[Gllug] trashman - a trash management utility for Linux filesystem

[Nix]

On Thu, 2 Feb 2006, Richard Jones mused:
> On Thu, Feb 02, 2006 at 08:24:22AM +0000, Nix wrote:
>> Ooo! I know classic ML was awful at it...
> 
> My classic ML is not so good.  Sticking with OCaml though ...
> 
> This C code:
> 
>   struct foo {
>     struct bar *ptr;
>   }
> 
>   foo_a->ptr = foo_b;
> 
> is exactly equivalent to this in OCaml:
> 
>   type foo = {
>     mutable ptr : bar;
>   }
> 
>   foo_a.ptr <- foo_b;

... although quite why one would use ML and then write C in it
is... unclear to me. ;)

> Even the in-memory representation and the code generated to
> read/update the pointer is identical.

Nifty! Plus of course OCaml is probably more amenable to optimizations
(not that that matters in disk-bound filesystem code!)

One of the reasons I'm using a lesser language is actually to encourage
me to use *no* non-transient in-memory data structures, so the stupider
the language's type system the happier I am, to a certain degree. If I
ensure that all storage is on the disk, then making Recant a distributed
filesystem is dog simple: just start up Recant daemons on other machines
and connect them to the same database. (No distributed locking, though;
FUSE doesn't pass the POSIX locking stuff down to userspace.)

Plus, threading it also becomes a doddle, especially with the aid of the
lovely `transactor' template in libpqxx, which repeatedly reinvokes a
deadlocking transaction until it succeeds (within certain bounds): of
course the code that implements that transaction should not modify any
global state!

But I'll have a look at OCaml anyway. :)

> The only thing OCaml won't let you do is to create a foo without
> initializing the ptr field (which you can obviously do in C just by

How simply *terrible*.

> using malloc and then not bothering to initialize the field).  This
> "limitation" of OCaml is probably to be regarded as a Good Thing ...

Yeah, I tend to dislike memory allocation bugs as much as the next man
:)

-- 
`I won't make a secret of the fact that your statement/question
 sent a wave of shock and horror through us.' --- David Anderson
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug