The "EchoLists" help menu only worked when you first enter the sub-menu and would get wiped out by subsequent child-menu help text.

Filled out the "Configuring an EchoList" help text.

Fixed issue seen where deleting an EchoList's Required Key didn't set the "dirty config" flag, so saving of changes was not prompted when exiting (if that was the only change made).

Also, it's not technically an option, so remove from the option list.

Fixes the Win32 build.

Done.

At Tracker1's request/suggestion.

jsexec -Itrue
...
Result (boolean): true

More elaborate expressions can be evaluated and white-space included by quoting the "-I expression".

Also, send all cmd-line usage errors to stderr and syntax help to stdout, always (Deuce's request).

Stop lying. Programming is hard enough already.

The extra spaces caused the centering of the cryptlib version to be offset (not exactly centered).

Now the default bar is INT_MAX, so it always gets corrected when
left as default.

This is good.

Instead of setting bar to zero when possible, set it to the number
of displayed options, putting the lightbar at the bottom of the
displayed list instead of at the top.

So here's what we're doing now... we set bar so that the first item
shown in the menu is the lowest index possible that still shows cur.
bar is therefore calculated as the displayed item count minus the
distance from cur to the end of the list.

Yeah, that's the easiest way to explain it.

scrn, showbuf, timedisplay, bottomline, getstrxy

iniGetNamedStringList() was returning NULL when attempting to get the list of keys in the "root" (unnamed) section and the file is actually empty/blank. If there were *any* characters in the file, it would return an empty list (not NULL). Now, always return the empty list, even when the file is blank.

Doing so prevents extending the prototype.

The default character set for outbound mail is now auto-determined (when not explicitly specified for a message) between UTF-8, ASCII, and CP437. The [mail] DefaultCharset setting (which fell-back to iso-8859-1 if blank) is no longer "a thing".

Also: specify 8-bit content-transfer-encoding for the (potentially UTF-8 or CP-437) plain text portion of a MIME-encoded message with file attachment (7-bit was wrong) and pass down the text sub-type (e.g. could be "html") for inclusion in the mime-part header (don't assume text/plain, but still use that as default).

Also: log an error when failing to delete an attached file (e.g. from data/file/*.out).

Some (important) File methods did not support .ini files that used the !include directive because they were using the xpdev iniRead* API (which performs no "pre-processing") instead of xpdev iniGet*.

Impacted methods:
- iniGetValue()
- iniGetKeys()
- iniGetObject()

The other existing ini* methods already worked fine with nested (!include'd) .ini files. It's possible there's a slight performance penalty with the new implementation since the entire .ini file is always read for each operation and previously it was possible that only a few "lines" were read to find the key(s) of interest. However, since .ini files are not typically huge and the iniRead/file-stream method likely read large (e.g. 8-32K) blocks anyway (which is usually the entire .ini file) - I don't actually suspect any observable impact to performance.

This change was needed for the new ctrl/modopts.d support.

Added new method useful for debugging nested .ini files:
- iniReadAll()

This allows auto-inclusion of sub-directories of config files (e.g. !include modopts.d/*.ini). This, along with the next commit to js_file.c, allows module authors (or sysops) to keep their module-specific settings in a module-specific file (rather than always merging-with/managing the monolithic/shared modopts.ini). The [modname] section heading is still required in the included .ini file.

Replaced use of non-thread-safe STRERROR() with safe_strerror().

Example:
Command-line: ?showargs   " a b c "d "e f"
argc = 3
argv[0] = ' a b c '
argv[1] = 'd'
argv[2] = 'e f'   

This resolves a long-standing TODO comment.

Also, fixed a problem where multiple spaces between the module name and the first argument would result in argv[0] being set to an empty string.

The first argument can now optionally be an array of strings (e.g. as read from File.readAll()), so that multiple searches of the same file (e.g. twitlist.cfg, while importing messages) does not require multiple *reads* of the same file.

If there was an error opening a file for attachment, the MIME part would be left unterminated.

MSVC detected heap corruption from this function when the attachment was 0-bytes in length. Good catch.

Consume the SMTP lines as fast as possible since the sender may be on the
local network (or even local/loopback interface!) and could get way ahead and
timeout otherwise, resulting in the "premature evacuation" error (and dumping
of the received message) on the receive side. This resolves an observed issue
with sending large attachments to the mail server at very high rates and the
sending client timing and disconnecting waiting for a response from the server
(which was throttling the receive using YIELD).

An unrelated optimization (elimination of an unnecessary use of fprintf) resulted in a new bug that combined all lines from SMTP-received mail messages into a single long line, thus breaking all decoding ability of multi-part MIME messages (where blank lines are significant).

Went ahead and replaced some other unnecessary uses of fprintf(), replaces with fputs() while at it.

Some terminals can send NUL (ASCII 0), e.g. hitting Ctrl-Space in Apple iTerm. Allow users of inkey() (in C++ or JS) to optionally detect the difference by specifying the K_NUL mode flag. In JS, console.inkey() will return null upon timeout (rather than "") when the K_NUL mode flag is used and return an empty string ("") upon receipt of NUL (ASCII 0).

The default is the previous behavior where a timeout and the receipt of a NUL character appear the same to the caller of inkey().

When a file library's "Access to Sub-directories" option was enabled, any auto-created directories would have their data dir initialized as "blank" (empty), so their data files (e.g. <code>.dat, <code>.ixb) would be created in the ctrl directory.

I guess nobody used this feature? It is pretty cool, but not really documented. If you were using it before this commit, you'll need to move your sub-directories data files to data/dirs or re-add the files (recreate the database files). And you can remove the old cruft from the ctrl dir.

This involved the removal of the content-transfer-decoding feature of the SMTP mail server since we need single-part MIME-attached file contents to be stored in their original encoded form (e.g. base64-encoded) and not in binary, for the message body text (where NULs aren't allowed, CR/LF is appended, etc).

I accidentally made this change to smbtxt.c in the new_file_base branch and then copied over here, so there's some unrelated innocuous changes (comment header, removal of SMBCALL) that hopefully won't cause a merge conflict later.

I don't actually receive single-part MIME attached files, so depending on others (e.g. Nelgin and Dream Master) to test for me.

Don't display at all if smb_getplaintext() returns a blank string.
Don't display "(null)" if there's no text sub-type.