nice, thanks Stéphane.
I would just add `'include ` to the example using placement new.
best,
Martin
On 19/06/17 09:28, Stéphane Letz wrote:
> Hi All,
>
> After several « back and forth » steps and test on a real use case (thanks
> Martin Klang for feedback on OWL impleme
that would work for us.
We have a program size limit of 64k, but 1M heap available, so heap
allocation is really important for us.
Martin
On 08/06/17 13:06, Stéphane Letz wrote:
>> Le 8 juin 2017 à 13:39, Kjetil Matheussen a écrit
>> :
>>
>>
>>
>> On Thu, Jun 8, 2017 at 1:14 PM, Stéphane Let
lopment of the backend…)
>
> - so we’ll have to rework this table memory allocation scheme in all backend
> to be perfectly clean. Added on the « TODO » list.
>
> Stéphane
>
>
>> Le 8 juin 2017 à 12:31, Martin Klang a écrit :
>>
>> I've got another quest
ll
works beautifully.
I imagine there's a reason why classInit() and the generated tables are
static.
Martin
On 06/06/17 15:58, Stéphane Letz wrote:
>> Le 6 juin 2017 à 16:50, Martin Klang a écrit :
>>
>> Hello list, this is my first post - Martin in London here!
>>
>
On 06/06/17 15:58, Stéphane Letz wrote:
> We are not generating any optimized version of math functions based on
> lookup tables
oh, okay!
I think I managed to confuse myself with different versions of patches
and Faust.
Looking at it again it seems clear that the tables I found are
expli
Hello list, this is my first post - Martin in London here!
I develop the OWL Pedal / Modular, an embedded platform which runs FAUST.
We're trying to get up to speed with the latest version, but have hit a
couple of snags.
Somewhere between v0.9.85 and 0.9.90 it seems that optimised versions of
ma
