How is it expected to work ? I see code that would attempt to hold
the previous output value when the trigger is not on, but it is not
implemented correctly, so the opcode outputs zero instead. So, should
holding the previous output be fixed, or should the existing code that
outputs zero when not triggered be kept (in this case the non-functional
parts could be removed) ? The manual does not define the behavior of
the opcode in this respect at all.
Note that all other opcodes in pycall.c.auto are affected as well,
but those in pyx.c.auto are apparently not.

I ran into the same problem/bug/feature when using pycall1t.
I used a goto to work around it, like this:

;*******
kUpdate		metro	1
if kUpdate == 0 kgoto end	; necessary hack to prevent the globals being overwritten by a zero between updates

kArgument	= 0		; bogus argument

; get date and time from python, write to csound globals
gkYear	pycall1t	kUpdate, "dateAndTime.getYear", kArgument
gkMonth	pycall1t	kUpdate, "dateAndTime.getMonth", kArgument
gkMDay	pycall1t	kUpdate, "dateAndTime.getMonthDay", kArgument
gkWDay	pycall1t	kUpdate, "dateAndTime.getWeekDay", kArgument
gkHour	pycall1t	kUpdate, "dateAndTime.getHour", kArgument
gkMinute	pycall1t	kUpdate, "dateAndTime.getMinute", kArgument
gkSecond	pycall1t	kUpdate, "dateAndTime.getSecond", kArgument

end:
;*******
Oeyvind




> From: Istvan Varga [istvan_v@fibermail.hu]
> Sent: 2005-10-16 13:01:23 CEST
> To: csound@lists.bath.ac.uk
> Subject: [Csnd] [Cs-dev] pycall1t trigger argument
> 
> How is it expected to work ? I see code that would attempt to hold
> the previous output value when the trigger is not on, but it is not
> implemented correctly, so the opcode outputs zero instead. So, should
> holding the previous output be fixed, or should the existing code that
> outputs zero when not triggered be kept (in this case the non-functional
> parts could be removed) ? The manual does not define the behavior of
> the opcode in this respect at all.
> Note that all other opcodes in pycall.c.auto are affected as well,
> but those in pyx.c.auto are apparently not.
> 
> -- 
> Send bugs reports to this list.
> To unsubscribe, send email to csound-unsubscribe@lists.bath.ac.uk
> 

So, how should the opcodes actually behave when the trigger is off ?

   1. output zero
   2. do not write to output variable
   3. hold previous output, not clearing internal state on
      note initialization
   4. hold previous output, which is cleared to zero on
      note initialization

Oeyvind Brandtsegg wrote:

> I ran into the same problem/bug/feature when using pycall1t.
> I used a goto to work around it, like this:
> 
> ;*******
> kUpdate		metro	1
> if kUpdate == 0 kgoto end	; necessary hack to prevent the globals being overwritten by a zero between updates
> 
> kArgument	= 0		; bogus argument
> 
> ; get date and time from python, write to csound globals
> gkYear	pycall1t	kUpdate, "dateAndTime.getYear", kArgument
> gkMonth	pycall1t	kUpdate, "dateAndTime.getMonth", kArgument
> gkMDay	pycall1t	kUpdate, "dateAndTime.getMonthDay", kArgument
> gkWDay	pycall1t	kUpdate, "dateAndTime.getWeekDay", kArgument
> gkHour	pycall1t	kUpdate, "dateAndTime.getHour", kArgument
> gkMinute	pycall1t	kUpdate, "dateAndTime.getMinute", kArgument
> gkSecond	pycall1t	kUpdate, "dateAndTime.getSecond", kArgument
> 
> end:
> ;*******
> Oeyvind

I think it would be nice if it works like alternative 4
>    4. hold previous output, which is cleared to zero on
>       note initialization

Oeyvind




> From: Istvan Varga [istvan_v@fibermail.hu]
> Sent: 2005-10-17 14:04:23 CEST
> To: csound@lists.bath.ac.uk
> Subject: Re: [Csnd] [Cs-dev] pycall1t trigger argument
> 
> So, how should the opcodes actually behave when the trigger is off ?
> 
>    1. output zero
>    2. do not write to output variable
>    3. hold previous output, not clearing internal state on
>       note initialization
>    4. hold previous output, which is cleared to zero on
>       note initialization
> 
> Oeyvind Brandtsegg wrote:
> 
> > I ran into the same problem/bug/feature when using pycall1t.
> > I used a goto to work around it, like this:
> > 
> > ;*******
> > kUpdate		metro	1
> > if kUpdate == 0 kgoto end	; necessary hack to prevent the globals being overwritten by a zero between updates
> > 
> > kArgument	= 0		; bogus argument
> > 
> > ; get date and time from python, write to csound globals
> > gkYear	pycall1t	kUpdate, "dateAndTime.getYear", kArgument
> > gkMonth	pycall1t	kUpdate, "dateAndTime.getMonth", kArgument
> > gkMDay	pycall1t	kUpdate, "dateAndTime.getMonthDay", kArgument
> > gkWDay	pycall1t	kUpdate, "dateAndTime.getWeekDay", kArgument
> > gkHour	pycall1t	kUpdate, "dateAndTime.getHour", kArgument
> > gkMinute	pycall1t	kUpdate, "dateAndTime.getMinute", kArgument
> > gkSecond	pycall1t	kUpdate, "dateAndTime.getSecond", kArgument
> > 
> > end:
> > ;*******
> > Oeyvind
> -- 
> Send bugs reports to this list.
> To unsubscribe, send email to csound-unsubscribe@lists.bath.ac.uk
>