From: jerry+a2@jpen.ca   
      
   Tom Thumb  writes:   
      
   > On Thursday, June 8, 2023 at 8:59:21 PM UTC-4, Tom Thumb wrote:   
   >> On Thursday, June 8, 2023 at 7:17:06 AM UTC-4, Dirk Froehling wrote:   
   >> > Am 02.06.23 um 19:15 schrieb Tom Thumb:   
   >> > > On Thursday, June 1, 2023 at 8:33:08 PM UTC-4, Tom Thumb wrote:   
   >> > >> On Thursday, June 1, 2023 at 2:05:06 PM UTC-4, stephen...@gmail.com   
   wrote:   
   >> > >>> On Thursday, June 1, 2023 at 11:25:59 AM UTC-5, Tom Thumb wrote:   
   >> > >>>> On Wednesday, May 31, 2023 at 3:17:30 PM UTC-4, Tom Thumb wrote:   
   >> > >>>>> Mike Westerfield's "Toolbox Programming in C" contains a routine:   
   >> > >>>>>   
   >> > >>>>> void HandleControl (void)   
   >> > >>>>> {   
   >> > >>>>> if ((myEvent.wmTaskData4 & 0xFFFF8000) == 0)   
   >> > >>>>> switch (myEvent.wmTaskData4) {   
   >> > >>>>>   
   >> > >>>>> ...   
   >> > >>>>>   
   >> > >>>>> I can find no explanation for doing this. I believe it's just   
   testing if any of   
   >> > >>>>> the bits 17 - 31 have been set to one?   
   >> > >>>>>   
   >> > >>>>> I'd be grateful if someone would explain why this is being done. I   
   imagine I've   
   >> > >>>>> missed something there or the IIgs Reference Manuals.   
   >> > >>>> Oh, it's checking to see if wmTaskData4 is less than 32768   
   >> > >>> Yes, that's correct. In this case wmTaskData4 holds a control ID   
   value, and all   
   >> > >>> the control IDs that the code wants to check for are less than 32768.   
   The "if"   
   >> > >>> check is OK, but it's not really necessary, since the code could just   
   check for   
   >> > >>> the control IDs of interest via the "switch".   
   >> > >>>   
   >> > >>> I suspect the "if" check is there because the code was originally   
   written in   
   >> > >>> Pascal for the Toolbox Programming in Pascal course and then later   
   translated to   
   >> > >>> C. The "case" statement in ORCA/Pascal can only handle values up to   
   32767, making   
   >> > >>> a check like this necessary, but the "switch" statement in C does not   
   have the   
   >> > >>> same limitation.   
   >> > >>>   
   >> > >>> --   
   >> > >>> Stephen Heumann   
   >> > >> Ahh, thank you very much.   
   >> > >   
   >> > > May I ask though; why go about it this way: if ((myEvent.wmTaskData4 &   
   0xFFFF8000)   
   >> > > == 0) rather than: if(myEvent.wmTaskData4 <= 32767) ?   
   >> > >   
   >> > > is there a reason? or just a choice?   
   >> > >   
   >> > It is easier to read. :)   
   >> >   
   >> > No, really. 32767 is meaningless in that context. wmTaskData4 contains a   
   >> > bunch of seperate settings coded as bits, not a single number.   
   >> >   
   >> > 0xFFFF8000 is 1111 1111 1111 1111 1000 0000 0000 0000 in binary form,   
   >> > and by ANDing wmTaskData4 with this number, you say "I am only   
   >> > interested in the upper 17 bits, ignore the lower bits."   
   >> > The IF condition then means "If none of the settings in the upper 17   
   >> > bits is set, do the following."   
   >> >   
   >> > Dirk   
   >> I get the ANDing to test if bits are set and that's what I thought first   
   too but as far   
   >> as I've read the only thing wmTaskData4 ever contains is a zero or a   
   control id; it,   
   >> wmTaskData4, isn't flags bits. Only after playing ANDing that I stumbled   
   upon   
   >> 00000000000000001111111111111111, 0x7FFF, 32768 triggers that condition   
      
   >>   
   >> Was just wondering why.   
   >   
   > He has a sense of humor. I think he was just messing with me.   
      
   Hmmm, I don't think he was messing with you.  The hex-style is a fairly   
   typical way to handle stuff like this.  It's common in microcontroller   
   code.   
      
      
   --   
   --   
   Jerry     jerry+a2 at jpen.ca   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)