"brian m. carlson" <sandals@xxxxxxxxxxxxxxxxxxxx> writes:
> The current code states "unknown hash algorithm length". This is
> slightly confusing, and moreover, we've gotten feedback that it's hard
> to translate into other languages. Since the case is that we cannot
> detect the hash algorithm, let's just say that. This is clearer and
> more direct, reflects the intent of the code (which calls the function
> detect_hash_algo just a few lines above), and will be easier to
> translate.
>
> Suggested-by: Matthias Rüster <matthias.ruester@xxxxxxxxx>
> Signed-off-by: brian m. carlson <sandals@xxxxxxxxxxxxxxxxxxxx>
> ---
> Matthias contacted me privately about this message and I thought since
> we're still in the RC period, it would make sense to include this in to
> help translators and those who aren't native speakers.
Yup, and yuck. Didn't we find any room in the bundle header to
sneak in the actual hash algorithm name in a backward compatible
way? I am actually OK if we declared that v2 bundle files are SHA-1
only and v3 bundle file has an item in the bundle header to say what
hash algorithm is used, than using "ah this uses a hash algorithm
that produces a 32-byte output, so it must be algorithm X", to be
honest.
> bundle.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/bundle.c b/bundle.c
> index 2a0d744d3f..69c807bda9 100644
> --- a/bundle.c
> +++ b/bundle.c
> @@ -66,7 +66,7 @@ static int parse_bundle_header(int fd, struct bundle_header *header,
> if (!header->hash_algo) {
> header->hash_algo = detect_hash_algo(&buf);
> if (!header->hash_algo) {
> - error(_("unknown hash algorithm length"));
> + error(_("unable to detect hash algorithm"));
> status = -1;
> break;
> }
