]> asedeno.scripts.mit.edu Git - PuTTY.git/log
PuTTY.git
21 years agoTentative merge of ben-mac-port (only dead for three years!) into the trunk.
Ben Harris [Tue, 19 Nov 2002 02:13:46 +0000 (02:13 +0000)]
Tentative merge of ben-mac-port (only dead for three years!) into the trunk.
This doesn't include any mkfiles.pl glue, and is missing one or two other
fixes.  The terminal emulator is kind of working, though, as, I believe, is
the store module.  Everything else is yet to be done.

[originally from svn r2226]

21 years agoWhen processing input, refer to CR and LF as \015 and \012 respectively,
Ben Harris [Mon, 18 Nov 2002 23:49:30 +0000 (23:49 +0000)]
When processing input, refer to CR and LF as \015 and \012 respectively,
rather than \r and \n.  The latter tend to get swapped around by Mac C
compilers.

[originally from svn r2225]

21 years agoISO-IR entry 157 appears to correspond to ISO/IEC 8859-10, and lists
Ben Harris [Mon, 18 Nov 2002 22:27:25 +0000 (22:27 +0000)]
ISO-IR entry 157 appears to correspond to ISO/IEC 8859-10, and lists
code point 3/13 as HORIZONTAL BAR, which agrees with unicode.org's mapping
table.  Change ours to match (it used to have EM DASH, courtesy of RDB).
This brings all our 8859-to-Unicode tables into line with unicode.org.

[originally from svn r2224]

21 years agoMore comment cleanup.
Ben Harris [Mon, 18 Nov 2002 20:06:46 +0000 (20:06 +0000)]
More comment cleanup.

[originally from svn r2223]

21 years agoDon't forget the IEC!
Ben Harris [Mon, 18 Nov 2002 18:24:35 +0000 (18:24 +0000)]
Don't forget the IEC!

[originally from svn r2222]

21 years agoBoth ISO and unicode.org now admit the existence of 8859-11 (:2001, FWIW),
Ben Harris [Mon, 18 Nov 2002 18:14:18 +0000 (18:14 +0000)]
Both ISO and unicode.org now admit the existence of 8859-11 (:2001, FWIW),
and unicode.org agrees with our mapping table.  Update text accordingly.

[originally from svn r2221]

21 years agoThanks to Hans-Juergen Petrich for spotting this tiny memory leak:
Simon Tatham [Sun, 17 Nov 2002 15:06:16 +0000 (15:06 +0000)]
Thanks to Hans-Juergen Petrich for spotting this tiny memory leak:
`otherbuf' should still be freed even if the RegEnumKey function
that was supposed to fill it with data failed. While I'm at it, also
remove the redundant check for its non-NULL-ness (what's the point
of having a malloc wrapper that dies rather than return NULL if you
then waste effort checking its return value for NULL _anyway_, eh?).

[originally from svn r2217]

21 years agoApple's C compilers don't think that putting parentheses around assignments
Ben Harris [Sun, 17 Nov 2002 02:00:06 +0000 (02:00 +0000)]
Apple's C compilers don't think that putting parentheses around assignments
in "if" conditions is enough.  Use an actual comparison against NULL instead.

[originally from svn r2216]

21 years agoChange a temporary buffer from unsigned chars to chars, avoiding a pointer-cast
Ben Harris [Sun, 17 Nov 2002 01:56:01 +0000 (01:56 +0000)]
Change a temporary buffer from unsigned chars to chars, avoiding a pointer-cast
warning.

[originally from svn r2215]

21 years agoBump latest version to 0.53b for release!
Simon Tatham [Tue, 12 Nov 2002 18:55:41 +0000 (18:55 +0000)]
Bump latest version to 0.53b for release!

[originally from svn r2210]

21 years agoUse <stddef.h> to get wchar_t, rather than <wchar.h> (or nothing, in putty.h).
Ben Harris [Sun, 10 Nov 2002 00:03:55 +0000 (00:03 +0000)]
Use <stddef.h> to get wchar_t, rather than <wchar.h> (or nothing, in putty.h).
Both are required to contain wchar_t in C99, but only <stddef.h> does in the
version of MPW I've got here.

[originally from svn r2206]

21 years agoRename CharWidth() to char_width(). The former name clashes with an API
Ben Harris [Sat, 9 Nov 2002 21:46:21 +0000 (21:46 +0000)]
Rename CharWidth() to char_width().  The former name clashes with an API
function in Mac OS.

[originally from svn r2205]

21 years agoImprove robustness in random seed file handling.
Simon Tatham [Thu, 7 Nov 2002 20:01:04 +0000 (20:01 +0000)]
Improve robustness in random seed file handling.

[originally from svn r2200]

21 years agoRobustness fixes for KEXINIT handling and others. In particular, I've
Simon Tatham [Thu, 7 Nov 2002 19:49:03 +0000 (19:49 +0000)]
Robustness fixes for KEXINIT handling and others. In particular, I've
created a self-mallocing variant of sprintf, to obviate any future
need for paranoid %.100s type stuff in format strings.

[originally from svn r2199]

21 years agoBorland makefile needs to define _WINDOWS; apparently this makefile
Simon Tatham [Thu, 7 Nov 2002 19:37:36 +0000 (19:37 +0000)]
Borland makefile needs to define _WINDOWS; apparently this makefile
hasn't worked since the Unix port started and nobody has noticed
until now :-)

[originally from svn r2198]

21 years agoPacket-level logging should now work properly in Unix Plink.
Simon Tatham [Tue, 5 Nov 2002 13:21:03 +0000 (13:21 +0000)]
Packet-level logging should now work properly in Unix Plink.

[originally from svn r2193]

21 years agoFix command-line error handling in pterm. (Hint: wrapping a multi-
Simon Tatham [Tue, 5 Nov 2002 13:20:42 +0000 (13:20 +0000)]
Fix command-line error handling in pterm. (Hint: wrapping a multi-
statement macro in `do ... while (0)' and putting a `continue'
within it don't go well together. Oops.)

[originally from svn r2192]

21 years agoSubstitute `-' (magical hyphen) for `\-' (inert minus sign) in the
Simon Tatham [Tue, 5 Nov 2002 09:37:31 +0000 (09:37 +0000)]
Substitute `-' (magical hyphen) for `\-' (inert minus sign) in the
manpage. Fixes Debian bug #167761.

[originally from svn r2190]

21 years agoHalf of Lars Gunnarsson's iXplorer compatibility patch: the PSFTP
Simon Tatham [Sun, 3 Nov 2002 08:46:35 +0000 (08:46 +0000)]
Half of Lars Gunnarsson's iXplorer compatibility patch: the PSFTP
login prompt should be fflushed (presumably fgets fails to implicitly
do this when stdin and stdout are redirected weirdly).

[originally from svn r2186]

21 years agoProbably about time we mentioned the nascent Unix port in the FAQ.
Simon Tatham [Sat, 2 Nov 2002 16:27:17 +0000 (16:27 +0000)]
Probably about time we mentioned the nascent Unix port in the FAQ.
Not that I desperately want to shout about it just yet, but I feel a
bit bad about the FAQ saying `we don't have a Unix port, anyone who
told you so was wrong'. :-)

[originally from svn r2185]

21 years agoImprove shadow bold mode: set the default shadow bold offset to +1
Simon Tatham [Sat, 2 Nov 2002 16:16:35 +0000 (16:16 +0000)]
Improve shadow bold mode: set the default shadow bold offset to +1
not -1 (it turns out _most_ X fonts prefer the former, though
irritatingly my favourite real X font used to prefer the latter
which was why I made the X version of my Font Of Choice do so too),
and also clip to the boundaries of the rectangle we should be
drawing text in. This still doesn't completely prevent display
corruption in the case where text drawn in one sweep is partially
overwritten in a future one, but gnome-terminal has this problem
too, and now we've got the right default SB offset _and_ offer the
opportunity to reconfigure it I think this is pretty good for now.

[originally from svn r2184]

21 years agoAnother signal-handling refinement from RJK: the SIGCHLD handler
Simon Tatham [Sat, 2 Nov 2002 16:05:26 +0000 (16:05 +0000)]
Another signal-handling refinement from RJK: the SIGCHLD handler
should be prepared to reap more than one child per invocation if
necessary, since we do after all have two.

[originally from svn r2183]

21 years agoImprove the noise collection for the internal random pool.
Simon Tatham [Sat, 2 Nov 2002 15:23:20 +0000 (15:23 +0000)]
Improve the noise collection for the internal random pool.

[originally from svn r2182]

21 years agoRJK's general signal-handling robustness patch. Should fix the weird
Simon Tatham [Sat, 2 Nov 2002 14:35:57 +0000 (14:35 +0000)]
RJK's general signal-handling robustness patch. Should fix the weird
spin behaviour occasionally seen after pterm's child process dies.

[originally from svn r2181]

21 years agos/public/private/ spotted by Clint Hastings; also change a "public key"
Jacob Nevins [Fri, 1 Nov 2002 21:50:35 +0000 (21:50 +0000)]
s/public/private/ spotted by Clint Hastings; also change a "public key"
to "key pair"

[originally from svn r2180]

21 years agoIntroduce program category U, for non-GTK-requiring Unix apps. Plink
Simon Tatham [Fri, 1 Nov 2002 18:51:55 +0000 (18:51 +0000)]
Introduce program category U, for non-GTK-requiring Unix apps. Plink
doesn't need to be linked with libgtk, libgdk, libX11 etc!

[originally from svn r2179]

21 years agoPrevent another segfault. Oops.
Simon Tatham [Fri, 1 Nov 2002 18:51:22 +0000 (18:51 +0000)]
Prevent another segfault. Oops.

[originally from svn r2178]

21 years ago.HLP files should be ignored as well. Oops.
Simon Tatham [Fri, 1 Nov 2002 13:49:25 +0000 (13:49 +0000)]
.HLP files should be ignored as well. Oops.

[originally from svn r2177]

21 years agoAdd a .cvsignore for the Unix subdir
Simon Tatham [Fri, 1 Nov 2002 13:37:39 +0000 (13:37 +0000)]
Add a .cvsignore for the Unix subdir

[originally from svn r2176]

21 years agoImprove handling of oobinline sockets; Plink in telnet mode now
Simon Tatham [Fri, 1 Nov 2002 13:36:48 +0000 (13:36 +0000)]
Improve handling of oobinline sockets; Plink in telnet mode now
doesn't hang when you hit ^C, which is nice. I think a better
solution would involve nonblocking sockets; as it stands it's a
little dependent on what may be quirks of the Linux socket layer.

[originally from svn r2175]

21 years agoAnother valgrind-caught error. This one has apparently been there
Simon Tatham [Fri, 1 Nov 2002 13:01:14 +0000 (13:01 +0000)]
Another valgrind-caught error. This one has apparently been there
since the Dawn O' Time, and consisted of me putting the two halves
of a short-circuiting bounds check the wrong way round: instead of
`p_in_range && *p', I had `*p && p_in_range'. Oops. valgrind rocks.

[originally from svn r2174]

21 years agovalgrind has caught two more uninitialised elements in the SSH
Simon Tatham [Fri, 1 Nov 2002 12:59:09 +0000 (12:59 +0000)]
valgrind has caught two more uninitialised elements in the SSH
context structure. I knew the Unix port would be a good idea!

[originally from svn r2173]

21 years agoImplement access to the SSH agent. This ought to make agent
Simon Tatham [Fri, 1 Nov 2002 12:55:13 +0000 (12:55 +0000)]
Implement access to the SSH agent. This ought to make agent
forwarding work as well, of course.

[originally from svn r2172]

21 years agoReplace a \r\n with an ordinary \n - this isn't Windows!
Simon Tatham [Fri, 1 Nov 2002 12:54:51 +0000 (12:54 +0000)]
Replace a \r\n with an ordinary \n - this isn't Windows!

[originally from svn r2171]

21 years agoStop the segfault on failure to resolve a host name.
Simon Tatham [Fri, 1 Nov 2002 12:54:22 +0000 (12:54 +0000)]
Stop the segfault on failure to resolve a host name.

[originally from svn r2170]

21 years agoAdd a \n at the end of an error message - oops.
Simon Tatham [Fri, 1 Nov 2002 12:54:03 +0000 (12:54 +0000)]
Add a \n at the end of an error message - oops.

[originally from svn r2169]

21 years agoMove an output newline to make username entry in Plink/SSH2 look
Simon Tatham [Fri, 1 Nov 2002 12:53:45 +0000 (12:53 +0000)]
Move an output newline to make username entry in Plink/SSH2 look
more sensible.

[originally from svn r2168]

21 years agoPass int rather than char to va_arg (the latter gets promoted).
Simon Tatham [Fri, 1 Nov 2002 12:53:10 +0000 (12:53 +0000)]
Pass int rather than char to va_arg (the latter gets promoted).

[originally from svn r2167]

21 years agoSplit a trigraph sequence in a string constant.
Simon Tatham [Fri, 1 Nov 2002 12:52:51 +0000 (12:52 +0000)]
Split a trigraph sequence in a string constant.

[originally from svn r2166]

21 years agoFirst attempt at a Unix port of Plink. Seems to basically work;
Simon Tatham [Thu, 31 Oct 2002 19:49:52 +0000 (19:49 +0000)]
First attempt at a Unix port of Plink. Seems to basically work;
doesn't yet use the SSH agent, no way to specify arbitrary config
options, no manpage yet, couple of other fiddly things need doing,
but it makes SSH connections and doesn't fall over horribly so I say
it's a good start. Now to run it under valgrind...

[originally from svn r2165]

21 years agoRJK's `pterm --help' patch. I _must_ find a better alternative to
Simon Tatham [Wed, 30 Oct 2002 18:22:37 +0000 (18:22 +0000)]
RJK's `pterm --help' patch. I _must_ find a better alternative to
this init sequence - it surely can't be right that `pterm --help'
with no DISPLAY complains at the lack of DISPLAY rather than giving
a help message!

[originally from svn r2164]

21 years agoRJK's `make install' patch for the Unix makefile.
Simon Tatham [Wed, 30 Oct 2002 18:17:56 +0000 (18:17 +0000)]
RJK's `make install' patch for the Unix makefile.

[originally from svn r2163]

21 years agoFurther deglobalisation: settings.c now has a more sensible interface.
Simon Tatham [Wed, 30 Oct 2002 18:12:46 +0000 (18:12 +0000)]
Further deglobalisation: settings.c now has a more sensible interface.

[originally from svn r2162]

21 years agoFix some consts.
Simon Tatham [Wed, 30 Oct 2002 18:12:22 +0000 (18:12 +0000)]
Fix some consts.

[originally from svn r2161]

21 years agoMore preparatory work: remove the <windows.h> include from lots of
Simon Tatham [Wed, 30 Oct 2002 17:57:31 +0000 (17:57 +0000)]
More preparatory work: remove the <windows.h> include from lots of
source files in which it's no longer required (it was previously
required in anything that included <putty.h>, but not any more).
Also moved a couple of stray bits of exposed WinSock back into
winnet.c (getservbyname from ssh.c and AF_INET from proxy.c).

[originally from svn r2160]

21 years agoPreparatory work before attempting a Unix port of plink: mkfiles.pl
Simon Tatham [Wed, 30 Oct 2002 17:56:05 +0000 (17:56 +0000)]
Preparatory work before attempting a Unix port of plink: mkfiles.pl
needs to be able to handle separate Recipe entries for the same
program with different types (plink [C] and plink [X] for example,
with different object lists).

[originally from svn r2159]

21 years agoYikes - put back a #include I accidentally removed two revs ago!
Simon Tatham [Tue, 29 Oct 2002 16:54:09 +0000 (16:54 +0000)]
Yikes - put back a #include I accidentally removed two revs ago!
That'd have been embarrassing. Mind you I'd like to remove this
particular #include for good anyway, but the time is not now...

[originally from svn r2158]

21 years agoHmm. Probably best initialise mainchan as well; oops.
Simon Tatham [Tue, 29 Oct 2002 14:41:10 +0000 (14:41 +0000)]
Hmm. Probably best initialise mainchan as well; oops.

[originally from svn r2157]

21 years agoTry actually initialising ssh->v1_compressing. With any luck this
Simon Tatham [Tue, 29 Oct 2002 13:05:20 +0000 (13:05 +0000)]
Try actually initialising ssh->v1_compressing. With any luck this
will be what was causing Owen's crash report today.

[originally from svn r2156]

21 years agoterminal.c should call the frontend beep() routine even with
Simon Tatham [Mon, 28 Oct 2002 21:58:07 +0000 (21:58 +0000)]
terminal.c should call the frontend beep() routine even with
mode==BELL_VISUAL, otherwise taskbar flashing won't happen on visual
bells. It's up to the frontend routine to spot BELL_VISUAL and avoid
making any noise.

[originally from svn r2155]

21 years agoCloseOnExit now defaults to COE_ALWAYS for pterm, bringing it back
Simon Tatham [Mon, 28 Oct 2002 17:39:21 +0000 (17:39 +0000)]
CloseOnExit now defaults to COE_ALWAYS for pterm, bringing it back
into line with most other xtermalikes. On Unix, the exit code of a
shell is the last exit code of one of its child processes, even if
it's an interactive shell - so some pterms will close and some will
not for no particularly good reason. Power-detaching a screen
session is especially bad for this.
COE_NORMAL is still useful for specialist purposes (running a single
command in its own pterm), but I don't think it's a sane default,
unfortunately.

[originally from svn r2154]

21 years agoFix Alt+numberpad in pterm. For a start, there was a bug whereby if
Simon Tatham [Mon, 28 Oct 2002 17:34:45 +0000 (17:34 +0000)]
Fix Alt+numberpad in pterm. For a start, there was a bug whereby if
a pterm came up while Alt was down, then releasing it would cause a
^@ to be generated. Also, though, I've decided that Alt plus a
single numberpad key should not generate a low-numbered control
code, because that's too easy to do by mistake and the codes are too
powerful. Anyone who really _wants_ to create a ^C or ^D from the
numberpad can do Alt-03 or Alt-04 easily enough; two-digit codes and
more such as Alt-65 are unaffected.

[originally from svn r2153]

21 years agoAnother deglobalisation bug: failed to initialise term->curstype.
Simon Tatham [Mon, 28 Oct 2002 17:30:47 +0000 (17:30 +0000)]
Another deglobalisation bug: failed to initialise term->curstype.
Thanks to valgrind for finding this one (aha! I knew there would be
benefits from doing a Unix port!).

[originally from svn r2152]

21 years agoFirst bug discovered as a result of global-removal: pasting into
Simon Tatham [Mon, 28 Oct 2002 09:38:28 +0000 (09:38 +0000)]
First bug discovered as a result of global-removal: pasting into
pterm caused a crash because I had the wrong prototype for the
selection_received event handler. Should be fixed.

[originally from svn r2150]

21 years agoArrgh! _Another_ missed commit. I really must do something about
Simon Tatham [Sun, 27 Oct 2002 09:24:47 +0000 (09:24 +0000)]
Arrgh! _Another_ missed commit. I really must do something about
this bad habit of mine :-/

[originally from svn r2149]

21 years agoRetire another global in favour of adding a feature to the
Simon Tatham [Sat, 26 Oct 2002 14:06:52 +0000 (14:06 +0000)]
Retire another global in favour of adding a feature to the
terminal.c interface.

[originally from svn r2148]

21 years agoYet more global-removal. The static variables in logging.c are now
Simon Tatham [Sat, 26 Oct 2002 12:58:13 +0000 (12:58 +0000)]
Yet more global-removal. The static variables in logging.c are now
absent, and also (I think) all the frontend request functions (such
as request_resize) take a context pointer, so that multiple windows
can be handled sensibly. I wouldn't swear to this, but I _think_
that only leaves the Unicode stuff as the last stubborn holdout.

[originally from svn r2147]

21 years agoRichard points out that it would probably help if I committed
Simon Tatham [Sat, 26 Oct 2002 12:42:07 +0000 (12:42 +0000)]
Richard points out that it would probably help if I committed
ldisc.h. Bah. :-)

[originally from svn r2146]

21 years agoX forwarding authentication is now invented on a per-SSH-connection
Simon Tatham [Sat, 26 Oct 2002 11:23:15 +0000 (11:23 +0000)]
X forwarding authentication is now invented on a per-SSH-connection
basis, so the statics are gone from x11fwd.c.

[originally from svn r2145]

21 years agoFix proxy.c so that the static variables become const.
Simon Tatham [Sat, 26 Oct 2002 11:12:24 +0000 (11:12 +0000)]
Fix proxy.c so that the static variables become const.

[originally from svn r2144]

21 years agoReorganised the Unicode layer somewhat: moved luni_send and
Simon Tatham [Sat, 26 Oct 2002 11:08:59 +0000 (11:08 +0000)]
Reorganised the Unicode layer somewhat: moved luni_send and
lpage_send out into the line discipline, making them _clients_ of
the Unicode layer rather than part of it. This means they can access
ldisc->term, which in turn means I've been able to remove the
temporary global variable `term'. We're slowly getting there.

[originally from svn r2143]

21 years agoPort forwarding module now passes backend handles around properly.
Simon Tatham [Sat, 26 Oct 2002 10:33:59 +0000 (10:33 +0000)]
Port forwarding module now passes backend handles around properly.
As a result I've now been able to turn the global variables `back'
and `backhandle' into module-level statics in the individual front
ends. Now _that's_ progress!

[originally from svn r2142]

21 years agoLine discipline module now uses dynamically allocated data. Also
Simon Tatham [Sat, 26 Oct 2002 10:16:19 +0000 (10:16 +0000)]
Line discipline module now uses dynamically allocated data. Also
fixed one or two other minor problems.

[originally from svn r2141]

21 years agoOops; remembering to call term_provide_resize_fn in the Unix front
Simon Tatham [Fri, 25 Oct 2002 22:00:22 +0000 (22:00 +0000)]
Oops; remembering to call term_provide_resize_fn in the Unix front
end would probably help. Thanks Colin.

[originally from svn r2140]

21 years agoThe Zlib module now uses dynamically allocated contexts. I think
Simon Tatham [Fri, 25 Oct 2002 13:26:33 +0000 (13:26 +0000)]
The Zlib module now uses dynamically allocated contexts. I think
that completes the static-removal in the crypto library. Ooh.

[originally from svn r2136]

21 years agoDiffie-Hellman key exchange now uses a dynamically allocated context.
Simon Tatham [Fri, 25 Oct 2002 13:08:01 +0000 (13:08 +0000)]
Diffie-Hellman key exchange now uses a dynamically allocated context.

[originally from svn r2135]

21 years agoFix some compiler warnings.
Simon Tatham [Fri, 25 Oct 2002 13:00:45 +0000 (13:00 +0000)]
Fix some compiler warnings.

[originally from svn r2134]

21 years agoTidy up and fix a compiler warning.
Simon Tatham [Fri, 25 Oct 2002 12:59:57 +0000 (12:59 +0000)]
Tidy up and fix a compiler warning.

[originally from svn r2133]

21 years agoSSH CRC attack detector now uses a dynamically allocated context.
Simon Tatham [Fri, 25 Oct 2002 12:58:21 +0000 (12:58 +0000)]
SSH CRC attack detector now uses a dynamically allocated context.

[originally from svn r2132]

21 years agoSSH2 MACs now use dynamically allocated contexts.
Simon Tatham [Fri, 25 Oct 2002 12:51:28 +0000 (12:51 +0000)]
SSH2 MACs now use dynamically allocated contexts.

[originally from svn r2131]

21 years agoSSH ciphers now use dynamically allocated contexts.
Simon Tatham [Fri, 25 Oct 2002 12:35:22 +0000 (12:35 +0000)]
SSH ciphers now use dynamically allocated contexts.

[originally from svn r2130]

21 years agoFix the nasty flashing-light-grey-on-resize problem, after MCV
Simon Tatham [Fri, 25 Oct 2002 11:58:59 +0000 (11:58 +0000)]
Fix the nasty flashing-light-grey-on-resize problem, after MCV
helpfully alerted me to the existence of gdk_window_set_background().

[originally from svn r2129]

21 years agopty backend now supports the changed function interface, so pterm
Simon Tatham [Fri, 25 Oct 2002 11:50:51 +0000 (11:50 +0000)]
pty backend now supports the changed function interface, so pterm
now compiles and runs again after the major destabilisation.
Unfortunately it wasn't feasible to actually encapsulate all of the
pty backend's data, since the utmp helper and the need to fork and
drop privileges before doing anything else at all rather confuses
matters. So the data handle passed around to the pty backend is a
null pointer, and the pty backend is just as global-ridden as it
always has been. Shame, but such is life.

[originally from svn r2128]

21 years agoMajor destabilisation, phase 2. This time it's the backends' turn:
Simon Tatham [Fri, 25 Oct 2002 11:30:33 +0000 (11:30 +0000)]
Major destabilisation, phase 2. This time it's the backends' turn:
each backend now stores all its internal variables in a big struct,
and each backend function gets a pointer to this struct passed to
it. This still isn't the end of the work - lots of subsidiary things
still use globals, notably all the cipher and compressor modules and
the X11 forwarding authentication stuff. But ssh.c itself has now
been transformed, and that was the really painful bit, so from here
on it all ought to be a sequence of much smaller and simpler pieces
of work.

[originally from svn r2127]

21 years agoNeed to stub frontend_keypress() in console.c as well as window.c.
Simon Tatham [Thu, 24 Oct 2002 14:48:08 +0000 (14:48 +0000)]
Need to stub frontend_keypress() in console.c as well as window.c.

[originally from svn r2126]

21 years agoOops - repercussions of the close-on-exit stuff which I forgot to
Simon Tatham [Thu, 24 Oct 2002 14:12:55 +0000 (14:12 +0000)]
Oops - repercussions of the close-on-exit stuff which I forgot to
check in. I must stop doing my Unix checkins in the Unix subdir :-(

[originally from svn r2125]

21 years agoMinor compiler nits:
Jacob Nevins [Wed, 23 Oct 2002 14:24:40 +0000 (14:24 +0000)]
Minor compiler nits:
 - use smalloc/sfree, not malloc/free
 - include <ctype.h>
 - include <string.h> (although this doesn't shut the compiler up about
   non-ANSI stricmp/strnicmp)

[originally from svn r2121]

21 years agoImplement handling of all Close On Exit modes. Default is to close
Simon Tatham [Wed, 23 Oct 2002 14:21:12 +0000 (14:21 +0000)]
Implement handling of all Close On Exit modes. Default is to close
only on clean exit, which is a departure from most xterm-alikes but
Ian reckons people will love me for it. If this turns out to be
wrong, we can always change the default for Unix.

[originally from svn r2120]

21 years agoFixes for (Backend)->size() changes -- internal declarations didn't include
Jacob Nevins [Wed, 23 Oct 2002 14:03:11 +0000 (14:03 +0000)]
Fixes for (Backend)->size() changes -- internal declarations didn't include
new arguments and neither did internal calls.

[originally from svn r2119]

21 years agoRemove apparently unnecessary declaration of typedef Config from winstuff.h --
Jacob Nevins [Wed, 23 Oct 2002 13:46:59 +0000 (13:46 +0000)]
Remove apparently unnecessary declaration of typedef Config from winstuff.h --
it was breaking mingw builds (gcc being more picky than Visual C).

[originally from svn r2118]

21 years agoCleanups from yesterday's destabilisation: lots of stuff in
Simon Tatham [Wed, 23 Oct 2002 12:41:35 +0000 (12:41 +0000)]
Cleanups from yesterday's destabilisation: lots of stuff in
terminal.c was apparently relying on implicit initialisation to
zero, and also I've removed the backends' dependency on terminal.h
by having terminal sizes explicitly passed in to back->size().

[originally from svn r2117]

21 years agoOops. Forgot to check in terminal.h from yesterday's work. There's
Simon Tatham [Wed, 23 Oct 2002 09:11:36 +0000 (09:11 +0000)]
Oops. Forgot to check in terminal.h from yesterday's work. There's
always one :-/

[originally from svn r2116]

21 years agoMajor destabilisation, phase 1. In this phase I've moved (I think)
Simon Tatham [Tue, 22 Oct 2002 16:11:33 +0000 (16:11 +0000)]
Major destabilisation, phase 1. In this phase I've moved (I think)
all the global and function-static variables out of terminal.c into
a dynamically allocated data structure. Note that this does not yet
confer the ability to run more than one of them in the same process,
because other things (the line discipline, the back end) are still
global, and also in particular the address of the dynamically
allocated terminal-data structure is held in a global variable
`term'. But what I've got here represents a reasonable stopping
point at which to check things in. In _theory_ this should all still
work happily, on both Unix and Windows. In practice, who knows?

[originally from svn r2115]

21 years agoMake -ut work the right way round! :-)
Simon Tatham [Tue, 22 Oct 2002 13:27:58 +0000 (13:27 +0000)]
Make -ut work the right way round! :-)

[originally from svn r2114]

21 years agoStop `pterm -ut-' leaving the unnecessary utmp helper as a zombie
Simon Tatham [Tue, 22 Oct 2002 13:26:57 +0000 (13:26 +0000)]
Stop `pterm -ut-' leaving the unnecessary utmp helper as a zombie
process (Debian bug #165887).

[originally from svn r2113]

21 years agoThanks to Richard B for pointing out that xterm has its own variants
Simon Tatham [Tue, 22 Oct 2002 10:31:23 +0000 (10:31 +0000)]
Thanks to Richard B for pointing out that xterm has its own variants
of the alternate-screen and save-cursor control sequences, with
subtly different semantics and entertaining interactions with the
usual ones. No thanks to xterm for doing so in the first place :-(
This checkin should sort it all out.

[originally from svn r2112]

21 years agoJustin Bradford's patch for increased proxy robustness.
Simon Tatham [Tue, 22 Oct 2002 09:40:38 +0000 (09:40 +0000)]
Justin Bradford's patch for increased proxy robustness.

[originally from svn r2111]

21 years agoMake sure SIGINT and SIGQUIT haven't been nobbled in our child
Simon Tatham [Mon, 21 Oct 2002 23:01:34 +0000 (23:01 +0000)]
Make sure SIGINT and SIGQUIT haven't been nobbled in our child
process by weird POSIX-required shell behaviour.

[originally from svn r2110]

21 years agoDon't bother closing fds 0-2 before dup2ing over them; there's no
Simon Tatham [Mon, 21 Oct 2002 23:00:18 +0000 (23:00 +0000)]
Don't bother closing fds 0-2 before dup2ing over them; there's no
need, and it means we always have a valid open stderr.

[originally from svn r2109]

21 years agoVT100 line drawing characters should only happen between 0x5F and
Simon Tatham [Mon, 21 Oct 2002 22:59:14 +0000 (22:59 +0000)]
VT100 line drawing characters should only happen between 0x5F and
0x7E, not everywhere else. Silly me thought nobody would bother to
depend on this :-)

[originally from svn r2108]

21 years agoImplement Richard's really clever idea about bell overload mode:
Simon Tatham [Sun, 20 Oct 2002 13:23:30 +0000 (13:23 +0000)]
Implement Richard's really clever idea about bell overload mode:
it's automatically deactivated by any keypress, so that command-line
beeps from (e.g.) filename completion don't suddenly stop occurring,
but it still provides a rapid response to an accidental spewing of a
binary to your terminal.

[originally from svn r2107]

21 years agoScrollOnKey wasn't working because I failed to set seen_key_event in
Simon Tatham [Sun, 20 Oct 2002 12:44:29 +0000 (12:44 +0000)]
ScrollOnKey wasn't working because I failed to set seen_key_event in
pterm.c.

[originally from svn r2106]

21 years agoReject unrecognised command-line options; thanks rjk.
Simon Tatham [Fri, 18 Oct 2002 15:26:54 +0000 (15:26 +0000)]
Reject unrecognised command-line options; thanks rjk.

[originally from svn r2102]

21 years agoFix typo in man page. Thanks Richard.
Simon Tatham [Fri, 18 Oct 2002 10:38:22 +0000 (10:38 +0000)]
Fix typo in man page. Thanks Richard.

[originally from svn r2101]

21 years agoOops - that fix wasn't _quite_ right, since it killed all
Simon Tatham [Thu, 17 Oct 2002 16:58:24 +0000 (16:58 +0000)]
Oops - that fix wasn't _quite_ right, since it killed all
non-function keys completely :-/

[originally from svn r2096]

21 years agoMake the shadow bold offset configurable, after discovering that
Simon Tatham [Thu, 17 Oct 2002 16:51:01 +0000 (16:51 +0000)]
Make the shadow bold offset configurable, after discovering that
7x13 goes the other way to all other X fonts I've ever seen. (Arrgh.)

[originally from svn r2095]

21 years agoThis should fix the bug causing Alt-Shift to generate Escape.
Simon Tatham [Thu, 17 Oct 2002 16:45:17 +0000 (16:45 +0000)]
This should fix the bug causing Alt-Shift to generate Escape.

[originally from svn r2093]

21 years agoAdd the -xrm command-line option, to allow specification of an
Simon Tatham [Wed, 16 Oct 2002 22:54:58 +0000 (22:54 +0000)]
Add the -xrm command-line option, to allow specification of an
arbitrary X resource which doesn't have a dedicated command-line
option.

[originally from svn r2089]

21 years agoAdd a man page.
Simon Tatham [Wed, 16 Oct 2002 22:43:35 +0000 (22:43 +0000)]
Add a man page.

[originally from svn r2088]

21 years agoTemporarily change the default for cut-and-paste of line drawing
Simon Tatham [Wed, 16 Oct 2002 16:32:17 +0000 (16:32 +0000)]
Temporarily change the default for cut-and-paste of line drawing
characters, under Unix only, because the stub Unicode layer makes
the usual default break moderately painfully.

[originally from svn r2087]