Re: [Caml-list] ocaml considered dangerous

[Gabriel Scherer <gabriel.scherer@gmail.com>]

I guess one could also mention that the no-magic implementation is
notably faster than the current implementation (as it avoids any
format string parsing at runtime), which is not necessarily a concern
for printf (for large output you're probably killed by I/O overhead
anyway) but can be for sprintf or bprintf.

Note that while the current proposal uses GADTs, Benoît also
experimented with reasonable similar designs using higher-order
function (this is the usual initial vs. final tagless interpreter
choice) that would have been implementable before 4.00.

On Sat, Jan 18, 2014 at 3:22 AM, Jeremy Yallop <yallop@gmail.com> wrote:
> Jacques Garrigue wrote:
>>> By the way, there is now an implementation of Printf that avoids most of the Obj.magic by using GADTs. It should be merged soon.
>
> Jon Harrop <jon@ffconsultancy.com> wrote:
>> Just curious: is that a drop-in replacement?
>
> Yes, except for some tiny incompatibilities, mostly fixes for bugs
> that the development of the GADT version uncovered in the original
> implementation.  The patch is here, along with a description and some
> discussion:
>
>    http://caml.inria.fr/mantis/view.php?id=6017
>
> The author, Benoît Vaugon, presented the design along with some
> performance figures at OCaml 2013:
>
>    http://ocaml.org/meetings/ocaml/2013/proposals/formats-as-gadts.pdf
>    http://ocaml.org/meetings/ocaml/2013/slides/vaugon.pdf
>
> --
> Caml-list mailing list.  Subscription management and archives:
> https://sympa.inria.fr/sympa/arc/caml-list
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs