1) Initialize the r2y array for xBR so it actually works.
2) Add a vertical (only) interpolation scaler for aspect ratio enforcement
3) Add a simple muliplier scaler, so that can be removed from x_event.c
4) Use a new graphics buffer free list, which allows tracking last
   drawn screen instead of last bitmap rectangle, removing various hacks
5) Share the Y'CbCr <-> R'dG'dB'd tables between xBR and scale.c

This fixes all the window adjustment "stuff", but the new aspect
ratio enforcement means that ALT-UP and ALT-DOWN are useless for
resizing the window unless it's not already snapped... use ALT-LEFT
and ALT-RIGHT to decrease/increase the window size.

Also, the 8x16 font at 80x25 is *not* a square pixel aspect ratio.
As a result, it now stays "proper", which is likely not what you
want.

This was discovered when I was trying to make DigitalMan happy
because he wanted the apsect ratio retained on scaling.

Sorry for the impending imminent scaling hassle DigitalMan!

Or detail level is MAX.

Sorry for the impending imminent merge hassle Deuce!

Do this in JS and use JSON for format instead of .ini.

The old "content" array property goes away.

Uses "pointyscale" for x3 and x5
Uses xBR from FFmpeg for x2 and x4
HQx is also included, but unused as it's too slow at x4.

<plt> May 13 00:39:40 sbbs synchronet: term Node 1 <plt> append key into keybuf: 4E (N)
<plt> May 13 00:39:40 sbbs synchronet: term Node 1 <plt> Executing external: /sbbs/exec/sexyz 14 -Telnet sz /mnt/disk1/dl/COCO/AGI/agi-xmas.zip
<plt> May 13 00:39:40 sbbs synchronet: term Node 1 <plt> sexyz: Synchronet External X/Y/ZMODEM  v3.0  master/23b741a1  Copyright Rob Swindell
<plt> May 13 00:39:40 sbbs synchronet: term Node 1 <plt> sexyz: !File skipped by receiver
<plt> May 13 00:39:40 sbbs synchronet: term Node 1 <plt> sexyz: !File Skipped
<plt> May 13 00:39:40 sbbs synchronet: term Node 1 <plt> sexyz: !0 Aborting Transfer (Sending ZABORT)
<plt> May 13 00:39:40 sbbs synchronet: term Node 1 !ERROR 11 receiving from socket 14
<plt> May 13 00:39:40 sbbs synchronet: term Node 1 input thread terminated (received 365 bytes in 35 blocks)

Not sure why this isn't happening for other sysops - perhaps plt was spying on the node?

This was really slowing down the upgrade_to_v319 and there's no current consumer of the data.

Consider adding back in JSON format later or just leave it to JS things to use for JSON-formatted metadata.

Don't allocate and memset a buffer of -1:
#1  0x0000559222bc167f in strListCombine (list=0x559224220720, buf=0x559224226550 "", maxlen=18446744073709551615,
    delimit=0x559222bcce0a "\r\n") at str_list.c:447
447             memset(buf, 0, maxlen);
(gdb) print maxlen
$1 = 18446744073709551615
(gdb) up
#2  0x0000559222ba3841 in smb_addfile_withlist (smb=0x7ffce63a7e90, file=0x7ffce63a5bc0, storage=1, extdesc=0x0,
    list=0x559224220720, path=0x7ffce63a6e80 "/xfer/loons/manhater.mp3") at smbfile.c:365
365                     strListCombine(list, content, size - 1, "\r\n");
(gdb) print list
$2 = (str_list_t) 0x559224220720
(gdb) print list[0]
$3 = 0x0

Found/reported by plt via irc. Thanks!

Don't use iniSet* since we know we're not updating existing ini entries. Use strListAppendFormat() instead.

Do multiple before divide, and use an lround() to get the same results
on Win32 with mingw32 as Clang on FreeBSD.

Was removed along with select() in e2094fb9

The fread() usage here did not need "addressing" (Commit 1b56dc96).

This caused messages posted via smbutil to be short (e.g. 0 or 1 bytes
in length). Reported by Accession via Idle Relay Chat.

Also, don't call strlen() on msgtxt which could be NULL, causing segfault
in final printf() callin in postmsg().

Being stritly less/greater causes a one pixel gap at the start/end
of the ellipse, meaning flood fills can escape.

Clamp coordinate translation max
Fill using native pattern, not scaled pattern
Add a bad, broken, and slow even/odd fill
Fix logging when RIP is enabled
Use the same ellipse algorithm for partial as full ellipses
Detect diagonal movement in ellipse drawing and don't make the edge fat
Only recurse once in the flood fill
Implement the v2 poly bezier command
Enable filled pie slices
Add RIP_POINT command (basically a scaled pixel using line width)

If the extended description and the normal (short) description are the same,
delete the extended (long) description.

This should address issue # 261.
Before April 19, 2016, area-linked nodes were required to be listed as nodes in sbbsecho.ini:
http://cvs.synchro.net/cgi-bin/viewcvs.cgi/src/sbbs3/sbbsecho.c?revision=3.6&view=markup
I'm not sure why the change was made to only require that there be a wildmatch for each area-linked node. So just make it a configuration option. <shrug>

Also removed the trailing "..." for every top level menu item in echocfg.

It's very tricky to defeat the auto-pause to have your own custom prompt at  rows-1. So going back to having  a blank line before the batch prompt.

If a sysop wants a *blank* line before the batch flag prompt, they can put that (or a conditional blank line, even better) in their text.dat (BatchFlagPrompt). For now, I'm thinking no blank line before the prompt is better use of the screen.

Consistently use map_rip_color()
Optimize flood fill to not clobber the stack
Fix small full ellipses to match RIPTerm 1.54

One in an error path (CID 331353), the other not.

as get_ini_globals() will de-ref the 'interfaces' element. CID 331352

Mostly tightening things up and fixed some text things.

Some archives contain exactly the same files as others, but in a different order. Believe it or not.

This also exposed problems with the find_section() utility function: when the "found" section was actually empty and the next immediate line in the ini file was the beginning of a new section, the "keys" indexed would actually be the next section. Now fixed.

More support for moving backwards through file searches/listing types.

e.g. when using the JS FileBase.update() method

When moving or renewing files, they're first removed and then added and they were being added with the DELETE attribute set. The filenames are displayed in red when staged for deletion/removal, so that's how I noticed. Oops.

Apparently these menu files were created exactly 80 columns wide. <shrug>

This fixes issue 257 reported by Keyop.