RE: [boost] Latest version of the profiler

From: Brian Braatz (brianb_at_[hidden])
Date: 2005-02-10 15:50:01

> -----Original Message-----
> From: boost-bounces_at_[hidden]
[mailto:boost-bounces_at_[hidden]]
> On Behalf Of christopher diggins
> 
> I have updated the profiler library to take into account many of the
> excellent suggestions. I have tried to give acknowledgements where it
is
> due, please let me know if I have overlooked anyone.
> 
> The latest version is at: http://www.cdiggins.com/profiler
> 
> The major changes are:
> - defining BOOST_PROFILING_OFF causes all member calls to become empty
> statements
> - basic_profiler now has three policies: logging, stats, timer
> - profiler now supports: restart, stop, pause, resume, generate_report
> - fixed QueryPerformanceCounter() api bug in HighResolutionTimer
> 
> If you have brought an issue up previously (or think of a new one)
which
> you
> feel is a significant omission of this library and would affect
whether or
> not you could use it, please feel free to bring it up again.
> 
> I am still looking for a posix high-resolution timer. Hartmut
suggested a
> couple of Posix timers which he got off of Google, which unfortunately
> were
> unusable: one was GPL'ed and the other lacked source code. Gennadiy
said
> there were some on the mailing list (I couldn't find ), and that he
uses
> one
> in the CVS but I don't know how to use the CVS. I also can not test
Posix
> timers, since I only use visual C++. I believe that perhaps
> boost::date_time
> contains a good resolution timer, but I can't tell.
> 
[Brian Braatz] 
Christopher,

        I really like the way this is shaping up. I encourage you onward
:) 

        I am in the thick of some deep work things right now and will
not be able to really play with this for a week or two, but by reading
the docs it looks like it would be very usable for the types of problems
I have with profiling.

        Suggestion \ Comments:
                A suggestion would be to make the TYPE for the name be a
template param. (or maybe a typedef in the policy).
                For my needs, I need to store a structure in this place
and not just a string.
                This goes back to the meta-data discussion awhile back.
I believe the users of the library need to be able to give you the type
to store for the id, this way I can use whatever ID concept I need to,
and it means I can store extra metadata in that (without having to look
it up in a separate map<> or something)