| |
::::::::::
:: HTTP ::
::::::::::
In contrast to most other solutions (e.g. CGI scripts) BCM uses a build in
HTTP server. That's why BCM HTTP interface can handle any kind of command
interactive. That means that commands are executed directly and users can
login with their own call. The HTTP server can be reached normally via
port 8080. Sysops can setup other ports if they want (see HELP HTTP_PORT).
At first connection with BCM HTTP interface the user gets a window
where login call and password must be entered.
Follow these rules:
- login call is every time your own call.
- password must be entered as following:
* if you connect via localhost [127.0.0.1] or via packet radio network then
password is same as your name known by BBS. So sysop must not setup a
password for each user. If a user has entered a user password, login via
HTTP is impossible.
* if you connect from other networks then password is your set TTY
password (see HELP ALTER TTYPW)
After successful login the BCM HTTP server sets a little cookie at users
webbrowser. The cookie is used to identify the user for later logins and so to
bypass login query. If webbrowser is forbidden to save a cookie, user have to
login with his user call and password each time he uses BCM HTTP server again.
The webbrowser must support frames; netscape products are a good choice.
All functions of BCM can be done via HTTP interface, but there
are some restrictions:
- if sending mails you have to add reference everytime. It's not working to
write a send command in command line and then to click on execute
- you can't use interactive runutils
- if you read mails with binary content, the content is converted
automatically (e.g. JPG pictures)
All other functions should work as normal. See also HELP HTTPNEWBIE
| |