Work on PR1848 has brought up one unresolved issue that has been present since we mainstreamed non-pow-two tables.

There are two problems:

1) There is currently no way to tell Csound to create a table with an extended guard point if we do not use a pow-two size.

2) Pow-two+1 sizes are pow-two lengths with extended guard point. There is no way to actually create a table with that size.

A solution for 1) is to find a way to indicate an extended guard point using some form of code (like we do to suppress rescaling)

We can't use negative table numbers because these were used in the past to allow np2 tables.

We can't use negative table numbers because these are used for internal tables.

Negative creation times may be a solution. While I don't like these adhoc codes, there may not be any other means.

Any suggestions, ideas?

Prof. Victor Lazzarini
Maynooth University
Ireland

good you are looking into this, victor.
i just want to confirm that the extended guard point is really hard to 
understand for "normal users",  for instance, creating a table for GEN02 
with 8 and 9 points both return 8 with ftlen:
	print(ftlen(ftgen(0,0,8,2,0)))
	print(ftlen(ftgen(0,0,9,2,0)))
if we could find a solution to make this more intuitive, it would be 
wonderful.


On 29/03/2024 18:48, Victor Lazzarini wrote:
> That's one idea. It might work.
> 
> 
> Prof. Victor Lazzarini
> Maynooth University
> Ireland
> 
>> On 29 Mar 2024, at 16:17, Michael Gogins  wrote:
>>
>> 
>>
>>
>>   *Warning*
>>
>> This email originated from outside of Maynooth University's Mail 
>> System. Do not reply, click links or open attachments unless you 
>> recognise the sender and know the content is safe.
>>
>> Fractional table number?
>>
>> -----------------------------------------------------
>> Michael Gogins
>> Irreducible Productions
>> http://michaelgogins.tumblr.com 
>> Michael dot Gogins at gmail dot com
>>
>>
>> On Fri, Mar 29, 2024 at 3:53 PM Victor Lazzarini 
>> > wrote:
>>
>>     Work on PR1848 has brought up one unresolved issue that has been
>>     present since we mainstreamed non-pow-two tables.
>>
>>     There are two problems:
>>
>>     1) There is currently no way to tell Csound to create a table with
>>     an extended guard point if we do not use a pow-two size.
>>
>>     2) Pow-two+1 sizes are pow-two lengths with extended guard point.
>>     There is no way to actually create a table with that size.
>>
>>     A solution for 1) is to find a way to indicate an extended guard
>>     point using some form of code (like we do to suppress rescaling)
>>
>>     We can't use negative table numbers because these were used in the
>>     past to allow np2 tables.
>>
>>     We can't use negative table numbers because these are used for
>>     internal tables.
>>
>>     Negative creation times may be a solution. While I don't like
>>     these adhoc codes, there may not be any other means.
>>
>>     Any suggestions, ideas?
>>
>>     Prof. Victor Lazzarini
>>     Maynooth University
>>     Ireland
>>

yes, that's confusing and there is not an easy solution as we want to keep backward compatibility

Prof. Victor Lazzarini
Maynooth University
Ireland

> On 29 Mar 2024, at 19:19, joachim heintz  wrote:
>
> good you are looking into this, victor.
> i just want to confirm that the extended guard point is really hard to understand for "normal users",  for instance, creating a table for GEN02 with 8 and 9 points both return 8 with ftlen:
>    print(ftlen(ftgen(0,0,8,2,0)))
>    print(ftlen(ftgen(0,0,9,2,0)))
> if we could find a solution to make this more intuitive, it would be wonderful.
>
>
>> On 29/03/2024 18:48, Victor Lazzarini wrote:
>> That's one idea. It might work.
>> Prof. Victor Lazzarini
>> Maynooth University
>> Ireland
>>>> On 29 Mar 2024, at 16:17, Michael Gogins  wrote:
>>>
>>> 
>>>
>>>
>>>  *Warning*
>>>
>>> This email originated from outside of Maynooth University's Mail System. Do not reply, click links or open attachments unless you recognise the sender and know the content is safe.
>>>
>>> Fractional table number?
>>>
>>> -----------------------------------------------------
>>> Michael Gogins
>>> Irreducible Productions
>>> http://michaelgogins.tumblr.com/ 
>>> Michael dot Gogins at gmail dot com
>>>
>>>
>>>> On Fri, Mar 29, 2024 at 3:53 PM Victor Lazzarini > wrote:
>>>
>>>    Work on PR1848 has brought up one unresolved issue that has been
>>>    present since we mainstreamed non-pow-two tables.
>>>
>>>    There are two problems:
>>>
>>>    1) There is currently no way to tell Csound to create a table with
>>>    an extended guard point if we do not use a pow-two size.
>>>
>>>    2) Pow-two+1 sizes are pow-two lengths with extended guard point.
>>>    There is no way to actually create a table with that size.
>>>
>>>    A solution for 1) is to find a way to indicate an extended guard
>>>    point using some form of code (like we do to suppress rescaling)
>>>
>>>    We can't use negative table numbers because these were used in the
>>>    past to allow np2 tables.
>>>
>>>    We can't use negative table numbers because these are used for
>>>    internal tables.
>>>
>>>    Negative creation times may be a solution. While I don't like
>>>    these adhoc codes, there may not be any other means.
>>>
>>>    Any suggestions, ideas?
>>>
>>>    Prof. Victor Lazzarini
>>>    Maynooth University
>>>    Ireland
>>>