Konstantin Vlasov wrote:

> I'm not yet sure how it happened that ERRORLEVEL was non-zero by the time
> gen_resp_file.bat was started. Even when it is 0 before I run nmake, it becomes
> 1 when the $(MODE) rule starts to execute. But even putting that question aside,
> this is not right to check an exit code from a script that does not change the
> exit code at all.

From:
    https://docs.microsoft.com/en-us/windows-server/administration/windows-commands/setlocal

   The setlocal command sets the ERRORLEVEL variable. If you
   pass {enableextensions | disableextensions} or
   {enabledelayedexpansion | disabledelayedexpansion}, the ERRORLEVEL
   variable is set to 0 (zero).

Seems related to your issue.

So try adding a 'setlocal enableextensions' at the start somewhere.

(untested since I do not use CMD. But rather JPsoft's
far superior 4NT).

-- 
--gv
-------------------------------------------------------------------
Unsubscribe: https://cool.haxx.se/list/listinfo/curl-library
Etiquette:   https://curl.haxx.se/mail/etiquette.html