[vtk-developers] Compi

Bill Lorensen bill.lorensen at gmail.com
Mon Apr 20 15:38:27 EDT 2009


Must depend on the compiler. (send me your test, i have several
compilers available). As I said at least for Borland they should be
equivalent. I guess an #ifdef should be used.

Thanks,

Bill

On Mon, Apr 20, 2009 at 3:26 PM, Philippe P. Pebay <pppebay at sandia.gov> wrote:
> Thanks Bill for having looked into it.
>
> However, it does not resolve the performance issue:
> ## Completed 1000000000 push_back vs. += vs. append operations.
>   Wall times:
>    push_back: 4 sec.
>    +=: 8 sec.
>    append: 8 sec.
>
>
> P.
>
>
>
> Bill Lorensen wrote:
>>
>> Folks,
>>
>> The compile errors for VS6:
>> http://www.cdash.org/CDash/viewBuildError.php?buildid=316621
>>
>> are easily fixed with the portable changes in the code below. I did
>> run all of the InfoVis tests. I did not check in the changes because I
>> can't tell if this code is covered by testing.  The current coverage
>> build:
>> http://www.cdash.org/CDash/index.php?project=VTK&date=2009-04-20
>> does not include InfoVis.
>>
>> FYI: push_back on strings is implemented as append on some stl
>> implementations and clear() is implemented as erase() on Borland.
>>
>> Could an InfoVis person verify that these changes are valid and please
>> check them into cvs?
>>
>> Thanks,
>>
>> Bill
>>
>>
>> void PackValues( const vtkstd::vector<vtkStdString>& values,
>>                 vtkStdString& buffer )
>> {
>>  buffer.erase(); // equivalent to buffer.clear()
>>
>>  for( vtkstd::vector<vtkStdString>::const_iterator it = values.begin();
>>       it != values.end(); ++ it )
>>    {
>>    buffer.append( *it );
>>    buffer.append(1, 0); // equilvalent to buffer.push_back( 0 )
>>    }
>> }
>> _______________________________________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/listinfo/vtk-developers
>>
>>
>
>
> --
> Philippe Pébay
> Sandia National Laboratories
>
>
>



More information about the vtk-developers mailing list