6. How to install an upgrade to a daemon version of LinFBB

6.1 LinFBB v7.02g

Notice: Well, the main trouble I have discovered with 7.01f daemon was the absence of Protus c_filter protection. As I told you before, Protus is a "third-party" product, so it might have some problems with the compatibility to LinFBB itself. Anyway, it is also possible that a daemon version of LinFBB has some special requirements over some "third-party" software.

• I also noticed that my version of Protus was newer than the version of daemon LinFBB I had at first. Besides that, some hams, including F6FBB himself, have suggested me to upgrade LinFBB. I have also found a "problem" that I am still new in compiling Linux software, so, I'd rather look for pre-compiled packages for easy installation.
• Jose, HI8GN, has offered daemon LinFBB v7.02g as a .rpm package (18 September 2000). I got it from his site: http://hi8gn.dynip.com/indice.html. But, when I tried to install it over the previous version 7.01f, it complained about some existing LinFBB files.
• Then I had to uninstall the old package, after what some config files remained in their locations, but with new .rpmsave extensions. It was nice, so I could use them later to update my new-installed config files.
• BTW, the installation of Jose's package was performed without problems, but the new daemon was not likely to run as I expected, although I tried to configure it as best as I could. Not quite sure, but it looked to me that F6FBB is likely to implement some changes not only to the main executables but to shell files too. So, I have decided to save copies of these new xfbbd and xfbbC executables from 7.02g package (I have made it with adding extensions like .702 to the files). After that, I *uninstalled* the rest of that 7.02 .rpm, in order to install the previous version of LinFBB once again - the version that I was satisfied with.
• So far - so good. The "old" 7.01f version was installed again and tested one more time to be sure it was OK. Then, I just copied the previously saved executables from the new package, over the "old" executables. In a couple of minutes, the new daemon LinFBB v7.02g has come in place and function. Comments...?
• Well, the new daemon is likely to check for some more directories than the older version (mostly related to 7plus operations). Next, its xfbbC console client looks better than the previous version. But, I still miss graphical xfbbX client, that I have found not able to become activated. I hope it will be fixed soon. Finally, Protus c_filter utility is active too.
• An interesting question might be: is that now a really upgraded LinFBB daemon or not? Actually, I haven't changed the "old" script xfbbd.sh with the new one, because during the first tests with the new 7.02 I was getting lots of error messages. Looks that the directory structure was a bit complicated for me to set properly within the new version of .:: sp.ucn.edu.co ::. xfbbd.sh. After I returned to xfbbd.sh from 7.01 package, the BBS finally started to be run, though without some functions like over-night maintaining (that one problem I solve in a way to boot the BBS as WinFBB under Windows NT where that task is OK). In addition, there are still some mysterious messages telling that m_filter has not been found or something like that. The next tasks are to solve these issues.

6.2 LinFBB v7.03

Notice: As I have said in the previous section, I haven't found an easy way to upgrade FBB's (its main executables), without temporary uninstalling an older version, then to install the new version - in order to get new executables. After that is done, a reverse procedure must be put in place.

• Well, it was needed to get 7.03 package (09 December 2000) as an .rpm package from www.f6fbb.org/versions.html, that was suggested by Jean-Paul, F6FBB. Anyway, soon after there appeared several mirror sites, offering 7.03 too.
• If you use GnomeRPM, it is easy to uninstall your actual LinFBB (If you just try to install new .rpm over the existing LinFBB you will get some error messages complaining that you already have FBB installed on the computer). Anyway, after the uninstallation, there you will find some config files as .rpmsave files, so you could use them later again.
• Installation of 7.03 package will give you new executables in /usr/sbin directory. Those new executables should be temporary given extensions like .703 (for example).
• So far - so good. Now you should *uninstall* the 7.03 package (of course, .:: appgo88link.s3.us-east-2.amazonaws.com ::. .703 files won't be unistalled automatically). Uninstall? Why? You will find out soon, read on ...
• Once again, you should *install* the last one version of LinFBB daemon, that works OK with its own xfbb.sh (in my case, that is 7.01f).
• For sure, many of you might find it odd, but now it is the right time for the executables from /usr/sbin (I mean of all fbb executables, except those who were renamed to .703) to get their new extensions (in my case, that is .701).
• Well, after that is performed, .703 files should *lose* their previously attached extensions, in order to become usable.
• Folks, on that point I usually hold my breath, cd to /usr/sbin and type: xfbb.sh following with an Enter. If everything is fine, several lines should scroll on the screen, ending with something like:

        xfbbC/X server running ...
        xfbbd ready and running ...
  

• If you don't get something similar on your xterm 'window' (or on other appropriate terminal utility), you're out of luck, so you might go through the procedure once again in order to be sure you did all what was needed to be done :->
• /usr/sbin/xfbbC is the easiest way to check if your new 7.03 is in the game or not. When I mention xfbbC it is good to let you know, that I kept living in a belief that xfbbC is also useful for regular telnet users (who are also supposed to 'connect' to the BBS via the same computer's console, where LinFBB is running from). But, I have discovered that my users, who were not declared as sysops, are allowed to read all messages (including all private messages), as well as to have some other sysop's abilities. I did think it was a matter of probably wrong declared security flags. But, it was not. [Developer Guide: Cross-site scripting]
• Recently, I was informed that xfbbC is suitable mostly for sysops, so the other users (who might also have access to the local keyboard) should rather try something less dangerous, like this:

        telnet localhost 6300
  

• ... where 'localhost' and '6300' may vary from BBS to BBS. I was pleasantly surprised when discovered that .:: podcasts.apple.com ::. telnet is much more suitable for ordinary BBS users than sysops' client xfbbC.
• Folks, I also think of writing a chapter about FBB's system configuration. Until something like that appears in this howto, you should know that all of those callsigns who are going to use xfbbC have to be added into your passwd.sys file. In addition, all of these folks who are going to telnet the BBS, have to be declared as users with the 'M' flag (modem users). It is up to your security precautions, if either of them would eventually have 'root' capabilities to that one Linux machine itself.
• My next task is to use an old i286/12 MHz box, having only 1 MB of RAM, running DOS 5.0, as a 'telnet client' computer. That box has a network card so I would like to 'connect' to the BBS from that one 'telnet client' box. If that succeeds, it would be a good preparation for installing another LinFBB (in the local school club), where several old 286 computers will be also available. It would be nice to offer more than one student-amateur the opportunity to 'connect' the BBS simultaneously, using a bunch of vintage 'telnet client' DOS computers.

6.3 LinFBB v7.04

Notice: Maybe I have already explained that I use Red Hat 6.2 at home. That's why I usually look for .rpm packages that have been made for that particular Linux distribution, but not only that. I have also tried to use Red Hat 7.1 but it seemed not to support an older Xwindow application, LinFBB 7.00g (04 August 1998). When I noticed that issue, I returned back to Red Hat 6.2.

• Well, I started by downloading the package xfbb-7.04-2.i386.rpm (07 August 2001) from www.f6fbb.org/versions.html
• Folks, this time I finally decided to install version 7.04 as a completely "fresh" installation, i.e. without some parts of any previously used "daemon" on the disk. It means that I have uninstalled the last daemon version I was using before, and, in addition, I also removed the old executables. Of course, before the uninstalation, I made the backup of some config files that are not version depending (like /etc/fbb.conf), in order to avoid editing the same "defaults" once again and again :-)
• The setup procedure has reported some dependency issues. I didn't want to get bored with them so I repeated the installation once again with "--force" and "--nodeps" options. [Scientific Data: IMAP]
• So far - so good. Then, I replaced a couple of "default" files with the saved ones. After that being accomplished, I mounted a FAT partition with WinFBB's system files, made a pray and started LinFBB's daemon. It was also an interesting new experience to try HI8GN's script /usr/sbin/fbb start (activated in an xterm) to start the server. Although there were no usual lines:

        xfbbC/X server running ...
        xfbbd ready and running ...
  

  on my screen, TNC's PTT lamp confirmed that a beacon was really transmitted.

• Then I wanted to try HI8GN's script /usr/sbin/monitor to see what's going on the frequency. Although I got something like:

        Connecting localhost ... Ok
        Authentication in progress ... Ok
        Monitoring channel 0 ...
  

  there appeared no traffic on the screen. In order to really monitor the channel, I had to start another xterm and type:

        telnet localhost 6300
  

  Bingo! From the usual FBB's prompt I entered the gateway and typed the familiar "M" ("Monitor") command. Interestingly, as soon as I "telnet-ed" to the BBS, .:: hedgedoc.dezentrale.space ::. /usr/sbin/monitor window, mentioned above, started to copy whatever was going on the telnet xterm as long as that telnet session was closed. I was in doubt if that was OK or not, because there I expected to see the traffic from the radio channel - regardless being connected to the system or not. Any suggestion here?

• Well, then I wanted to use /usr/sbin/bbs, in order to connect to the client's (or better to say: sysop's) console (xfbbC). Looks that there was a line in HI8GN's script:

        xfbbC -c -f -h localhost -i [callsign] -w [password]
  

  with missing ./ (dot slash) before xfbbC, so the script was not likely to be executed. Instead of that it reported that command couldn't be found. Anyway, xfbbC v3.01 itself appeared to work nice. It is still possible to monitor the working channel too (using the "M" command from within the gateway), but this is not a valuable solution because while "Monitor ON", it is not confortable to do anything else within the gateway. Once again, solutions are welcomed!

• Although the active xfbbC session can be easily terminated using "B" ("Bye") command, a fooled /usr/sbin/monitor can not. The user has to find its process number, (PID), using ps ax command and then kill that process.
• At the end of the game, daemon itself should be stopped. HI8GN's script /usr/sbin/fbb stop returns:

        Shutting down xfbbd:          [OK]
  

  but /usr/sbin/fbb status reports:

        Checking, the FBB daemon
        xfbbd (pid) is running...
  

  Looks that /usr/sbin/fbb stop does not terminate daemon *every* time the command is executed, but re-start it (the only difference is the new PID of the process and .:: podcasts.apple.com ::. .:: logopond.com ::. ps ax can show that new PID). So, there is a question why it returns that [OK] when it is obvious that daemon is not stopped, but rather re-started.

• Well, if you are like me, you may also want to experiment with some special sysop's commands, from within an xfbbC session. For example, "/R" command ("Re-boot PC") shuts down xfbbC and /usr/sbin/fbb status reports:

        Checking, the FBB daemon
        xfbbd dead but subsys locked
  

  while "/A" command ("Stop BBS") returns:

        Stop-request accepted, no connection.
  

  before shutting down xfbbC client itself.

  Further attempts to re-start either xfbbC client or xfbbd server (using /usr/sbin/fbb start) are not successful, unless an additional /usr/sbin/fbb stop is executed. The result is:

        Shutting down xfbbd:          [FAILED]
  

  Now another /usr/sbin/fbb status reports:

        Checking, the FBB daemon
        xfbbd is stopped
  

  so finally, daemon might be re-started again. Here it is also a mystery why it returns that [FAILED] when it is obvious that daemon is really stopped (maybe it is a "failure" when we try to stop the same thing twice).

  There are some other commands: "/K" (Re-boot BBS with housekeeping), "/M" (Re-boot BBS immediately) and "/L" (Re-boot BBS, waiting users to disconnect) - all of them with slight different behavior. Anyway, those three commands have something in common: they all re-start the daemon (with different PIDs, of course).

• Finally, what I would like to have is to manage housekeeping and other maintaining tasks. Until now, that is not accomplished. I suppose that I should make some more fine customization in system paths. Any suggestion about is welcomed. .:: telegra.ph ::.