a.out Sven Hartrumpf (06 Jul 2016 21:20 UTC)
Re: a.out Marc Nieper-Wißkirchen (07 Jul 2016 06:07 UTC)
Re: a.out Pierpaolo Bernardi (07 Jul 2016 07:43 UTC)
Re: a.out Marc Nieper-Wißkirchen (07 Jul 2016 14:48 UTC)

Re: a.out Pierpaolo Bernardi 07 Jul 2016 07:43 UTC

On Thu, Jul 7, 2016 at 8:07 AM, Marc Nieper-Wißkirchen
<xxxxxx@nieper-wisskirchen.de> wrote:
> Sven Hartrumpf <xxxxxx@gmx.net> schrieb am Mi., 6. Juli 2016 um 23:20
> Uhr:
>>
>> Thanks for this SRFI.
>>
>> Just a minor comment.
>> a.out is the traditional output file of a compiler and is the default
>> in this SRFI.
>> But is not an informative name much more convenient?
>
>
> Any name but a.out is probably more informative, and a.out was just chosen
> for tradition. If we want to use a name that is derived from the program
> source file, we would have to specify an algorithm how the name is derived.
>
> The algorithm could be: Strip off the extension of the source program file
> name and replace it with an extension appropriate for executable files on
> the host system (e.g. no extension on POSIX systems, .exe on Windows).
>
> However, this algorithm may produce the same file name that was fed into the
> algorithm, e.g. if the source program file had no extension at all.

In this unlikely case, maybe it could be mandated that this
coincidence of names is detected, an error is signalled, and the whole
procedure is aborted?