English version
Accueil     À propos     Téléchargement     Ressources     Contactez-nous    

Ce site est rarement mis à jour. Pour les informations les plus récentes, rendez-vous sur le nouveau site OCaml à l'adresse ocaml.org.

Browse thread
ocamlbuild is VERY nifty
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: 2007-03-07 (21:01)
From: Jon Harrop <jon@f...>
Subject: Re: [Caml-list] beta-test of OCaml 3.10.0
On Wednesday 07 March 2007 15:53, Joel Reymont wrote:
> Deep into my first project I confirm that ocamlbuild is VERY nifty.

Excellent news. Make just can't hack it when things get non-trivial and I'm 
bored of recompiling everything all the time, even if ocamlopt is a million 
times faster than gcc. :-)

Here are my results (timings are old vs new) for my 2.2GHz AMD64x2:

OCaml itself still doesn't parallel build but serial builds in around 4mins.

Ray tracer builds fine: 3.963s vs 4.000s run time

n-th nearest neighbour example from OFS builds and runs fine: 8.142s vs 8.320s

Smoke builds fine and runs fine: 12.39s vs 11.56s build time, 68.5fps vs 
72.5fps running speed

Presenta builds fine: 39.354s vs 39.524s

I've only had one problem and I'm not sure if that is due to the new version 
of ocamlopt: Presenta screws up font rendering when using the experimental 
stencil buffer code instead of the usual GLU tesselator code.

So my conclusion is that 3.10 looks fine. Compile times are slightly faster in 
some cases. Running times are basically indifferent.

Dr Jon D Harrop, Flying Frog Consultancy Ltd.
OCaml for Scientists