Xref: utzoo comp.sys.hp:3522 comp.lang.pascal:2721
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!samsung!uakari.primate.wisc.edu!xanth!mcnc!rti!bnrunix!rick
From: rick@bnrunix.UUCP (Rick Johns)
Newsgroups: comp.sys.hp,comp.lang.pascal
Subject: Re^2: Separate Compilation With HP/UX Pascal
Message-ID: <421@bnrunix.UUCP>
Date: 22 Nov 89 21:51:05 GMT
References: <416@bnrunix.UUCP> <9110005@hpfcso.HP.COM>
Organization: BNR Inc., RTP, NC
Lines: 39

Sorry to post this but you can't get there from here....

bigelow@hpfcso.HP.COM (Jim Bigelow) writes:

>Have you tried reversing the files in the search path? e.g.,

>	$search 'lib1.a,lib2.a,...,foo.o'$

>This way libraries are used first to resolve imported modules.


Oops, sorry, what I was trying to get at is that under normal circumstances I
want to use the libraries.  However, on occasion I need to change a module in
a library, so I compile it locally, without putting it back into the library.
Then, I want foo.o to be found first.  I guess I could remove it from the 
library, but other people will expect to find it there.  What I really need is
a way for the compiler to say, 'ok, couldn't find foo.o, let's keep looking."

Steve Taylor's reply on comp.sys.hp made me feel better about *why* the $SEARCH 
directive was included, although I'm porting to HP Pascal from a Pascal that
doesn't make it necessary.  However, I still think the $SEARCH directive could 
be smarter.

Thanks for your interest--

Rick Johns

BNR
Research Triangle Park, NC
(919)991-7191
rti!bnrunix!rick@mcnc.org
uunet!mcnc.org!bnrunix!rick@uunet.UU.NET
rjohns@bnr.ca

Obnoxious Quote:
================================================================================
            The word is "its".  "It's" is something else entirely.  
================================================================================
Disclaimer: These opinions are mine and not BNR's.