Re: [Chicken-users] Blobs: a (modest?) proposal

[Alaric Snell-Pym]

On 24 Jan 2009, at 2:43 pm, felix winkelmann wrote:

> On Sat, Jan 24, 2009 at 1:10 PM, Alaric Snell-Pym
> <address@hidden> wrote:
> > What do people think of this? Would it be welcome in the chicken core
> > once it's proven itself?
>
> Sure. The use of strings for binary data is indeed something that
> should
> be reduced. There is also a lot of copying involved. But one has to be
> careful not to re-implement memory management completely: Use
> srfi-4 bytevectors where appropriate and blobs where the data is
> truly opaque (but I guess this is self-evident).

Indeed... my desire to base srfi-4 vectors on top of blobs is purely
so they can share an underlying memory management structure so that
raw I/O functions - memory map, reads, random C libraries such as
spread and gdbm and compression and encryption libraries, can use the
blob system to encapsulate the opaque buffers they return; then srfi-4
vectors, being wrappers around blobs that specialise their semantics,
can just sit on top of the foreign memory management of the blob.

> cheers,
> felix

ABS

--
Alaric Snell-Pym
Work: http://www.snell-systems.co.uk/
Play: http://www.snell-pym.org.uk/alaric/
Blog: http://www.snell-pym.org.uk/?author=4