Re: native code syntax extensions

[Anil Madhavapeddy <anil@xxxxxxxxxx>]

This is on your 32-bit Mac? That pretty much covers all the combinations then, phew...

The FS test is David's so it might be slightly out of date; I'll look at it shortly after getting some more of the slides done and merging your/Raphael's Lwt tutorials.

Anil

On 17 Sep 2011, at 18:43, Balraj Singh wrote:

Builds for me too - went from 2 min 10, to 1 min 15 seconds.

The fs regression does not compile though (Error: Unbound value OS.Blkif.enumerate).

Balraj

On Sat, Sep 17, 2011 at 8:22 AM, Anil Madhavapeddy <anil@xxxxxxxxxx> wrote:

Great! What are the warnings?

Anil

On 17 Sep 2011, at 16:22, Raphael Proust wrote:

> Build ok here! (a few warnings though)
>
> On 9/17/11, Anil Madhavapeddy <anil@xxxxxxxxxx> wrote:
>> I've finally added support for loading all the syntax extensions as native
>> code, instead of byte code (although that option remains in the myocamlbuild
>> in case of disaster).
>>
>> Since we're up to 10 extensions now (Bitstring, Lwt, Ulex, JSON, CSS, HTML,
>> XML, Dyntype, Type_conv, and eventually the ORM), this has hugely decreased
>> build time: by about 3x-4x for the main tree, and more for statically linked
>> in modules.
>>
>> But since it's so close to the tutorial, would be great if everyone could
>> test their trees by pulling from mine and making sure the dynamic code
>> loading works on their platform (I've tested on MacOS x64 and Linux x64).
>> You just need to pull, make && make install, and perhaps build mirage-www.
>>
>> Anil
>>
>
>
> --
> _______
> Raphael
>