Re: In <E5x7wF.6r7@nonexistent.com>, Jacob Salomon wrote:

Felix Koshy Mathews wrote: } } In <E5x7wF.6r7@nonexistent.com>, Jacob Salomon wrote: } >Hi Folks, } > } . } . } >In every [blessed] 4GL source module in our application, we have } >commented lines that look like: } > } ># $RCSfile: main.4gl,v $ } ># $Date: 1996/08/01 18:21:47 $ } ># $Revision: 1.1 $ } > } >Is there any way in 4GL that we might use these string values in } >variables? We'd like to be able to pass this to the error-handling } >routines. Would another source-code control program make this possible? } . } . } . } } Jacob, } } Why do you want to print version-control information into error-log. If } your executables are writing error messages with exec_name and } func_names wherever necessary, you could use "what exec_name", of UNIX, } to figure out which 4GL file is writing the message and its RCS version. } Isn't that enough to track down the problem. } Of course, only if you define a variable and assign it the version control keywords. Comments are not compiled into the executable, so putting the keywords in comments yields no 4gl version information with the what command (other than the linked libraries, which brings me to point #2. It sucks to wade through the 100+ version numbers output by a compiled 4gl program (due to all of the linked libraries), so some take the easy way out in this manner. Also, it's nice to see the version info in the error log without having to do separate commands. This is particularly helpful when combined with tools that provide centralized help-desk error reporting of multiple systems to a separate help-desk system. Tim