[m-rev.] for review: stream typeclasses

[Julien Fischer]

On Wed, 25 Oct 2006, Peter Ross wrote:

> On 10/25/06, Julien Fischer <juliensf at csse.unimelb.edu.au> wrote:
>> 
>> On Mon, 23 Oct 2006, Peter Ross wrote:
>> 
>> > I think streams should support non-blocking I/O as well.
>> > I can't forsee any problems due to the current hierachy of typeclasses
>> > to supporting non-blocking I/O, but I haven't thought about it deeply.
>> > Have you?
>> 
>> The following diff adds support for non-blocking streams (however
>> see the XXX on the non-blocking version of put).
>> 
>> Comments?
>> 
> First comment I prefer the inst scheme with one type for blocking and
> non-blocking, so I like this diff.
>
> My only thoughts are what does an async write mean.  Does it mean,
> attempt to write the unit and then later check to see if write has
> been successful or does it mean attempt to write and if the device is
> busy already don't write?

Since it seems we can add support for non-blocking streams in a
backwards compatible fashion I suggest we leave them out of the
stream module until (1) we have addressed the above issues and (2)
we actually have some examples of non-blocking streams.

Julien.
--------------------------------------------------------------------------
mercury-reviews mailing list
Post messages to:       mercury-reviews at csse.unimelb.edu.au
Administrative Queries: owner-mercury-reviews at csse.unimelb.edu.au
Subscriptions:          mercury-reviews-request at csse.unimelb.edu.au
--------------------------------------------------------------------------