[gnutls-devel] [PATCH 2/2] guile: doc: Mention bytevectors.

[Nikos Mavrogiannopoulos]

Both pushed, thank you.

On Mon, Mar 14, 2016 at 10:44 AM, Ludovic Courtès <ludo at gnu.org> wrote:
> * doc/gnutls-guile.texi (Representation of Binary Data): Mention bytevectors.
> (Input and Output): Likewise.
> ---
>  doc/gnutls-guile.texi | 18 ++++++++++--------
>  1 file changed, 10 insertions(+), 8 deletions(-)
>
> diff --git a/doc/gnutls-guile.texi b/doc/gnutls-guile.texi
> index 8a085be..0bb7995 100644
> --- a/doc/gnutls-guile.texi
> +++ b/doc/gnutls-guile.texi
> @@ -288,13 +288,14 @@ procedure applies to session.
>  Many procedures operate on binary data.  For instance,
>  @code{pkcs3-import-dh-parameters} expects binary data as input.
>
> + at cindex bytevectors
>  @cindex SRFI-4
>  @cindex homogeneous vector
> -
> -Binary data is represented on the Scheme side using SRFI-4 homogeneous
> -vectors (@pxref{SRFI-4,,, guile, The GNU Guile Reference Manual}).
> -Although any type of homogeneous vector may be used, @code{u8vector}s
> -(i.e., vectors of bytes) are highly recommended.
> +Binary data is represented on the Scheme side using bytevectors
> +(@pxref{Bytevectors,,, guile, The GNU Guile Reference Manual}).
> +Homogeneous vectors such as SRFI-4 @code{u8vector}s can also be
> +used at footnote{Historically, SRFI-4 @code{u8vector}s are the closest
> +thing to bytevectors that Guile 1.8 and earlier supported.}.
>
>  As an example, generating and then exporting Diffie-Hellman parameters
>  in the PEM format can be done as follows:
> @@ -385,9 +386,10 @@ applications do their own buffering.
>  @findex record-receive!
>
>  A lower-level I/O API is provided by @code{record-send} and
> - at code{record-receive!} which take an SRFI-4 vector to represent the
> -data sent or received.  While it might improve performance, it is much
> -less convenient than the above and should rarely be needed.
> + at code{record-receive!} which take a bytevector (or a SRFI-4 vector) to
> +represent the data sent or received.  While it might improve
> +performance, it is much less convenient than the session record port and
> +should rarely be needed.
>
>
>  @node Exception Handling
> --
> 2.6.3
>