[Cuis-dev] Float32Array>at: and at:put:
Luciano Notarfrancesco
luchiano at gmail.com
Wed Nov 19 06:04:54 PST 2025
Oh, I just realized in Cuis Float64Array is variableWordSubclass, but it
should be variableDoubleWord right? And the we wouldn’t need to convert
endianness either. Or you did that for compatibility with 32 bit images?
Anyway, I don’t need Float64Array for now, I’m only using Float32Array.
On Wed, Nov 19, 2025 at 20:38 Juan Vuletich <juan at cuis.st> wrote:
> Hi Luciano,
>
> This is neat!
>
> Works perfectly for Float32Array. However the numbered primitives don't
> work for Float64Array. I added another cs to restore the old behavior for
> them. Now @GitHub.
>
> Thanks!
> On 2025-11-19 7:37 AM, Luciano Notarfrancesco via Cuis-dev wrote:
>
> Please review...
>
> On Wed, Nov 19, 2025 at 5:11 PM Luciano Notarfrancesco <luchiano at gmail.com>
> wrote:
>
>> Found it, primitives 238 and 239, the performance almodt doubled :)
>>
>> On Wed, Nov 19, 2025 at 10:06 Luciano Notarfrancesco <luchiano at gmail.com>
>> wrote:
>>
>>> I would imagine there’s a bytecode for this, and that at some point with
>>> Sista we’ll be able to optimize loops that access float arrays, but calling
>>> the plugin seems to get in the way…
>>>
>>> On Wed, Nov 19, 2025 at 09:52 Luciano Notarfrancesco <luchiano at gmail.com>
>>> wrote:
>>>
>>>> We're still calling the FloatArrayPlugin for accessing Float32Array.
>>>> But now that we have immediate floats, isn't there a way to do this better
>>>> (faster)?
>>>>
>>>
> --
> Juan Vuletichwww.cuis.stgithub.com/jvuletichresearchgate.net/profile/Juan-Vuletichindependent.academia.edu/JuanVuletichpatents.justia.com/inventor/juan-manuel-vuletich
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cuis.st/mailman/archives/cuis-dev/attachments/20251119/e045d742/attachment-0001.htm>
More information about the Cuis-dev
mailing list