Re: [boost] [GSoC] [Boost.Hana] Formal review request

Subject: Re: [boost] [GSoC] [Boost.Hana] Formal review request
From: Louis Dionne (ldionne.2_at_[hidden])
Date: 2014-07-30 16:19:44

Paul A. Bristow <pbristow <at> hetp.u-net.com> writes:

>
> [...]
>
> For me, the killer argument for using Quickbook is the ability to use *code
> snippets*.
> These ensure that the bits of code you show have actually been through the
> (current) compiler.

That's what I currently do with Doxygen; all the snippets in the reference 
are taken from the example/ subdirectory of the project, and all those files 
are compiled and ran just like the other unit tests. It's really great.

> I see no problems converting to Quickbook. Conversion is not too painful -
> especially if you have done it before.
> 
> However, I don't think you should worry about the docs now - they are fine 
> for a review.
> 
> IMO your priority is to get some real-life users able to make an informed
> review.

Agreed.

> Paul
> 
> PS For me, the library name is a big turn off - it doesn't say what the 
> library does.

Heterogeneous combiNAtors. I agree the mapping is not as direct as, say, MPL, 
but it still beats Spirit, Phoenix and Proto (unless those are non-obvious 
acronyms).

> And, for me, the names of functions are enough to condemn the library as
> unacceptable for Boost. 

I have reasons to be uncomfortable with changing names in the library:

  - While names are inconsistent with the usual C++, they are consistent 
    inside the library. Changing one name can break this consistency if 
    I'm not careful. Further, there are names which just don't have a 
    C++-friendly name, so I can't hope to change _all_ the names. We'll
    have to deal with either 100% FP names or 50% C++ / 50% FP, but 
    100% C++ just can't be done.

  - Some, C++ names imply some kind of mutation of the structure. Since Hana 
    does not do any mutation, I must be careful not to choose a name that 
    suggests something that's not the reality.

> This really, really impressive and obviously really useful library is for 
> C++ users, not Haskell users.

Louis