Creating new effects?

[Shea Levy]

Hi all,

Is it possible to create new effects of sort eff? For example, none of
the existing cases at [1] seem to cover the side effects of the 'open'
system call, but that is surely not pure!

[1]: https://github.com/githwxi/ATS-Postiats/wiki/effects

~Shea

[gmhwxi]

The sort 'eff' is currently not extensible.

The effect system in ATS is not meant to track effects like 'open'.
Tracking such effects makes programming intractable in general. The
primary use of the effect system in ATS is to ensure that an implemented
proof function is indeed pure.

[Shea Levy]

Ah I see, thanks. Under what circumstances should I care about
explicitly annotating functions with their effects?

~Shea

> --
> You received this message because you are subscribed to the Google Groups "ats-lang-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ats-lang-user...@googlegroups.com.
> To post to this group, send email to ats-lan...@googlegroups.com.
> Visit this group at http://groups.google.com/group/ats-lang-users.
> To view this discussion on the web visit https://groups.google.com/d/msgid/ats-lang-users/f7eddf8a-ba78-4777-a0f1-4f63aa885472%40googlegroups.com.

[Hongwei Xi]

It is hard to say.

I occasionally use effect-tracking to catch bugs causing non-termination.

When implementing (pure) algorithms, one may have the luxury to be very
thorough. In general, I think effect-tracking only works on code of very small
scale.

To view this discussion on the web visit https://groups.google.com/d/msgid/ats-lang-users/20140831181825.GD2237%40nixos.hsd1.nh.comcast.net.

[Shea Levy]

I see, thanks. I think for now I will just work without effect
annotations until I find a use case.

~Shea

> > https://groups.google.com/d/msgid/ats-lang-users/20140831181825.GD2237%40nixos.hsd1.nh.comcast.net
> > .

> >
>
> --
> You received this message because you are subscribed to the Google Groups "ats-lang-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ats-lang-user...@googlegroups.com.
> To post to this group, send email to ats-lan...@googlegroups.com.
> Visit this group at http://groups.google.com/group/ats-lang-users.

> To view this discussion on the web visit https://groups.google.com/d/msgid/ats-lang-users/CAPPSPLrNzsokUcUHAy90tTWznjFaUNp%2BWFv8uevGQNcqZP2xCA%40mail.gmail.com.

[gmhwxi]

I would say that is the right way to proceed.

ATS is feature-rich. It is very important to make sure that
you do not get distracted by features at this stage.

[Yannick Duchêne]

Le dimanche 31 août 2014 03:11:45 UTC+2, gmhwxi a écrit :

The sort 'eff' is currently not extensible.

The effect system in ATS is not meant to track effects like 'open'.

Better use a linear type instead? 

[Hongwei Xi]

Yes, using linear proofs (of views) is another possibility. Another possibility

is to adopt a monadic style of programming.

--

You received this message because you are subscribed to the Google Groups "ats-lang-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ats-lang-user...@googlegroups.com.
To post to this group, send email to ats-lan...@googlegroups.com.
Visit this group at http://groups.google.com/group/ats-lang-users.

To view this discussion on the web visit https://groups.google.com/d/msgid/ats-lang-users/f3f8e29c-4ae8-4b18-a3dd-7a8ba56d8b88%40googlegroups.com.