Sebastian suggests using some syntax other than pattern
matching to express the isomorphism involved in a newtype.
I can't see any advantage in this.
Further, Simon PJ claims that if someone has written
data Age = Age Int
foo (Age n) = (n, Age (n+1))
that we want to be able to
In a recent message Sebastian Hunt suggests a solution to the 'newtype'
problem. Let me recall another approach which can cure several things at a
time (probably introducing new problems though).
Some time ago Mark Jones wrote a paper " From Hindley-Milner Types to
Modular Structures". He sugges
John P. Haskell writes:
> Howdy!
> Yes, you never thought it was possible but your old-fashioned
> brother is learning how to do research on the internet and I have an
> e-mail account. Yippee! [remainder deleted]
To the subscribers of the Haskell Mailing List <[EMAIL PROTECTED]>:
T
The proposed change should not break too much code. The `.' is only
treated specially after a constructor. This will break:
tstpatp9= (P_cons P_write.P_cons P_read) P_empty_list
because of the P in P_write but this will be fine:
proc_emulator f = prm_em.f
Since constructors appear r
I think that the following points have emerged from the recent discussion
about the proposed newtype declaration:
1) Pattern matching against strict constructors will result in
functions which are strict in the annotated constructor argument. For
example:
data T = A !Int
Jules Gilbert writes:
>>Huh?
I don't know what this letter of yours is in response to. I have been
getting funny misdirected e-mail recently so if you are answering
something that appears to come from me I would be interested to see
the details.
Richard Platek
John P Haskell writes:
>>Howdy!
>> Yes, you never thought it was possible but your old-fashioned
>>brother is learning how to do research on the internet and I have an
>>e-mail account. Yippee!
>> This is really quite fun. I've e-mailed people here, in canada
>>and now that I've fi
Yesterday I had sent certain suggestion for the Haskell prelude.
Then I recalled it and found it rubbishy.
It concerned changing `Ordering' to `Compare' for the result type
of `compare' in Haskell 1.3 and the function foo having the type
Foo.
Improving a confusion:
`Ordering' is not a typ
John:
You got the wrong steve. There are two Haskells: A Jennie and a
Patrick.
--
Steve (really "D. E.") Stevenson [EMAIL PROTECTED]
Department of Computer Science,(803)656-5880.mabell
Clemson University, Clemson, SC 29634-1906
Wanted: Sterbenz, P. Floating Point
I just had a look in the proposed changes for Haskell 1.3 again, being
stuck in what I was otherwise trying to do, (more of that later maybe)
when I found the Qualified names:
> Qualified names are defined in the lexical syntax. Thus,
> `Foo.a' and `Foo . a' are quite different. No whitespace
10 matches
Mail list logo
