>>
>> Best,
>>   Johan
>>
>
> True, we can even do it on the fly. For example, post pone drag
> computation until the simulation time is larger then the last sample
> in aero_f.dat
>
> While on the subject, we need to rethink aero_f.dat, for example
> lustre is not about the small amount of data written every timestep.
>
>
> Niclas
>

I think the culprit here is the:

        forceFile.flush();

statement. If we remove it then the output should be buffered (it's
also possible to set the buffer size if necessary). I think that
should take care of it or?

Best,
  Johan