Yes, argument to higher rank functions are probably the top reason why
MonoLocalBinds is a nuisance.
As of now I think the best thing is to do (1), but define type synonyms that
abbreviate the oft-repeated signatures. That should make the signatures much
onerous.
Simon
| -Original
Ok, I've got a patch that fixes this segfault. In the process I looked
at all patches to Cg* modules after Nov 2009 and looked for changes that
weren't applied to the new codegen. I skipped the LLVM patch, but picked
up the rest of the blackhole changes. There are, however, two hunks that
I am
Hello,
I'm trying to recover my opensolaris builder machine after disk crash,
but after reinstall I'm not able to build any GHC there. I'm trying head
and now also 6.12.3 as a reference (as I'm able to build it on my
workstation with the same OS). The problem I see here is when I invoke
Hello,
Another design-pattern which sometimes works pretty well is to
encapsulate commonly used polymorphic types in ordinary data-types
(i.e., use the rank-2 style). Then, the data-type constructors provide
a quick way to---essentially---write a type signature. It seems that
this should work well
Is it safe to consider type families and associated type families
extensions for ghc as stable ? Wich related extensions (flexible
contexts, undecidable instanses and so on) may be deprecated or changed
in near (2-3 years) future and wich may not?
___
