caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Yotam Barnoy <yotambarnoy@gmail.com>
To: Paolo Donadeo <p.donadeo@gmail.com>
Cc: OCaml mailing list <caml-list@inria.fr>
Subject: Re: [Caml-list] Fwd: Proposal: re-design of ocaml headers
Date: Fri, 27 Sep 2013 14:41:58 -0400	[thread overview]
Message-ID: <CAN6ygOmS01G16DGXJQtyT2ivAP-gh_tV1i1De1wDa9oPX89_3Q@mail.gmail.com> (raw)
In-Reply-To: <CAPzAKVARRPNA0Wed=BbaGYUN=1zCCKm=T87+NY6FRvC6tQdN+Q@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 1067 bytes --]

On Fri, Sep 27, 2013 at 2:15 PM, Paolo Donadeo <p.donadeo@gmail.com> wrote:

> Many bindings deal with memory allocation. The binding of Lua deeply
> interacts with the garbage collector and inspects the actual type
> (tag) of OCaml value passed.
>
> In principle you are right but the reality is that there is no much
> "abstraction" in C :-)
>
> It's true that coming from the C side, abstraction is a problem. However,
the kind of bindings you're describing (Lua) essentially inhibit any
improvement to the GC (such as parallelization, which is bound to happen
some day) or to the runtime. It seems like it would be much better if
bindings from C (Lua, for example) called more general functions or
bindings present in the runtime's C code aka an API. Given that it seems
like some bindings currently don't use an API but inspect values closely, I
don't see any solution other than adapting the things that require changes.
Hopefully the majority of C bindings (at least from the ocaml side) can be
moved to ctypes, and can therefore be made more generic.

-Yotam

[-- Attachment #2: Type: text/html, Size: 1520 bytes --]

  reply	other threads:[~2013-09-27 18:42 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-09-27 14:05 [Caml-list] " Yotam Barnoy
2013-09-27 15:08 ` Dmitry Grebeniuk
     [not found]   ` <CAN6ygOmuCX6HLfSns0tXQCF3LWMANqhpnSN0vGWcNg0one2QzQ@mail.gmail.com>
2013-09-27 15:25     ` [Caml-list] Fwd: " Yotam Barnoy
2013-09-27 16:20       ` Dmitry Grebeniuk
2013-09-27 18:08         ` Yotam Barnoy
2013-09-27 18:12           ` Yotam Barnoy
2013-09-27 18:15           ` Paolo Donadeo
2013-09-27 18:41             ` Yotam Barnoy [this message]
2013-09-27 15:31   ` [Caml-list] " Anthony Tavener
2013-09-27 15:37     ` Yotam Barnoy
2013-09-27 16:50     ` Dmitry Grebeniuk
2013-09-30 14:48 ` Goswin von Brederlow
2013-09-30 15:31   ` Yotam Barnoy
2013-10-08 10:52     ` Goswin von Brederlow
2013-10-11 15:48       ` Yotam Barnoy
2014-01-30 20:53         ` Yotam Barnoy
2014-02-01 15:27         ` Goswin von Brederlow
2013-10-06 10:39 ` Florian Weimer

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAN6ygOmS01G16DGXJQtyT2ivAP-gh_tV1i1De1wDa9oPX89_3Q@mail.gmail.com \
    --to=yotambarnoy@gmail.com \
    --cc=caml-list@inria.fr \
    --cc=p.donadeo@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).