[Cs-dev] how can this happen?
| Date | 2010-08-23 16:25 |
| From | Victor Lazzarini |
| Subject | [Cs-dev] how can this happen? |
*p->hptr = (MYFLT) ((unsigned long)p->handle);
csound->Message(csound, "handle %p => %u => %f => %x \n",
p->handle, (unsigned long) p->handle, *p->hptr,
(unsigned long) *p->hptr);
prints out....
handle 0x183473c8 => 406090696 => 406090688.000000 => 183473c0
why is this assignment being rounded up? (NB: MYFLT => float)
------------------------------------------------------------------------------
This SF.net email is sponsored by
Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net |
| Date | 2010-08-23 16:35 |
| From | Richard Dobson |
| Subject | Re: [Cs-dev] how can this happen? |
On 23/08/2010 16:25, Victor Lazzarini wrote: > *p->hptr = (MYFLT) ((unsigned long)p->handle); > csound->Message(csound, "handle %p => %u => %f => %x \n", > p->handle, (unsigned long) p->handle, *p->hptr, > (unsigned long) *p->hptr); > > prints out.... > > handle 0x183473c8 => 406090696 => 406090688.000000 => 183473c0 > > > why is this assignment being rounded up? (NB: MYFLT => float) > A 32bit float only resolves to seven(-ish) accurate decimal digits, whatever the magnitude. The above maps a large 32bit int to a smaller float with at best 25bit precision. Richard Dobson ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net |
| Date | 2010-08-23 17:12 |
| From | Victor Lazzarini |
| Subject | Re: [Cs-dev] how can this happen? |
yes, I suspected overflow as I typed the e-mail... On 23 Aug 2010, at 16:35, Richard Dobson wrote: > On 23/08/2010 16:25, Victor Lazzarini wrote: >> *p->hptr = (MYFLT) ((unsigned long)p->handle); >> csound->Message(csound, "handle %p => %u => %f => %x \n", >> p->handle, (unsigned long) p->handle, *p->hptr, >> (unsigned long) *p->hptr); >> >> prints out.... >> >> handle 0x183473c8 => 406090696 => 406090688.000000 => 183473c0 >> >> >> why is this assignment being rounded up? (NB: MYFLT => float) >> > A 32bit float only resolves to seven(-ish) accurate decimal digits, > whatever the magnitude. The above maps a large 32bit int to a smaller > float with at best 25bit precision. > > Richard Dobson > > > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by > > Make an app they can't live without > Enter the BlackBerry Developer Challenge > http://p.sf.net/sfu/RIM-dev2dev > _______________________________________________ > Csound-devel mailing list > Csound-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/csound-devel ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net |