[Csnd] array size question

[Csnd] array size question

Date	2019-03-25 12:30
From	Eduardo Moguillansky
Subject	[Csnd] array size question
	What would you expect the first printf to be? kA[] fillarray 0, 1, 2, 3, 4, 5 printarray kA, 1 printf "kSize 1 = %d \n", 1, lenarray:k(kA) kB[] fillarray 0, 1, 2, 3 kA = kB * 2 printarray kA, 1 printf "kSize 2 = %d \n", 1, lenarray:k(kA) turnoff I would have assumed that it would print 6, but it prints 4. What actually seems to happen right now is that the assignment kA = kB * 2 has an init-time function which ensures the size of kA to be the size of kB at init time, which makes the size of kA to be wrong for the first printf. I personally think that this i-time behaviour should be removed and checks should be made at each k-pass. At the very least, this should be an error. What do you think? best eduardo Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-25 12:35
From	john
Subject	Re: [Csnd] array size question
	There was an explicit request for the current behaviour so this would need a major discussion. On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: > What would you expect the first printf to be? > > kA[] fillarray 0, 1, 2, 3, 4, 5 > printarray kA, 1 > printf "kSize 1 = %d \n", 1, lenarray:k(kA) > > kB[] fillarray 0, 1, 2, 3 > kA = kB * 2 > printarray kA, 1 > printf "kSize 2 = %d \n", 1, lenarray:k(kA) > turnoff > > I would have assumed that it would print 6, but it prints 4. What actually > seems to happen right now is that the assignment kA = kB * 2 has an init-time > function which ensures the size of kA to be the size of kB at init time, > which makes the size of kA to be wrong for the first printf. I personally > think that this i-time behaviour should be removed and checks should be made > at each k-pass. At the very least, this should be an error. What do you > think? > > best > eduardo > > Csound mailing list > Csound@listserv.heanet.ie > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here > Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-26 09:18
From	Eduardo Moguillansky
Subject	Re: [Csnd] array size question
	To be clear on this issue, I am not saying that the assignment "kA = kB * 2" should not change the perceived size of the array, but that this change happens at init time, so that any k-time operation on the array is performed on the wrong size. In my opinion the correct and expected behaviour is to call tabensure at every k-pass, only resizing upwards if necessary. In this way, the array would have the expected behaviour for the first printf and still be correct for the second. On Mo, Mar 25, 2019 at 1:35 PM, john wrote: > There was an explicit request for the current behaviour so this would > need a major discussion. > > On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: > >> What would you expect the first printf to be? >> >> kA[] fillarray 0, 1, 2, 3, 4, 5 >> printarray kA, 1 >> printf "kSize 1 = %d \n", 1, lenarray:k(kA) >> >> kB[] fillarray 0, 1, 2, 3 >> kA = kB * 2 >> printarray kA, 1 >> printf "kSize 2 = %d \n", 1, lenarray:k(kA) >> turnoff >> >> I would have assumed that it would print 6, but it prints 4. What >> actually seems to happen right now is that the assignment kA = kB * >> 2 has an init-time function which ensures the size of kA to be the >> size of kB at init time, which makes the size of kA to be wrong >> for the first printf. I personally think that this i-time behaviour >> should be removed and checks should be made at each k-pass. At the >> very least, this should be an error. What do you think? >> >> best >> eduardo >> >> Csound mailing list >> Csound@listserv.heanet.ie >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> Send bugs reports to >> https://github.com/csound/csound/issues >> Discussions of bugs and features can be posted here >> > > Csound mailing list > Csound@listserv.heanet.ie > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-26 10:40
From	Rory Walsh
Subject	Re: [Csnd] array size question
	if necessary. In this way, the array would have the expected behaviour for the first printf and still be correct for the second. I think this is what most user would expect? I certainly failed to answer the first question Eduardo posed. @Joachim Heintz what are your thoughts on this? On Mo, Mar 25, 2019 at 1:35 PM, john <jpff@CODEMIST.CO.UK> wrote: > There was an explicit request for the current behaviour so this would > need a major discussion. > > On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: > >> What would you expect the first printf to be? >> >> kA[] fillarray 0, 1, 2, 3, 4, 5 >> printarray kA, 1 >> printf "kSize 1 = %d \n", 1, lenarray:k(kA) >> >> kB[] fillarray 0, 1, 2, 3 >> kA = kB * 2 >> printarray kA, 1 >> printf "kSize 2 = %d \n", 1, lenarray:k(kA) >> turnoff >> >> I would have assumed that it would print 6, but it prints 4. What >> actually seems to happen right now is that the assignment kA = kB * >> 2 has an init-time function which ensures the size of kA to be the >> size of kB at init time, which makes the size of kA to be wrong >> for the first printf. I personally think that this i-time behaviour >> should be removed and checks should be made at each k-pass. At the >> very least, this should be an error. What do you think? >> >> best >> eduardo >> >> Csound mailing list >> Csound@listserv.heanet.ie >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> Send bugs reports to >> https://github.com/csound/csound/issues >> Discussions of bugs and features can be posted here >> > > Csound mailing list > Csound@listserv.heanet.ie > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-26 22:02
From	joachim heintz
Subject	Re: [Csnd] array size question
	as far as i understand, it is the ambiguity of the = operator (whether it refers to i- or k-rate) which makes things complcated here. for strings, we have strcpy and strcpyk to distinguish. perhaps we need something similar here? On 26/03/19 11:40, Rory Walsh wrote: > if necessary. In this way, the array would have the expected behaviour > for the first printf and still be correct for the second. > > > I think this is what most user would expect? I certainly failed to > answer the first question Eduardo posed. @Joachim Heintz > what are your thoughts on this? > > > > > > On Mo, Mar 25, 2019 at 1:35 PM, john > wrote: > > There was an explicit request for the current behaviour so this would > > need a major discussion. > > > > On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: > > > >> What would you expect the first printf to be? > >> > >> kA[] fillarray 0, 1, 2, 3, 4, 5 > >> printarray kA, 1 > >> printf "kSize 1 = %d \n", 1, lenarray:k(kA) > >> > >> kB[] fillarray 0, 1, 2, 3 > >> kA = kB * 2 > >> printarray kA, 1 > >> printf "kSize 2 = %d \n", 1, lenarray:k(kA) > >> turnoff > >> > >> I would have assumed that it would print 6, but it prints 4. What > >> actually seems to happen right now is that the assignment kA = kB * > >> 2 has an init-time function which ensures the size of kA to be the > >> size of kB at init time, which makes the size of kA to be wrong > >> for the first printf. I personally think that this i-time behaviour > >> should be removed and checks should be made at each k-pass. At the > >> very least, this should be an error. What do you think? > >> > >> best > >> eduardo > >> > >> Csound mailing list > >> Csound@listserv.heanet.ie > >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > >> Send bugs reports to > >> https://github.com/csound/csound/issues > >> Discussions of bugs and features can be posted here > >> > > > > Csound mailing list > > Csound@listserv.heanet.ie > > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > > Send bugs reports to > > https://github.com/csound/csound/issues > > Discussions of bugs and features can be posted here > > Csound mailing list > Csound@listserv.heanet.ie > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here > > Csound mailing list Csound@listserv.heanet.ie > > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to > https://github.com/csound/csound/issues Discussions of bugs and features > can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-26 22:15
From	Eduardo Moguillansky
Subject	Re: [Csnd] array size question
	In this case there is no ambiguity, since k[] * constant could never be an i-rate operation. But since it was mentioned that the current behaviour was explicitely asked for, could someone point to a case where this behaviour is the desired and expected one? On Di, Mar 26, 2019 at 11:02 PM, joachim heintz wrote: > as far as i understand, it is the ambiguity of the = operator > (whether it refers to i- or k-rate) which makes things complcated > here. > > for strings, we have strcpy and strcpyk to distinguish. perhaps we > need something similar here? > > > > On 26/03/19 11:40, Rory Walsh wrote: >> if necessary. In this way, the array would have the expected >> behaviour >> for the first printf and still be correct for the second. >> >> >> I think this is what most user would expect? I certainly failed to >> answer the first question Eduardo posed. @Joachim Heintz >> what are your thoughts on this? >> >> >> >> >> >> On Mo, Mar 25, 2019 at 1:35 PM, john > > wrote: >> > There was an explicit request for the current behaviour so >> this would >> > need a major discussion. >> > >> > On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: >> > >> >> What would you expect the first printf to be? >> >> >> >> kA[] fillarray 0, 1, 2, 3, 4, 5 >> >> printarray kA, 1 >> >> printf "kSize 1 = %d \n", 1, lenarray:k(kA) >> >> >> >> kB[] fillarray 0, 1, 2, 3 >> >> kA = kB * 2 >> >> printarray kA, 1 >> >> printf "kSize 2 = %d \n", 1, lenarray:k(kA) >> >> turnoff >> >> >> >> I would have assumed that it would print 6, but it prints 4. >> What >> >> actually seems to happen right now is that the assignment kA >> = kB * >> >> 2 has an init-time function which ensures the size of kA to >> be the >> >> size of kB at init time, which makes the size of kA to >> be wrong >> >> for the first printf. I personally think that this i-time >> behaviour >> >> should be removed and checks should be made at each k-pass. >> At the >> >> very least, this should be an error. What do you think? >> >> >> >> best >> >> eduardo >> >> >> >> Csound mailing list >> >> Csound@listserv.heanet.ie >> >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> >> Send bugs reports to >> >> https://github.com/csound/csound/issues >> >> Discussions of bugs and features can be posted here >> >> >> > >> > Csound mailing list >> > Csound@listserv.heanet.ie >> > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> > Send bugs reports to >> > https://github.com/csound/csound/issues >> > Discussions of bugs and features can be posted here >> >> Csound mailing list >> Csound@listserv.heanet.ie >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> Send bugs reports to >> https://github.com/csound/csound/issues >> Discussions of bugs and features can be posted here >> >> Csound mailing list Csound@listserv.heanet.ie >> >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to >> https://github.com/csound/csound/issues Discussions of bugs and >> features >> can be posted here > > Csound mailing list > Csound@listserv.heanet.ie > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-27 09:59
From	joachim heintz
Subject	Re: [Csnd] array size question
	i am not that deep in it but perhaps john was referring to this: https://github.com/csound/csound/issues/1039 On 26/03/19 23:15, Eduardo Moguillansky wrote: > In this case there is no ambiguity, since k[] * constant could never be > an i-rate operation. But since it was mentioned that the current > behaviour was explicitely asked for, could someone point to a case where > this behaviour is the desired and expected one? > > On Di, Mar 26, 2019 at 11:02 PM, joachim heintz > wrote: >> as far as i understand, it is the ambiguity of the = operator (whether >> it refers to i- or k-rate) which makes things complcated here. >> >> for strings, we have strcpy and strcpyk to distinguish. perhaps we >> need something similar here? >> >> >> >> On 26/03/19 11:40, Rory Walsh wrote: >>> if necessary. In this way, the array would have the expected >>> behaviour >>> for the first printf and still be correct for the second. >>> >>> >>> I think this is what most user would expect? I certainly failed to >>> answer the first question Eduardo posed. @Joachim Heintz >>> what are your thoughts on this? >>> >>> >>> >>> >>> >>> On Mo, Mar 25, 2019 at 1:35 PM, john >> > wrote: >>> > There was an explicit request for the current behaviour so this >>> would >>> > need a major discussion. >>> > >>> > On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: >>> > >>> >> What would you expect the first printf to be? >>> >> >>> >> kA[] fillarray 0, 1, 2, 3, 4, 5 >>> >> printarray kA, 1 >>> >> printf "kSize 1 = %d \n", 1, lenarray:k(kA) >>> >> >>> >> kB[] fillarray 0, 1, 2, 3 >>> >> kA = kB * 2 >>> >> printarray kA, 1 >>> >> printf "kSize 2 = %d \n", 1, lenarray:k(kA) >>> >> turnoff >>> >> >>> >> I would have assumed that it would print 6, but it prints 4. What >>> >> actually seems to happen right now is that the assignment kA >>> = kB * >>> >> 2 has an init-time function which ensures the size of kA to >>> be the >>> >> size of kB at init time, which makes the size of kA to be >>> wrong >>> >> for the first printf. I personally think that this i-time >>> behaviour >>> >> should be removed and checks should be made at each k-pass. >>> At the >>> >> very least, this should be an error. What do you think? >>> >> >>> >> best >>> >> eduardo >>> >> >>> >> Csound mailing list >>> >> Csound@listserv.heanet.ie >>> >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >>> >> Send bugs reports to >>> >> https://github.com/csound/csound/issues >>> >> Discussions of bugs and features can be posted here >>> >> >>> > >>> > Csound mailing list >>> > Csound@listserv.heanet.ie >>> > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >>> > Send bugs reports to >>> > https://github.com/csound/csound/issues >>> > Discussions of bugs and features can be posted here >>> >>> Csound mailing list >>> Csound@listserv.heanet.ie >>> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >>> Send bugs reports to >>> https://github.com/csound/csound/issues >>> Discussions of bugs and features can be posted here >>> >>> Csound mailing list Csound@listserv.heanet.ie >>> >>> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to >>> https://github.com/csound/csound/issues Discussions of bugs and features >>> can be posted here >> >> Csound mailing list >> Csound@listserv.heanet.ie >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> Send bugs reports to >> https://github.com/csound/csound/issues >> Discussions of bugs and features can be posted here > > Csound mailing list > Csound@listserv.heanet.ie > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here > Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-27 10:30
From	Eduardo Moguillansky
Subject	Re: [Csnd] array size question
	That is a very long discussion... But what I inderstood from skimming through it is that the critique was actually the same as here: an array action that is supposed to act at k-rate has an init component which collides with other init components (presumably to avoid having to do allocations at each k-pass). But since an array has an allocated and an apparent size, leaving the init part out would solve this and the situation mentioned in that discussion. At each k-pass the array's apparent size would be set to the needed size, growing it if necessary. On Mi, Mar 27, 2019 at 10:59 AM, joachim heintz wrote: > i am not that deep in it but perhaps john was referring to this: > https://github.com/csound/csound/issues/1039 > > > > On 26/03/19 23:15, Eduardo Moguillansky wrote: >> In this case there is no ambiguity, since k[] * constant could never >> be >> an i-rate operation. But since it was mentioned that the current >> behaviour was explicitely asked for, could someone point to a case >> where >> this behaviour is the desired and expected one? >> >> On Di, Mar 26, 2019 at 11:02 PM, joachim heintz >> wrote: >>> as far as i understand, it is the ambiguity of the = operator >>> (whether >>> it refers to i- or k-rate) which makes things complcated here. >>> >>> for strings, we have strcpy and strcpyk to distinguish. perhaps we >>> need something similar here? >>> >>> >>> >>> On 26/03/19 11:40, Rory Walsh wrote: >>>> if necessary. In this way, the array would have the expected >>>> behaviour >>>> for the first printf and still be correct for the second. >>>> >>>> >>>> I think this is what most user would expect? I certainly failed to >>>> answer the first question Eduardo posed. @Joachim Heintz >>>> what are your thoughts on this? >>>> >>>> >>>> >>>> >>>> >>>> On Mo, Mar 25, 2019 at 1:35 PM, john >>> > wrote: >>>> > There was an explicit request for the current behaviour so >>>> this >>>> would >>>> > need a major discussion. >>>> > >>>> > On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: >>>> > >>>> >> What would you expect the first printf to be? >>>> >> >>>> >> kA[] fillarray 0, 1, 2, 3, 4, 5 >>>> >> printarray kA, 1 >>>> >> printf "kSize 1 = %d \n", 1, lenarray:k(kA) >>>> >> >>>> >> kB[] fillarray 0, 1, 2, 3 >>>> >> kA = kB * 2 >>>> >> printarray kA, 1 >>>> >> printf "kSize 2 = %d \n", 1, lenarray:k(kA) >>>> >> turnoff >>>> >> >>>> >> I would have assumed that it would print 6, but it prints >>>> 4. What >>>> >> actually seems to happen right now is that the assignment >>>> kA >>>> = kB * >>>> >> 2 has an init-time function which ensures the size of kA to >>>> be the >>>> >> size of kB at init time, which makes the size of kA to >>>> be >>>> wrong >>>> >> for the first printf. I personally think that this i-time >>>> behaviour >>>> >> should be removed and checks should be made at each k-pass. >>>> At the >>>> >> very least, this should be an error. What do you think? >>>> >> >>>> >> best >>>> >> eduardo >>>> >> >>>> >> Csound mailing list >>>> >> Csound@listserv.heanet.ie >>>> >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >>>> >> Send bugs reports to >>>> >> https://github.com/csound/csound/issues >>>> >> Discussions of bugs and features can be posted here >>>> >> >>>> > >>>> > Csound mailing list >>>> > Csound@listserv.heanet.ie >>>> > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >>>> > Send bugs reports to >>>> > https://github.com/csound/csound/issues >>>> > Discussions of bugs and features can be posted here >>>> >>>> Csound mailing list >>>> Csound@listserv.heanet.ie >>>> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >>>> Send bugs reports to >>>> https://github.com/csound/csound/issues >>>> Discussions of bugs and features can be posted here >>>> >>>> Csound mailing list Csound@listserv.heanet.ie >>>> >>>> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports >>>> to >>>> https://github.com/csound/csound/issues Discussions of bugs and >>>> features >>>> can be posted here >>> >>> Csound mailing list >>> Csound@listserv.heanet.ie >>> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >>> Send bugs reports to >>> https://github.com/csound/csound/issues >>> Discussions of bugs and features can be posted here >> >> Csound mailing list >> Csound@listserv.heanet.ie >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> Send bugs reports to >> https://github.com/csound/csound/issues >> Discussions of bugs and features can be posted here >> > > Csound mailing list > Csound@listserv.heanet.ie > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here

Date	2019-03-27 17:50
From	Mauro Giubileo
Subject	Re: [Csnd] array size question
	I'm the user who opened that issue on github and it was solved by making slicearray:k[] work only at k-time. Yes, it appears to be a similar problem... In your example, the user expects that an array assignment operation with a k-array output would run only at k-time, so it's very confusing that something like "kA = kB" can change the kA array size at i-time too. Let's consider this: kA[] fillarray 0, 1, 2, 3, 4, 5 ; i-time printk 0, lenarray:k(kA) ; k-time kB[] fillarray 0, 1, 2, 3 ; i-time kA = kB ; i-time (UNEXPECTED!) AND k-time printk 0, lenarray:k(kA) ; k-time turnoff The printed results will be 4 for both the printk (instead of 6 and 4) because the "kA = kB" unexpectedly runs at i-time too... I don't know how it's implemented the "=" operation in csound, but maybe in this case too it could be solved by running the assignment opcode for k-rate output arrays only at k-time. Regards, --- Mauro Il 2019-03-27 11:30 Eduardo Moguillansky ha scritto: That is a very long discussion... But what I inderstood from skimming through it is that the critique was actually the same as here: an array action that is supposed to act at k-rate has an init component which collides with other init components (presumably to avoid having to do allocations at each k-pass). But since an array has an allocated and an apparent size, leaving the init part out would solve this and the situation mentioned in that discussion. At each k-pass the array's apparent size would be set to the needed size, growing it if necessary. On Mi, Mar 27, 2019 at 10:59 AM, joachim heintz <jh@JOACHIMHEINTZ.DE> wrote: i am not that deep in it but perhaps john was referring to this: https://github.com/csound/csound/issues/1039 On 26/03/19 23:15, Eduardo Moguillansky wrote: In this case there is no ambiguity, since k[] * constant could never be an i-rate operation. But since it was mentioned that the current behaviour was explicitely asked for, could someone point to a case where this behaviour is the desired and expected one? On Di, Mar 26, 2019 at 11:02 PM, joachim heintz <jh@JOACHIMHEINTZ.DE> wrote: as far as i understand, it is the ambiguity of the = operator (whether it refers to i- or k-rate) which makes things complcated here. for strings, we have strcpy and strcpyk to distinguish. perhaps we need something similar here? On 26/03/19 11:40, Rory Walsh wrote: if necessary. In this way, the array would have the expected behaviour for the first printf and still be correct for the second. I think this is what most user would expect? I certainly failed to answer the first question Eduardo posed. @Joachim Heintz <mailto:jh@joachimheintz.de> what are your thoughts on this? On Mo, Mar 25, 2019 at 1:35 PM, john <jpff@CODEMIST.CO.UK <mailto:jpff@CODEMIST.CO.UK>> wrote: > There was an explicit request for the current behaviour so this would > need a major discussion. > > On Mon, 25 Mar 2019, Eduardo Moguillansky wrote: > >> What would you expect the first printf to be? >> >> kA[] fillarray 0, 1, 2, 3, 4, 5 >> printarray kA, 1 >> printf "kSize 1 = %d \n", 1, lenarray:k(kA) >> >> kB[] fillarray 0, 1, 2, 3 >> kA = kB * 2 >> printarray kA, 1 >> printf "kSize 2 = %d \n", 1, lenarray:k(kA) >> turnoff >> >> I would have assumed that it would print 6, but it prints 4. What >> actually seems to happen right now is that the assignment kA = kB * >> 2 has an init-time function which ensures the size of kA to be the >> size of kB at init time, which makes the size of kA to be wrong >> for the first printf. I personally think that this i-time behaviour >> should be removed and checks should be made at each k-pass. At the >> very least, this should be an error. What do you think? >> >> best >> eduardo >> >> Csound mailing list >> Csound@listserv.heanet.ie <mailto:Csound@listserv.heanet.ie> >> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND >> Send bugs reports to >> https://github.com/csound/csound/issues >> Discussions of bugs and features can be posted here >> > > Csound mailing list > Csound@listserv.heanet.ie <mailto:Csound@listserv.heanet.ie> > https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND > Send bugs reports to > https://github.com/csound/csound/issues > Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie <mailto:Csound@listserv.heanet.ie> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie <mailto:Csound@listserv.heanet.ie> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here Csound mailing list Csound@listserv.heanet.ie https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND Send bugs reports to https://github.com/csound/csound/issues Discussions of bugs and features can be posted here