X-Git-Url: http://ncurses.scripts.mit.edu/?p=ncurses.git;a=blobdiff_plain;f=INSTALL;h=7917e299abae0beff72f82b374bd27531afe5482;hp=a329b7835381438d36200cc747f83a45a7806ef4;hb=6a5e1ae3015763934a358df2680a77458ca3e342;hpb=0c9774ef662e2137933ac0c79077eaa9c8981357

diff --git a/INSTALL b/INSTALL
index a329b783..7917e299 100644
--- a/INSTALL
+++ b/INSTALL
@@ -25,7 +25,7 @@
 -- sale, use or other dealings in this Software without prior written        --
 -- authorization.                                                            --
 -------------------------------------------------------------------------------
--- $Id: INSTALL,v 1.116 2007/02/17 21:53:16 tom Exp $
+-- $Id: INSTALL,v 1.124 2008/03/29 18:07:32 tom Exp $
 ---------------------------------------------------------------------
              How to install Ncurses/Terminfo on your system
 ---------------------------------------------------------------------
@@ -149,6 +149,9 @@ INSTALLATION PROCEDURE:
     recent implementation of libiconv.  We have built this configuration on
     Linux using libiconv, sometimes requiring libutf8.
 
+    If you configure using the --with-pthread option, a "t" is appended to
+    the library names (e.g., libncursest.a, libncursestw.a).
+
     If you do not specify any models, the normal and debug libraries will be
     configured.  Typing `configure' with no arguments is equivalent to:
 
@@ -304,6 +307,11 @@ SUMMARY OF CONFIGURE OPTIONS:
 	return deceptive results, so you may have to override the configure
 	script.  Or you may be building tic for a smaller machine.
 
+    --disable-big-strings
+	Disable compile-time optimization of predefined tables which puts
+	all of their strings into a very long string, to reduce relocation
+	overhead.
+
     --disable-database
 	Use only built-in data.  The ncurses libraries normally read terminfo
 	and termcap data from disk.  You can configure ncurses to have a
@@ -334,6 +342,16 @@ SUMMARY OF CONFIGURE OPTIONS:
 	For testing, compile-in code that frees memory that normally would not
 	be freed, to simplify analysis of memory-leaks.
 
+	Any implementation of curses must not free the memory associated with
+	a screen, since (even after calling endwin()), it must be available
+	for use in the next call to refresh().  There are also chunks of
+	memory held for performance reasons.  That makes it hard to analyze
+	curses applications for memory leaks.  To work around this, build
+	a debugging version of the ncurses library which frees those chunks
+	which it can, and provides the _nc_free_and_exit() function to free
+	the remainder on exit.  The ncurses utility and test programs use this
+	feature, e.g., via the ExitProgram() macro.
+
     --disable-lp64
 	The header files will ignore use of the _LP64 symbol to make chtype
 	and mmask_t types 32 bits (they may be long on 64-bit hosts, for
@@ -483,12 +501,23 @@ SUMMARY OF CONFIGURE OPTIONS:
 	may not be accurate, or that your stty settings have disabled the use
 	of tabs.
 
+    --enable-mixed-case
+	Controls whether the filesystem on which the terminfo database resides
+	supports mixed-case filenames (normal for UNIX, but not on other
+	systems).  If you do not specify this option, the configure script
+	checks the current filesystem.
+
     --enable-no-padding
 	Compile-in support for the $NCURSES_NO_PADDING environment variable,
 	which allows you to suppress the effect of non-mandatory padding in
 	terminfo entries.  This is the default, unless you have disabled the
 	extended functions.
 
+    --enable-reentrant
+	Compile experimental configuration which improves reentrant use of the
+	library by reducing global and static variables.  This option is also
+	set if --with-pthread is used.
+
     --enable-rpath
 	Use rpath option when generating shared libraries, and (with some
 	restrictions) when linking the corresponding programs.  This originally
@@ -669,10 +698,16 @@ SUMMARY OF CONFIGURE OPTIONS:
 
 	See also --without-dlsym
 
-    --with-hashed-db
+    --with-hashed-db[=XXX]
 	Use a hashed database for storing terminfo data rather than storing
 	each compiled entry in a separate binary file within a directory
 	tree.
+	
+	In particular, this uses the Berkeley database 1.8.5 interface, as
+	provided by that and its successors db 2, 3, and 4.  The actual
+	interface is slightly different in the successor versions of the
+	Berkeley database.  The database should have been configured using
+	"--enable-compat185".
 
 	If you use this option for configuring ncurses, tic will only be able
 	to write entries in the hashed database.  infocmp can still read
@@ -684,6 +719,12 @@ SUMMARY OF CONFIGURE OPTIONS:
 	You cannot have a directory containing both hashed-database and
 	filesystem-based terminfo entries.
 
+	Use the parameter value to give the install-prefix used for the
+	datbase, e.g.,
+		--with-hashed-db=/usr/local/BigBase
+	to find the corresponding include- and lib-directories under the
+	given directory.
+
 	See also the --enable-getcap option.
 
     --with-install-prefix=XXX
@@ -779,6 +820,11 @@ SUMMARY OF CONFIGURE OPTIONS:
 	Generate profile-libraries These are named by adding "_p" to the root,
 	e.g., libncurses_p.a
 
+    --with-pthread
+	Link with POSIX threads, set --enable-reentrant.  The use_window() and
+	use_screen() functions will use mutex's, allowing rudimentary support
+	for multithreaded applications.
+
     --with-rcs-ids
 	Compile-in RCS identifiers.  Most of the C files have an identifier.