Haskell report (August release)

[Simon Peyton-Jones]

A reasonable suggestion, but I'm afraid I don't want
to make structural changes to the Report.

Simon

| -----Original Message-----
| From: Mark Tullsen [mailto:mtullsen@cse.ogi.edu]=20
| Sent: 27 September 2001 20:35
| To: Simon Peyton-Jones
| Cc: haskell@haskell.org
| Subject: Re: Haskell report (August release)
|=20
|=20
| Simon,
|=20
| Here's a minor quibble with the organization of the Report=20
| and Library, not with the content.  Sorry if this has been=20
| brought up before.
|=20
| In section 4.3.3, Derived Instances, of the Report there is
|=20
|   The only classes in the Prelude for which derived instances are
|   allowed are Eq, Ord, Enum, Bounded, Show, and Read, all mentioned in
|   Figure 5, ...  Classes defined by the standard libraries may also be
|   derivable.
|=20
| In the introduction to the Library Report there is
|=20
|   Classes defined in libraries may be derivable. This report includes
|   the derivation of such classes when appropriate.
|=20
| Now, unless I missed something, the only class in the Library=20
| Report which is derivable is Ix.
|=20
| I would argue for bringing the Ix class into the Report for=20
| these reasons
| =20
|   * One does not have to search through the Library Report to=20
| determine
|     what is derivable.
|=20
|   * I think one would expect that the Libraries contain stuff=20
| that could be
|     implemented in Haskell by the user.  Until Haskell has=20
| the ability=20
|     to allow for user-defined derivable classes, Ix cannot be
|     defined by the user.
|=20
| - Mark
|=20