<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN"> <HTML> <HEAD> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8"> <META NAME="GENERATOR" CONTENT="GtkHTML/3.26.0"> </HEAD> <BODY> On Wed, 2010-12-29 at 15:51 -0500, Trevor Woerner wrote: <BLOCKQUOTE TYPE=CITE> <PRE> For this entire patch series I still can't understand why we need to define an explicit _SET variable which defines whether or not the base variable is defined, or not. For every one of the new variables you have defined you have included the following: > +# States if the user has exported BINDIR > +if [ X"$BINDIR" != X ]; then > + BINDIR_SET=yes > +fi And then you subsequently check to see if (using the specific variables for this specific patch) BINDIR_SET is defined to decide whether or not to use the value of BINDIR. But why? If BINDIR is not defined then BINDIR_SET will not be defined. So if you want to check whether BINDIR is defined why not just check to see if BINDIR is defined (instead of creating a new variable (BINDIR_SET) and checking whether it is defined or not)? Maybe there's something about what you're doing that I don't understand. But if you're defining BINDIR_SET only if BINDIR is defined and then using BINDIR_SET to indicate whether or not BINDIR is defined then I think you're adding complexity for no reason. PS. I have nothing against _what_ is being done, just the way in which it is being done. </PRE> </BLOCKQUOTE> I think I found a way to expose the need for this _SET variable. Perhaps you can then help me make it clearer as others will most likely have the same question, or risk breaking it when maintaining the script. <BLOCKQUOTE> <PRE> # States if the user has exported BINDIR if [ X"$BINDIR" != X ]; then     BINDIR_SET=yes fi </PRE> </BLOCKQUOTE> The comment is important. This variable is to tell us if the user has exported an environment variable called BINDIR, not to tell if it is empty or not. <BLOCKQUOTE> <PRE> # Set the default value for BINDIR if [ X"$BINDIR_SET" = X ]; then     BINDIR=$EPREFIX/bin fi </PRE> </BLOCKQUOTE> Th above part says: if the user has not exported the BINDIR variable, then assign it a default value. So far, no need for BINDIR_SET, I agree. Note that from now on, BINDIR is always set, always. <PRE>     # Set the path so that locally built apps will be found and used     PATH=${DESTDIR}${BINDIR}${PATH+:$PATH}     export PATH </PRE> The above line needs BINDIR to hold a value, always. <PRE>     ${BINDIR_SET:+--bindir="$BINDIR"} </PRE> This line above says: if the user has exported the BINDIR variable, add "--bindir=$BINDIR" to the configure command line. It also says: do not add --bindir if the user has not exported the BINDIR environment variable. Recall that BINDIR will always have a value, always. There may be a more elegant way of handling this situation. The variable should perhaps be names BINDIR_USER_SET. In the case of PREFIX, it can bet set by the user as an env var or on the command line. There are 2 places where PREFIX_SET is set. I appreciate you are not giving up, having code that works is good, having maintainable code that works is even better. Gaetan </BODY> </HTML>