command to send.

- typo "tickit" not "ticket"
- convert ini content array to object to create map

the find-text string.

data is external, in the exec/init-tickit.ini file).

length of option[].

terminated with a 0-length option string).
There still another crash I'm investigating that occurs if you type a string
that isn't found. It appears to deref past the end of the option array.

"ONLY", add "SYSOP" to the access requirements automatically.
Exception: when importing SUBS.TXT (since it includes Access Requirements
already).

For example, this is a valid single response:
220-I'm so clever
221 Yes, yes I am
220 So very clever

 Error 24 opening '../fido/outbound/5e7a2520.pkt'.  Not sending.
instead of this:
 Unable to open '../fido/outbound/5e7ab832.pkt'.  Not sending.

Error 24 is "too many open files", btw. Another problem to solve.

1) Fix ACCT command
2) RMD response is 250, not 221
3) When sending a file, throw an exception if a 226 response is not received
4) Before sending a command, if there is data available on the socket,
   log an error (don't throw an exception).

a new message with smb_addmsg(), retry (using the configured delay/timeout
values).
On Windows, importing a QWK packet into SMB messagebases stored on a remote
Samba server, for some unknown reason, the read of the msgbase status header
here would fail about 1 out of 100 times, with the error:
 evnt ERROR 13 (Permission denied) (WinError 33) in qwktomsg.cpp line 515
 (sbbs_t::qwk_import_msg) writing "msgbase" access=-203 info=smb_getstatus
 reading status
... as if another task had the msgbase status hdr locked (which, they don't),
and we (this thread) already had it locked (just 8 lines up in this same
function), so shouldn't even be possible. A single retry appeared to be
sufficient, but I went ahead and put the timed-loop with the delay in here.

Requested by: DigitalMan.

data is external, in the exec/init-tickit.ini file).

The alternative is JSON, really, but I like .ini for this type of list.

file containing only the good or bad messages, leaving the orignal *.pkt file
intact.
The -recover option will *only* create a *.good file (renamed from .recovered).
Now, if the recovered (good) or bad output file has no messages or is the exact
same size as the original pkt file, the output file is automatically deleted.

valid pkdmsg type.
This will allow better re-sync with a badly corrupted packet.
I waffled about whether it should try to read the variable header fields if the
DateTime field of the fixed header is invalid. Currently, it does. The pkdmsg
header does not have very good identifying marks to re-sync with, so it's
still possible that a valid msg that follows a corrupted one could get skipped.
I would need some good set of varying examples of corrupted packets to test
with, but I'm sure it's possible to "do better".

is called without a "mode" specified. This uses the new default SCAN_INDEX
mode. If you need the old behavior (does not invoke the listmsgs_mod), specify
a different mode value (e.g. SCAN_READ). All of the mode values (other than
SCAN_INDEX) will not start the list message module, but rather use the internal
C++ logic.
This provides an intuitive way to actually invoke the list messages module
from JS without hard-coding bbs.exec("?module") into a cmd shell, for example.

duplicates of all the parsed packet files, but excluding all corrupted packed
messages. The recovered packet files are named *.recovered.
Also improved the corrupted packed message header detection to make this
feature more useful.
Created at the request of Nelgin who promises to test and use, for better or
for worse, in sickness and in health, from this day forward, yadda yadda yadda.

If the RECIPIENTTOLIST header fields exists and the RFC822 header field
does not exist, that means the original RFC822 "To" field was not MIME-encoded,
which is the majority of email. The routine that converted an smbmsg_t to an
RFC822 header did not use this new field however (ignored it). Whoops. So
multiple recipients were reflect in the To field and the original destination
address(es) in the To field weren't passed on either. Thanks, Frank!

mouse button is pressed.

This makes the X10 mode much better, allows copy (but not paste), and
prevents weirdness with button releases when dragging.

While this may or may not be what xterm does, it's clearly the best choice.

input buffer when we stop the mouse.

command from passing the mouse release event in transit (and getting garbage
input to a next command).

We may need a one-shot mouse thing or something like that...