Hi,
FUD is good if it leads to improvments ;)

indices never match ... as compare and swap is atomic and pretty fast, the way a
starvation may occure for a given thread is that he never has a CPU ...
or for less cycles than he needs to loop twice or thrice in the CAS loop.

It also means that your application is doing nothing else than logging
;)

I've already almost done everything about the logging and object stuff.
Many thanks to vim and sed ...

I'm now working on the filesystem handling basing everything on
QtCore/QDir QtCore/QFile etc, and putting everything in a filesystem
small abstraction. I want to end with no path and file relative functions
elsewhere than in Filesystem class, mostly static methods.

see http://cgit.asynk.ch/cgi-bin/cgit/hydrogen/ branch unstable

( haven't pushed the Filesystem class yet )

Stress tests ... great !! I usually write a lot of tests cases when I'm coding 
in
ruby, but in C ... I find it terribly boring ;((

I'll test it tomorrow.

Regards

On Monday 04 January 2010  09:52, Gabriel M. Beddingfield wrote :
> 
> 
> On Wed, 30 Dec 2009, Jérémy Zurcher wrote:
> 
> >damn fast and seems to be reliable enough ...
> 
> Sorry for the delay...
> 
> Looks nice... though I have some doubts about the reliability of the
> way you're doing __read_from and __write_to manipulations... but
> that's just good old-fashioned FUD talk.  :-)  The biggest
> reservation is this:  what if the indices never match because
> there's always messages being written to the buffer?
> 
> But otherwise it looks nice.  Now you just need to go through and
> replace stuff like this:
> 
>    INFOLOG( QString( "Pan: %1L, %2R" )
>             .arg(pan_L)
>             .arg(pan_R) );
> 
> With stuff like this:
> 
>    INFOLOG( "Pan: %.4gL %.4gR", pan_L, pan_R );
> 
> In the 842 places that it occurs.  :-)
> 
> FYI, some dudes a year or two ago wrote a ring-buffer "stress test."
> Wouldn't be bad to use it to test your class.  The URL is here:
> 
>     http://svn.samalyse.com/misc/rbtest/ [1]
> 
> Peace,
> Gabriel
> 
> [1] Ref. http://lalists.stanford.edu/lad/2008/10/0093.html
> 
> !DSPAM:4b420fac71381727237172!

> ------------------------------------------------------------------------------
> This SF.Net email is sponsored by the Verizon Developer Community
> Take advantage of Verizon's best-in-class app development support
> A streamlined, 14 day to market process makes app distribution fast and easy
> Join now and get one step closer to millions of Verizon customers
> http://p.sf.net/sfu/verizon-dev2dev 
> 
> !DSPAM:4b420fac71381727237172!

> _______________________________________________
> Hydrogen-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/hydrogen-devel
> 
> 
> !DSPAM:4b420fac71381727237172!


-- 
Jérémy Zurcher
rte de Cully 29
1091 Grandvaux
+41 (0) 79 599 84 27


------------------------------------------------------------------------------
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
_______________________________________________
Hydrogen-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/hydrogen-devel

Reply via email to