Not actually part of CTerm, but documented here anyway.

This destroys graphics.  Fixes SF bug 94.

sdl_setwinsize() and eliminate another race.

Pass window size in SDL_USEREVENT_SETVIDMODE
- Fixes race condition where SDL_WINDOWEVENT_RESIZED or
  SDL_WINDOWEVENT_SIZE_CHANGED arraves while
  SDL_USEREVENT_SETVIDMODE is in the queue

Set new scaling based on old scaling multiplier
- Will resize window on mode changes, but tries to keep integer
  multiplier similar.  May break fullscreen modes, dunno.

Have aspect_fix() return higher rather than lower
- This may make it the same as aspect_correct(), but I'm too lazy
  to figure that out.

Add additional stderr debugging.
Don't try to correct the aspect ratio of the window.

Previously, b64_decode was being asked to decode past the end of
the application string.

This should get the original RIP stuff working properly.
LORD icon files are ideally handled a bit differently:
1) If the remote supports the SyncTERM cache operations, icons
   are transferred directly to the cache using them.  This requires
   a SyncTERM build including commit fa51ebe7.
2) If LORD is running in-process under Synchronet, it will try to
   initiate a YModem-G download for the files using
   bbs.send_file(name, 'G').  This means YModem-G must be enabled,
   and must use the 'G' command key.

Having the *correct* icon files will still work, but they must be
the expected file size and be the same modified date or newer to
be used by LORD JS.

Ideally, there would be a way to configure sexyz or something to
do icon transfers, but that's a lot harder, so I'm not going to
bother.

Fix off-by-one error with b64 cache store buffer size.

This allows ensuring we don't expand CRLF into CRLFLF.  This is
usually harmless, but there's no real need to put up with it in
a program with a single thread using the expand function.

This is what was breaking uploads via telnet.  Not only was it sending
the unexpanded buffer, it was also sending the length of the expanded
buffer out of it (wandering into gibberish or at worst, outside of the
buffer).

Fixes icon displays in RIPs that don't use the .ICN extension.

It seems Linux has deprecated siginterrupt(), and they've been
aggressive about removing deprecated C functions lately.

already in binary mode.

Should fix Sourceforge issue 91

Fixes SyncTERM SF bug 93

The changes for Ghost support broke URL parsing.

This doesn't actually contain things you can modify, but rather
answers the most FAQ "Where are the files stored?"

Before this change, dynamic parent windows wouldn't be redrawn as
active after showbuf() returns.

Just in case there's some clock wrap around issue  causing issue #468. Seems unlikely with a 64-bit clock, but there are a lot of conditions in the xp_timer64() implementation that could evaluate differently on different versions of Windows or on different computers.

This is the issue behind CVE-2022-24130 in xterm.
Reported against SyncTERM by Autumn Lamonte (Thanks!)
Discussion here: https://gitlab.com/klamonte/jexer/-/issues/105

Closes SyncTERM bug 79.

Previously, a file transfer would unconditionally enable RIP.

Quote line wrapping for DDMsgReader and SlyEdit

See merge request !235

This conversion was done automatically via git clone because of
the latest .gitattributes settings for these file types.

Fixes issue #463