Re: Final version of newmake.bat

[Kevin Greiner <kgreiner@xpediantsolutions.com>]

Frank Schmitt <usereplyto@Frank-Schmitt.net> writes:

> Kevin Greiner <kgreiner@xpediantsolutions.com> writes:
>
>> The compilation buffer ends with
>> "Compilation exited abnormally with code 1 at Wed Mar 19 08:51:57"
>
> No idea about this one, however this is an Emacs message and I don't
> think we can blame newmake.bat for this. This one comes before the
> processing of info files, does it?

No. This is the last line in the compilation buffer.  Emacs is
reporting the process status returned by the external process just
executed (i.e. newmake).  Emacs follows the Unix convention that a
zero status indicates normal completion while a non-zero status
indicates an error.


>>> The error message above means that newmake.bat couldn't find the files
>>> message-1 and message-2 in the texi directory so it seems that
>>> generation of info files failed.
>>
>> It appears, to me at least, that message.texi generated only one file "message".
>
> Indeed, if infohack.el is used to create the info files, only message
> is created, while makeinfo creates message, message-1 and message-2.
>
> This is weird and looks like a bug to me, but again I don't think that
> newmake.bat is responsible but either infohack.el or message.texi.

Bug is probably too harsh of a word.  Infohack simply didn't format the
message file exactly like makeinfo.

I commented out the xcopy ... message-? line without changing the exit
status.

Adding 'edit /B 0' to the end of newmake fixes the exitlevel.  The
compilation buffer now ends with
  "Compilation finished at Wed Mar 19 18:22:46".

However, I understand that that syntax wasn't added until Windows
2000.

Kevin