#5744: List layouts
---------------------------------+------------------------------------------
Reporter: nsch | Owner: nsch
Type: feature request | Status: new
Priority: normal | Milestone:
Component: Compiler | Version: 7.2.1
Keywords: | Os: Unknown/Multiple
Architecture: Unknown/Multiple | Failure: None/Unknown
Difficulty: Unknown | Testcase:
Blockedby: | Blocking:
Related: |
---------------------------------+------------------------------------------
Comment(by ehird):
Here's some justifications of this feature from my point of view (talking
about the version generalised to monoids, as I think it's the best
option):
There are quite a few widely-used libraries that use a monad for what is
essentially an optimised `Writer w ()` over a specific ''w''. Examples:
* [http://jaspervdj.be/blaze/ blaze-html] uses a
[http://hackage.haskell.org/packages/archive/blaze-
html/0.4.3.1/doc/html/src/Text-Blaze-Internal.html#HtmlM laws-breaking
Monad instance] so it can [http://jaspervdj.be/blaze/tutorial.html use
layout]
* [http://hackage.haskell.org/package/cereal cereal] has an
[http://hackage.haskell.org/packages/archive/cereal/0.3.5.0/doc/html/Data-
Serialize-Put.html optimised Writer monad] for serialisation; the user
doesn't care about the PutM monad itself, just the `type Put = PutM ()`
alias
So, one thing this is useful for is ''output'' and specification of
complex nested structures. These monads are exclusively used in a monoidal
style (indeed, if they're only used with the `()` result type (as e.g.
blaze-html must be) they're equivalent to monoids), have no use for do-
notation binds, and would be far more comfortable with a notation
specifically designed for monoids.
You could just use `mconcat [...]` everywhere, but writing long lists over
multiple lines is ugly. A nice thing about a monoid notation is that a lot
of the libraries can be seen (and can be implemented as) imperatively
streaming their output to a handle, so an imperative sequencing notation
and operations like `put :: Word8 -> M` are reasonable. In light of this,
and because the other proposed names seem too likely to clash to me, my
favourite name so far is `mdo`, for "monoidal do". The only worry there is
that it would clash with the obsolete recursive do notation, but it *is*
deprecated, and its history of being a reserved keyword will mean that
clashes are unlikely.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/5744#comment:9>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
