the default/all target.
So a single "make" command in src/sbbs3 should build these nested targets
as well.
Caveat: "make clean" does not clean these sub-make targets.

uedit.c:1475:5: warning: this 'if' clause does not guard...
     if(uifc.changes)

but it's parsed as a number.  Use the JS Date type for the values.

we are always freeing the pointer that was asigned the text.dat list and things
will crash (thanks for the report, Rastus).

Also add add_public_key() and get_public_key() methods to CryptKeyset.
Not tested yet, but this should be everything we need to finish up
LetSyncrypt.

This class allows saving and loading private keys.  It doesn't currently
allow saving or loading *public* keys, since they require certificate support
which we don't yet have in our JavaScript Object Model.

I'll fix the JSDocs build in a minute.

L (capital 'L'): Lock a message base
U (capital 'U'): Unlock a message base

These commands maybe used to insure that a message base is not modified while
being backed-up (but don't forget to unlock them when you're done!).

Also note: if you back up all the files (including *.lock), then the backed-up
message bases will also be locked and will need to be unlocked (i.e. *.lock
removed) before they may be used again if/when restored.

allow the "last_msg" value to be changed as well.
When dumping a msgbase index, show 'M' for messages (not blank).
When running maintenance, read the entire index in one shot.
Don't display progress (percentage counting) during maintenance (just
slows it down).
When reading messages, get the plain-text version (MIME decode if needed).
When a message base fails to open, continue processing additional message bases
if specified on the command-line.

Don't force set the label, require the caller to set it.
Fix up ASN.1 parsing of RSA certificates.

(Coverity clean-up) - apparently this function was already erroneously freeing
the object private pointer (p) and I copy/pasted that error into more places.

the format required by ACMEv2/Let's Encrypt.

Yes, this seems to be the easiest way to exfiltrate a private key from
a cryptlib context.

Reduce certificate checking... the default level will not validate the
certificate used by acme-staging-v02.api.letsencrypt.org.  Presumably, that
means other Google API stuff won't work either.

(without requireing the use of -renumber).

for very old message bases).

return an error.

-U (capital 'U') command-line option (e.g. '-U0' or '-U2' or '-U077').

numbers to message with "duplicate message numbers" when they are at the
end of the message base (newly created message headers).

the highest message number indexed.