Re: Minor bugs concerning translations

[Nick Clifton]

Hi Tommy,

> I am a member of the Swedish translation team <address@hidden>, 
> and I have just completed a Swedish translation of
> binutils-2.12-pre020121.pot.  I would like to comment on a
> few things I have found.

Thanks very much for submitting these comments.  I have applied fixes
for most of these problems to the CVS sources except for:

> |          if (string_table == NULL)
> |            printf ("%d", unwsec->sh_name);
> |          else
> |            printf ("'%s'", SECTION_NAME (unwsec));
> |        }
> 
> I also think most other
> messages in binutils use quotes like `this' and not like
> 'this'.

This is true, but readelf is internally consistent in it use of the
single quote, and personally, I think that the `form' is less
attractive than the 'form'.  So unless people violently object I am
going to leave things as they are.

> binutils/readelf.c:5988
> 
> |       printf (_("  DWARF Version:               %d\n"), info.li_version);
> |       printf (_("  Prologue Length:             %d\n"), 
> info.li_prologue_length);
> |       printf (_("  Minimum Instruction Length:  %d\n"), 
> info.li_min_insn_length);
> |       printf (_("  Initial value of 'is_stmt':  %d\n"), 
> info.li_default_is_stmt);
> |       printf (_("  Line Base:                   %d\n"), info.li_line_base);
> |       printf (_("  Line Range:                  %d\n"), info.li_line_range);
> |       printf (_("  Opcode Base:                 %d\n"), 
> info.li_opcode_base);
> 
> I think this quote should be `is_stmt'.

Same reasoning applies here.

> In binutils/objcopy.c:setup_section() there is a word
> puzzle with
> 
>   err = _("foo");
>   goto loser;
> 
>   [...]
> 
>   err = _("bar");
>   goto loser;
> 
>   [...]
> 
>   loser:
>     non_fatal (_("some message with %s in it", err));
> 
> If you don't have access to the source code it is almost
> impossible to get such translations right, and even with the
> source code, it can be a puzzle to make it work in another
> language.

This has been fixed in the current CVS sources.

Once again thanks very much for your input.

Cheers
        Nick