There's still outstanding issues in Coverity around sftp_state and
sftp_channel that need some deep thought, but this one is trivial.

Prevents buffer overlow with manually edited or corrupted files.

This way, WMs that enforce aspect ratio etc, won't get potentially
block the update.

A hex nybble is not 8 bits, and C strings need to be termianted.

This is a very crazy lock/unlock pattern... an extra assignment
makes it easier to follow the logic.

This should also resolve a conspiracy Coverity has about two input
threads running at the same time.

This is likely what make the old broken values appear to work.

does when compiled with mingw32.

Use "anything else" to convert hex to char.

Fixes crash when built without X11 or SDL, but with GDI (not default)

Changing a command shell or an external editor's internal code will revert
the preference of any user's that have selected that shell or editor as their
default/preference. A sysop may not realize this, so warn them with lots of
detail in the help text if they need it. See issue #721 for details.

Also updated the help text for command shell internal codes since it still
only mentioned Baja/.bin modules.

CSI=255h was not entering doorway mode anymore.

Basically everywhere was off by one.

Also, fix the get dpi size thingie to correct for window decorations.
We never actually use the result of this, but Windows might for
outline dragging or something... you can never know.

The extraction destination directory must be specified with the 'w=' option.

As Fernando Toledo (@Ragnarok) pointed out:
#524 (comment 4867)

So do more terrible hackery.

This means ciolib utilities built with MSVC won't have the right
manifest information, but the documentation is terrible and I can't
figure it out.

Looks like this was where it messed up first.

It appears they just set the default values, so try that out.

Presumably, "something" is setting it earlier so it can't be set
again because reasons.  To work around this, add a manifest to
ciolib library that asserts the binary is DPI aware and requests
Windows not do GDI scaling.

Not sure if manifest .o files in a lib are linked in by things, but
SyncTERM pulls the .o out and explicitly links it, so this should
fix SyncTERM at least.

We will now try SetProcessDpiAwarenessContext() first, then
SetProcessDpiAwareness() second, and finally SetProcessDPIAware() [sic].

Also, "handle" the WM_DPICHANGED event.

This may solve problems initializing SyncTERM on systems with two
displays that have different DPI settings.

SlyEdit: Update for the isPrintableChar() function to allow more for UTF-8 (it's a simplistic function).  Also updated the version number in slyedcfg.js.

See merge request !411

SlyEdit: Update for the isPrintableChar() function to allow more for UTF-8 (it's a simplistic function).  Also updated the version number in slyedcfg.js.

This allows easy copy/paste without scrollback hacks on BBSs that
insist on taking over the mouse just so you can move the mouse and
click a character rather than press one key because it's 3l337..

When using $^, all the auto-dependencies are included in the link
command-line, which includes sbj.c and all the .h files, causing
the link to fail if dependencies exist.

This won't matter now since only one thread ever touches it, but
when cool stuff happens, it could.

This allows ALT-B from dialing directory to show the last seen
thing, not the extra blank screen added to the end.

Doesn't fix anything, but a huge QoL improvement for people who use
ALT-B after disconnecting.