Sebastion,

Given the nature of audio platforms, I do not think excluding cgo is
feasible.
There always seems to be very important layers in userland (though
OS-privileged)
There is a really neat package which does ALSA without the accompanying
library already, github/yobert/alsa, whose headline is:

"""
This is a golang ALSA client implementation, without cgo! Unfortunately,
doing it without cgo means throwing away many years of compatibility work
that has been put into libalsa
"""

I don't think it's practical to eliminate cgo.  Also for darwin/iOS, the OS
comes with good support linked to the kernel which is only really available
at the C level.

But these are not really in the same category as portaudio, sox, ffmpeg,
etc.  I think that with time good 100% Go support can be found here, and
moreover I hope that some future codec developers consider doing their work
in Go to begin with, they could publish C support via -build-shared and
friends.

For FFT: Honestly it's been so long since I wrote that I don't remember the
reason I chose not to use go-num.  Perhaps one day go-num's fft can replace
it.  I've been looking at doing MDCT, and it would be nice if go-num had
that.  In the meantime I think the audio-specific parts of zikichombo's fft
justify it living apart.

Scott




On 8 August 2018 at 20:36, Sebastien Binet <seb.bi...@gmail.com> wrote:

> Scott,
>
> One thing that wasn't completely clear to me in the blog post was whether
> zikichombo was considering investigating a pure Go alsa compatible package.
> If I am not mistaken, right now, on Linux and on Darwin, the package
> actually performing audio stuff is using cgo.
>
> Also, while I understand the fft code currently part of zikichombo is
> specifically tailored for the audio use cases.
> I was wondering whether you had considered using gonum/fft and dropped it
> for some reason (and I'd be interested in knowing what that was) out
> whether you just didn't know about its existence (it's pretty new, but
> based on the netlib implementation, IIRC)
>
> Cheers,
> -s
>
> sent from my droid
>
> On Wed, Aug 8, 2018, 20:25 Zellyn <zel...@gmail.com> wrote:
>
>> Scott,
>>
>> Thank you for taking on this thorny and important area. I hope you manage
>> to include the various interested parties, and reach a consensus that is
>> both efficient and Go-like.
>>
>> Good luck!  I look forward to reading updates on the project blog.
>>
>> Zellyn
>>
>>
>> On Tuesday, August 7, 2018 at 11:33:48 PM UTC-4, Scott Cotton wrote:
>>>
>>> Hello all,
>>>
>>> I am pleased to announce a new project dedicated to rendering "awesome"
>>> sound processing and i/o in Go, ZikiChombo <http://zikichombo.org>,
>>>  The project is in alpha, and this announcement is mostly geared toward
>>> people potentially interested
>>> in helping.
>>>
>>> (you may want to check out the launch blog post
>>> <http://zikichombo.org/blog/launch> for a more personal introduction).
>>>
>>> Best,
>>> Scott
>>>
>>>
>>>
>>> --
>> You received this message because you are subscribed to the Google Groups
>> "golang-nuts" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to golang-nuts+unsubscr...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>>
>


-- 
Scott Cotton
President, IRI France SAS
http://www.iri-labs.com

-- 
You received this message because you are subscribed to the Google Groups 
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to golang-nuts+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to