[Emacs-diffs] /srv/bzr/emacs/trunk r100513: Remove NO_SOCK_SIGIO references.

[Glenn Morris]

------------------------------------------------------------
revno: 100513
committer: Glenn Morris <address@hidden>
branch nick: trunk
timestamp: Wed 2010-06-02 22:50:48 -0700
message:
  Remove NO_SOCK_SIGIO references.
  
  * src/m/template.h (NO_SOCK_SIGIO): Remove, no longer used.
  
  * nextstep/DEV-NOTES: Remove out-of-date information.
modified:
  nextstep/DEV-NOTES
  src/ChangeLog
  src/m/template.h

=== modified file 'nextstep/DEV-NOTES'
--- a/nextstep/DEV-NOTES        2009-02-10 03:28:13 +0000
+++ b/nextstep/DEV-NOTES        2010-06-03 05:50:48 +0000
@@ -64,13 +64,8 @@
 
 Currently ctrl-g is not detected in as many circumstances as other emacsen.
 It is not certain whether this is due to the means of event loop integration,
-or errors of omission in the NS code.  One area for exploration is the
-NO_SOCK_SIGIO define.  When it is defined, ctrl-g seems to be picked up more
-often, but there are some annoying side effects.  Currently it is left off by
-default, unless the --enable-cocoa-experimental-ctrl-g option is passed to
-configure [option removed Feb 2009].  (Has no effect under GNUstep.)
-This is an area for improvement.  Also, see the article here and its
-containing thread:
+or errors of omission in the NS code.  This is an area for improvement.
+Also, see the article here and its containing thread:
 
 http://article.gmane.org/gmane.emacs.devel/92021/match=handling%5fsignal
 

=== modified file 'src/ChangeLog'
--- a/src/ChangeLog     2010-06-03 05:41:49 +0000
+++ b/src/ChangeLog     2010-06-03 05:50:48 +0000
@@ -1,5 +1,7 @@
 2010-06-03  Glenn Morris  <address@hidden>
 
+       * m/template.h (NO_SOCK_SIGIO): Remove, no longer used.
+
        * m/hp800.h (alloca) [__NetBSD__ && __GNUC__]: No need to define it,
        now that AH_BOTTOM does it.
 

=== modified file 'src/m/template.h'
--- a/src/m/template.h  2010-05-18 02:59:15 +0000
+++ b/src/m/template.h  2010-06-03 05:50:48 +0000
@@ -58,17 +58,6 @@
    code will not be sharable; but that's better than failing completely.  */
 #define NO_REMAP
 
-/* Some really obscure 4.2-based systems (like Sequent DYNIX)
-   do not support asynchronous I/O (using SIGIO) on sockets,
-   even though it works fine on tty's.  If you have one of
-   these systems, define the following, and then use it in
-   config.h (or elsewhere) to decide when (not) to use SIGIO.
-
-   You'd think this would go in an operating-system description file,
-   but since it only occurs on some, but not all, BSD systems, the
-   reasonable place to select for it is in the machine description file.  */
-#define NO_SOCK_SIGIO
-
 /* After adding support for a new machine, modify the large case
    statement in configure.in to recognize reasonable
    configuration names, and add a description of the system to