The problem with the STL is that the performance of these libraries is
truly horrible. The approach that is currently planned is to attempt to
provide very similar functionality - just with some good performance. It is
expected that the STL will replace these libraries when it grows up.
There are a number of places where these standard, industry-wide standards
just do not live up to the kind of performance that is expected. With each
new technology, we take a significant performance hit.
When we first started using vxWorks we found that their locking semaphores
used 75 usecs even when the semaphore was available. Our rewritten one took
3 usecs when the semaphore was available. It took one small design meeting
about 4 hours to design the code and prove that it was safe. It was worth it.
Bob
- References:
- Let's use the STL! [Was: about the abstractData.h] Benjamin Franksen
- Navigate by Date:
- Prev:
MPF and mpfSerial Marty Kraimer
- Next:
Re: Performance of portable CA server ? Bernhard Kuner
- Index:
1994
1995
1996
1997
1998
1999
<2000>
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
- Navigate by Thread:
- Prev:
Let's use the STL! [Was: about the abstractData.h] Benjamin Franksen
- Next:
Re: Let's use the STL! [Was: about the abstractData.h] Kay Kasemir
- Index:
1994
1995
1996
1997
1998
1999
<2000>
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
|