On Mon, May 7, 2018, 4:24 AM gmhwxi wrote:
> Yes. It prompted me to take a look at LGPLv3.
>
> GPL/LGPL is tailored for compiling C into object code. ATS is a bit
> different. It is important to make sure that any C code generated by
> ATS/Postiats (compiler + libraries) is, by default, not licen
That would greatly complicate the semantics of ATS.
On Sunday, May 6, 2018 at 3:05:45 PM UTC-4, Steinway Wu wrote:
>
> I was playing with a utility library based on type classes, implemented as
> templates. As you mentioned, I experienced "interface explosion". For
> things like a fold, or a fma
Yes. It prompted me to take a look at LGPLv3.
GPL/LGPL is tailored for compiling C into object code. ATS is a bit
different. It is important to make sure that any C code generated by
ATS/Postiats (compiler + libraries) is, by default, not licensed under
GPL/LGPL.
On Thursday, May 3, 2018 at 11:32
I was playing with a utility library based on type classes, implemented as
templates. As you mentioned, I experienced "interface explosion". For
things like a fold, or a fmap, I have a t@ype version that is probably
garbage collected, a vt@ype version that is call-by-value, and a vt@ype
version
Yes. Should probably use some more informative syntax in ATS3.
On Sun, May 6, 2018 at 2:15 PM, Steinway Wu wrote:
> And the `free` method is defined to be the second argument of $ldelay? Is
> that understanding correct?
>
> On Sunday, May 6, 2018 at 12:19:55 AM UTC-4, gmhwxi wrote:
>>
>>
>> A l
And the `free` method is defined to be the second argument of $ldelay? Is
that understanding correct?
On Sunday, May 6, 2018 at 12:19:55 AM UTC-4, gmhwxi wrote:
>
>
> A linear stream is a bit like an object with two methods:
> one for eval and the other for free; !xs calls the eval method
> and ~