If no description is given, just make the default/short description, literally, "No description given" - rather than refusing to add the file.

... instead of a comma-separated list. .ini lines are limited in length,
MD5sums are kind of long and would limit the number of .exe versions we
could support. Also, no we now have a way to correlate and report version
numbers/strings and URLs with each exe version. The "ver" and "url" keys
are not yet displayed or used for anything and are thus optional.

Mlong/foodfite

See merge request !53

This should guard against a RST being sent on success.  It's possible
that the remote has sent data (ie: M_EOB) we would recv() after we
commit to ending the session.  If the script terminates while there
is data to be read, this would result in sending a RST.

To prevent this, we call shutdown(sock, SHUT_WR) via setting is_writeable
to false (because that's how we roll), then recv() all data until the
remote closes the session, or the timeout passes.

I don't think the function xtrn_access() *ever* existed, but this prototype seems to have been around forever.

Like bbs.text(), except the "system" object is more widely available (e.g. in JSexec, mail server, web server, services) - in case any text.dat strings are useful in those execution environments tool. Requested by mlong.

Also cleaned up the argument validation in some of these other system methods (throw useful error exceptions rather than just returning false).

Update INSTALL.txt - cvs vs git

See merge request !45

more worldz default install for solomoriah's war

See merge request !47

Added the ability to optinally edit a message before forwarding it.

See merge request !52

It enables the keyboard insert mode by default (for Nelgin).

The "DOS" ARS keyword was always evaluating to false, on all platforms, for the C version of chk_ar(), which is used for populating JS *_area objects and for the User.compare_ars() implementation.

Unfortunately, the startup (sbbs.ini) "NO_DOS" option is not recognized here (yet), so it'll report true (e.g. for Win64 or Linux systems with DOSemu) even if/when the NO_DOS option is set.

e.g. don't log the error "term !ERROR resolve_ip example.com failed with error -2", just let the script deal with the failure and log an error if it wishes.

Pre-increment would've worked too.

prextrn.js keys off of the current node action to detect if the user is
currently in the logon process. Defeat that check in this case (there is no
logon process for auto-xtrn's).

add global head and tail for both section menu and external program menu

See merge request !49

use options.clear_screen for clear instead of hardcoded value in SelectItemHdr
remove redundant code for input in main section menu

... or when passed an invalid external program code or number, rather just just returning false. The underlying exec_xtrn() C++ function may return false (failure) for other reasons, so let's not obfuscate the other potential failure causes.

Previously, the xtrn*_tail menu was only displayed in the programmatic menu
mode.

Also, using the P_NOERROR mode flag for bbs.menu() to simplify all the if/else
logic.

And use require() instead of load().

Nelgin/xtrn tail

See merge request !48

Nigel Reed authored 4 years ago and Rob Swindell committed 4 years ago

display the file after the list of external programs.

cmdstr() is often called for non-executable command-lines (e.g. for @include: file name/path arguments), so don't default to native unless a mode value was specified. Hopefully addresses issue reported by bobrob without breaking anything else (again).

This reverts commit 94517fdf.

Always check the command-line for a '*' or '?' prefix and compare the executed filename against the native program list. For DOSEMU magic drive/path name command-line specifier substitution, but only when invoking DOS programs. Reported by bobrob via IRC.

xmodem_cancel() would return an error if putcom() failed, *before*
setting the "cancelled" [sic] flag. xmodem_cancel() in response to the
receipt of a remote cancel request (^X character).

Safe-string handling in C/C++.

Bug fix for net type when forwarding messages to internet email or fidonet netmail

See merge request !43

Eric Oulashin authored 4 years ago and Rob Swindell committed 4 years ago

When forwarding a message, it now correctly sets sets the to_net_type property in the
message header to FidoNet or internet for those types of message destinations

This can be used to defeat the automatic inclusion of the sbbs3
"setcap" build target when building on Linux. This should address
issue #175.