On 5/27/12 2:04 PM, Daniel Bünzli wrote:
>
> Le dimanche, 27 mai 2012 à 18:53, Hongbo Zhang a écrit :
>
>> Well, I don't think it's good to downplay camlp4. The problem is we need
>> more documentation and more tests.
>
> camlp4 is one of the (conceptually) ugliest component of the OCaml system. Pre-processors in general are a wrong solution to a real problem; meta-programming facilities should be part of the core language and play by its scoping rules, they should not be layered on top of it.
Come on, Parsetree.structure->Parsetree.structure does not solve the 
scope rule problem.
It's essentially a less powerful p4 tool and does not solve the problem 
at all. It just fragment the community again.

> The aformentioned proposal is certainly not the best I'm hoping for (the definition and semantics of the attributes is defined by the build system). However it may be a step in the right direction. For one thing by preventing the extension of OCaml's concrete syntax and making attributes part of it, using them doesn't break other tools that act on OCaml sources files; the first one being your syntax highlighter.
>
> Best,
>
> Daniel
>
>