Mercurial > hg > xemacs-beta
annotate src/s/isc2-2.h @ 1648:712931b4b71d
[xemacs-hg @ 2003-08-27 18:06:54 by youngs]
2003-08-28 Steve Youngs <youngs@xemacs.org>
* README.packages: Update.
2003-08-28 Steve Youngs <youngs@xemacs.org>
* PACKAGES: Update.
2003-08-28 Steve Youngs <youngs@xemacs.org>
* xemacs-faq.texi (Q2.0.2): Rewrite, mentioning the correct way to
remove a package.
(Q3.8.2): big-menubar is in the edit-utils package.
(Q4.3.2): Add a comment about not needing TM for things like Gnus,
MH-E and VM.
(Q5.3.3): State correct location of ps-print.el.
* xemacs/packages.texi (Packages): Remove "Creating Packages" menu
entry.
(Package Terminology): Whitespace clean up.
(Installing Packages): Whitespace clean up and add some @code
formatters.
Re-organise the menu so that installation via PUI is first and
Sumo is last.
(Automatically): mule-base is no longer a requirement for using
PUI.
Mention optionally requiring mailcrypt.
(Note): Removed.
(Manually): Move to below the PUI installation method.
(Sumo): Move to below the manual installation method.
(Which Packages): Add mailcrypt.
(Building Packages): Remove duplicated stuff that is in
lispref/packaging.texi, xref to it instead.
(Local.rules File): xref to the appropriate node in
lispref/packaging.texi.
(Available Packages): Update to current reality.
(all): Removed.
(srckit): Removed.
(binkit): Removed.
* xemacs/reading.texi (Reading Mail): Mention Gnus and MEW.
* new-users-guide/custom2.texi (Init File): big-menubar.el is in
the edit-utils package.
* lispref/packaging.texi (Packaging):
(The User View):
(The Library Maintainer View):
(Infrastructure):
(Control Files):
(Obtaining):
(The Package Release Engineer View):
(Package Terminology):
(Building Packages):
(Makefile Targets):
(packages): New.
(Local.rules File):
(XEMACS_PACKAGES): Removed.
(XEMACS_INSTALLED_PACKAGES_ROOT): New.
(NONMULE_PACKAGES): New.
(EXCLUDES): New.
(Creating Packages):
(BATCH): New.
(VERSION): Removed.
(AUTHOR_VERSION): Removed.
(MAINTAINER): Removed.
(PACKAGE): Removed.
(PKG_TYPE): Removed.
(REQUIRES): Removed.
(CATEGORY): Removed.
(ELS): Removed.
(ELCS): Removed.
(all): Removed.
(srckit): Removed.
(binkit): Removed.
(are): New.
(STANDARD_DOCS): New.
(ELCS_1_DEST): New.
(example): New.
(PACKAGE_SUPPRESS): New.
(EXPLICIT_DOCS): New.
(DATA_DEST): New.
(Documenting Packages):
Not quite a total rewrite, but a fairly thorough audit
nonetheless.
| author | youngs |
|---|---|
| date | Wed, 27 Aug 2003 18:07:10 +0000 |
| parents | c69610198c35 |
| children |
| rev | line source |
|---|---|
| 428 | 1 /* Synched up with: FSF 19.31. */ |
| 2 | |
| 3 /* system description file for Interactive (ISC) Unix version 2.2 on | |
| 4 the 386. */ | |
| 5 | |
| 6 #include "usg5-3.h" | |
| 7 | |
| 8 /* select (in -linet) works okay on X ptys, but not on the serial port. | |
| 9 karl@cs.umb.edu says that with that select call, subprocesses made by | |
| 10 (e.g.) M-x grep don't exit cleanly, they just hang. Similar problems | |
| 11 have been observed in ISC 3.0. */ | |
| 12 #define BROKEN_SELECT_NON_X | |
| 13 | |
| 14 /* karl@cs.umb.edu says that ISC's socket support (in -linet) isn't | |
| 15 what Emacs needs; it makes interrupt-shell-subjob and the like do | |
| 16 nothing. But that appears to have been another manifestation of | |
| 17 the broken select, so it should now be safe to define this again. */ | |
| 18 #define HAVE_SOCKETS | |
| 19 | |
| 20 #define NO_SOCKETS_IN_FILE_SYSTEM | |
| 21 #define NEED_NET_ERRNO_H | |
| 22 | |
| 23 /* This keeps the .cdbx section that gcc puts out when generating | |
| 24 stabs-in-coff output, so Emacs can be debugged. --karl@cs.umb.edu. */ | |
| 25 #define USG_SHARED_LIBRARIES | |
| 26 | |
| 27 #define NO_FCHMOD | |
| 28 | |
| 29 #define MAXNAMLEN 512 | |
| 30 #define O_NDELAY O_NONBLOCK | |
| 31 #define MEMORY_IN_STRING_H | |
| 32 | |
| 33 /* Tell gmalloc.c that we don't have memmove (system include files to the | |
| 34 contrary!). */ | |
| 35 #define MEMMOVE_MISSING | |
| 36 | |
| 37 /* -lPW is only needed if not using Gcc. We used to include -lcposix here | |
| 38 for the rename function, but some people say ISC's rename doesn't | |
| 39 work correctly with Emacs so we use Emacs' emulation instead. */ | |
| 40 #if defined (__GNUC__) | |
| 41 # define LIB_STANDARD_1 -lcposix | |
| 42 #else /* !__GNUC__ */ | |
| 43 # define LIB_STANDARD_1 -lPW | |
| 44 #endif /* !__GNUC__ */ | |
| 45 | |
| 46 /* LIB_STANDARD_1 is used both here and in LIBS_SYSTEM | |
| 47 (the latter for the sake of configure). */ | |
| 48 #define LIB_STANDARD LIB_STANDARD_1 -lc | |
| 49 | |
| 50 #define NO_X_DESTROY_DATABASE | |
| 51 | |
| 52 /* -linet may be needed to avoid undefined symbols such as gethostname, | |
| 53 inet_addr, gethostbyname, socket, connect, ... */ | |
| 54 #define LIBS_SYSTEM -linet LIB_STANDARD_1 | |
| 55 | |
| 56 /* Inhibit asm code in netinet/in.h. Strictly speaking, only necessary | |
| 57 when -traditional is being used, but it doesn't hurt to | |
| 58 unconditionally define this. */ | |
| 59 #define NO_ASM | |
| 60 | |
| 61 /* -traditional is not necessary if the system header files are fixed to | |
| 62 define getc and putc in the absence of _POSIX_SOURCE. GCC's from 2.4.4 | |
| 63 on do this. */ | |
| 64 #if !defined (__GNUC__) || __GNUC__ < 2 | |
| 65 # define C_SWITCH_SYSTEM "-traditional" | |
| 66 #endif | |
| 67 | |
| 68 /* Some versions of ISC are said to define S_IFLNK even tho | |
| 69 they don't really support symlinks. */ | |
| 70 #undef S_IFLNK |