(the upgrade renames binkit.ini and ftn_domains.ini to *.old).

Use "jsexec binkit upgrade" to upgrade your existing .ini files if you're
already running BinkIT 1.x.
echocfg v3.04 may now be used to set all the BinkIT settings and domains
(in sbbsecho.ini).

(note: some of the key names were changed), run "jsexec binkit upgrade"
to upgrade your existing .ini's (if you have them).
Removed support for sbbsecho.cfg.

(e.g. "Zones=1,2,3,4,5,6").

if the thread returns before js_initcx() is called.

Reported by bgdjr.

discard).
Also, trim trailing white-spaces.

sdl_con.c: In function ‘sdl_mouse_thread’:
sdl_con.c:1424:1: warning: no return statement in function returning non-void

around a broken CRAM-MD5 implementation.  For now, retrains the double-test
for Mystic/1.12A39 until we see what version number the fixed builds have.

If the major version is less than one, returns true.

If the major version is greater than one, returns false.

If the minor version is less than 12, returns true.

If the minor version is greater than 12, returns false.

If there is no character after the minor versions, returns false.
(ie: 1.12 is newer than 1.12A39 and is assumed to be fixed)

If the character after the minor version is not an 'A', returns false.
(ie: 1.12B1 and 1.12.1 are newer than 1.12A39 and assumed fixed)

If the number after the 'A' is less than or equal to 39, returns true.

In all other cases, returns false.

This prevents lowering security by having two allowed CRAM-MD5 responses
for some remote software versions.  Hopefully the fixed build will have
version 1.12A40 and we'll be able to remove the last version that does that.

channel we ignore/close, we would get a separate warning for each chunk of
data on that channel until we read past the close.  Instead, if we're closing
a channel, we note the channel ID, and simply discard the information
immediately on the most recently closed channel.

A remote could still cause a bunch of log output by multiplexing two or more
writes to invalid channels, but nobody seems to do that, so this should be
good for now.

syncprog and the commit log again).

of the #if !NO_SOCKET_SUPPORT block. Fixes Windows build of addfiles
and other utils.

line 1040: TypeError: ini.GetObject is not a function

Version 0.36 Beta: For non-lightbar voting, updated to use BallotHdr (791) instead of SelectHdr (501).

the output_thread which may occur after an SSH-related error in the input
thread, e.g.:
Sun Mar 25 2018 01:45 pm bbs.synchro.net
Node 6 SSH ERROR 'Bad argument, parameter 2' (-2) setting channel

Sun Mar 25 2018 01:45 pm bbs.synchro.net
Node 6 SSH ERROR 'Bad argument, parameter 2' (-2) setting channel

Sun Mar 25 2018 01:45 pm bbs.synchro.net
Node 6 SSH ERROR 'Bad argument, parameter 2' (-2) setting channel

Sun Mar 25 2018 01:45 pm bbs.synchro.net
Node 6 SSH ERROR 'Bad argument, parameter 2' (-2) setting channel

Sun Mar 25 2018 01:45 pm bbs.synchro.net
Node 6 SSH ERROR 'Bad argument, parameter 2' (-2) setting channel

Sun Mar 25 2018 01:45 pm bbs.synchro.net
Node 6 SSH ERROR 'Bad argument, parameter 2' (-2) setting channel

... by introducing a new sbbs_t member variable: terminate_output_thread
which is only set to true when the input_thread is terminating.

If there is any possibility that the input_thread could startup and get
an SSH error before the output_thread even starts, then we will want to
move the terminate_output_thread=false initialization to the sbbs_t
constructor.

a temp variable.  Mostly impacting the setting of video flags.

Fix-up the user/####.subs ini style (only one blank line between sections).
Also, don't re-write the .subs file if nothing has changed.

(data/user/*.subs - in familar .ini syntax)
indexed by the sub-board internal rather than the old flat binary file using
the troublesome "pointer index" (ptridx) as an offset. The old files
(data/user/ptrs/*.ixb) will still be read in when the new file doesn't exist,
for upgrading users automatically to the new method of pointer storage.
This should resolve long-standing issues around users' new-scan pointers and
configurations getting corrupted when the sysop inserts/remove sub-boards.
However, if you change a sub-board's internal code, that invalidates the users'
pointers and scan configurations - so that may need to be addrsesed via SCFG.
The subs' ptridx values are still be used for storing QWKnet hub pointers, for
now, but that too will soon be addressed with a new file format for those ptrs
too.

stuff in binkp.js, only the protocol.  BinkIT now does the Mystic check.

increase post-count width supported from 4 to 5 digits (beautifully).

networking technology for 20 years. :-(

or not).

severity) whenever there is an SSH session establishment failure, because
some of the SSH failure errors are DEBUG level severity and won't be visible
with "Normal" (informational) log output filtering. This resolves the issue
of seing "SSH connection accepted from <ip-address> port <port>" messages
with no follow-up messages as to what happened to that connection.

messages. There's a lot of places in this script where we just send a
(somewhat explanatory) M_ERR string, but don't log anything locally unless
debug-level logging is enabled, now we'll at least log those messages with
a bit higher severity.