Here are some really old news entries that never got migrated to the database.
Return to main news page
225 (NEWS 416) - Domain and Website
By Alara: And we're back up! A month later. And a few more updates: I've put this server behind Cloudflare, just as a thing to play around and experiment with. If there are any connection issues, this may be the cause.
Another small item of note - in the next week or so, we may lose DNS, as I've initiated a transfer of the domain to another registrar. I'll try to get everything hooked back up asap after this takes effect.
Finally, "Connected Players: #" on the website now links to a pretty-print link (https://tfgmush.org/players/connected) which is a special case that shows online players rather than the original URL format.
--Wed, 06 Feb 2019 11:19:56 MST
224 (NEWS 415) - Great News
By Alara: Best news ever, in fact. That is the RPi is nearly recovered fully! I have a few outstanding items to take care of, but right now they are minor. But all websites hosted on this tiny $30 device are now working, and I now have a plan to have an active mirrored backup for everything as soon as I can justify another hard drive to run a second Pi.
--Sun, 06 Jan 2019 11:51:57 MST
223 (NEWS 414) - Fix the thing
By Alara: Timezones are still not being written into SQL correctly (for the ATOM feed). See if this fixes it??
--Sat, 05 Jan 2019 00:57:00 MST
222 (NEWS 413) - Further Fixes
By Alara: A quick note to let y'all know that all SQL has been recovered, and backup scripts are still working. The only major items left are actual web hosts, letsencrypt and hooking everything up from them. This should not be too bad, though. At least I hope not too bad...
--Sat, 05 Jan 2019 00:52:48 MST
221 (NEWS 412) - Upgrades, Breakages, and Fixes
By Alara: Here's an update on what's going on this week.
Alexander is visiting MAGFest this week! HOORAY! I wish I were there too. I told him to try to attend the Demoparty, because I hear those thing are pretty underground and cool.
I completely screwed up the server on Tuesday while running a suggested upgrade to MariaDB 10.1. Turns out Debian (and Raspbian) has a demo OS release out which when upgrading to that (because I thought they followed the same major release schedule as Ubuntu, Windows 10, etc.), completely borked the system as far as MySQL, Apache, and PHP goes. Basically my LAMP install bit the dust. So I am spending the remainder of this week refreshing ALL of Raspbian from scratch.
This includes upgrading to a RPi3+ which I've been hoarding at home as well as migrating to HDD-Only mode (this is apparently programmable in the RPi3's internal flash, so you can pull the SD card and it will still function)! Additionally I got Raspbian Stretch installed by default, as well as MariaDB 10.1 running as the base SQL server.
Compiling PennMUSH last night was a real pain. Turns out it couldn't guess that the SQL client libraries were located in /usr/include/mysql, and so I followed PennMUSH's SQL instructions to "specify mysql_config via --with-mysql=
" but did not realize that was supposed to point to /usr/bin/mysql_config. I thought it was supposed to be /etc/mysql/my.cnf or something!
Well, after that was clarified around 11pm, the engine FINALLY compiled and I spent the next hour merging MUSH configs in with the newest distributions.
Outstanding issue: All old web hosts are currently under construction. I need to reintegrate all the old website files and reconfigure SSL (probably will just generate new certs from scratch, rather than try to recover the old ones, since certbot is free). Additionally, I need to get cron back up and running in order to properly perform daily backups and other periodic activities.
MUSH-wise I decided to turn on the config option to warn on all rooms disconnected from #0 rather than on rooms without exits leading ino them. This means there were about 50 rooms this morning that were complaining about disconnects and unfloating. I am slowly working my way through them which leads to today's big(ish) in-game update:
The Sky is now mapped! I thought for sure I had already gotten this, but apparently not. Plus working my way through disconnected rooms, I ran into a few old grid projects which I'm going to gradually open up. A few are easter-eggy, so I don't want to spoil everything by giving you all details here. Just have to go searching, or in ten more years, randomly run across them!
--Sat, 05 Jan 2019 00:41:46 MST
220 (NEWS 411) - Christmas Box and GMCP
By Alara: Hiya, all! I was just working on a few fun things. And annoying things. And frustrating things. But I think everything works now!
1. The Christmas Box was not correctly releasing all the wandering creatures.
More realistically, the wandering Christmas creatures could not find the door because they didn't realize they could walk through the only exit out of the Christmas Box due to the inability of testlock() testing EVAL-type locks. Turns out elock() works much better for this.
2. GMCP is actually a working thing right now! I only have one module available: Exits!
Set OOB`EXITS=1 on your character to turn on this option. Of course, your client will need to be parsing incoming json data over Telnet Subnegotiation.
Check out HELP OOB and https://www.gammon.com.au/gmcp for more information on this. I will see about also publishing my MUSHClient scripts on the website for the saavy folks to pick apart.
--Sun, 09 Dec 2018 12:46:03 MST
219 (NEWS 410) - Crashes
By Alara: I've been attempting to track down what's been causing the game to crash about once a week. So far it's been completely untraceable, due to logs telling me that 'it was the parent process. Oh no!' and the crash happening at very weird times of the day. Therefore, I've done a little digging into my backup scripts and have added some datestamp logging to a few key points in the process. Then reading through mush.cnf, I found a line that said that 'forking database dump doubles your memory usage. Do not use if you are low on memory.' I have no idea what my memory usage is on my little RPi, but I imagine it's not THAT big. Therefore I've turned off the forking, just to see if I can run for an extended period of time. This also means that once more, the game will freeze for a few seconds while we back up the database every twelve or so hours.
--Mon, 08 Oct 2018 08:17:04 MST
217 (NEWS 409) - HTTP
By Alara: Woo!!! 1.8.7 totally adds this HTTP thing, as well as a websocket (which I have yet to figure out). HTTP Handling is totally awesome! This is WAY better than my plan for a REST API via MySQL. By creating an "HTTPHandler" bot player who is assigned as such in the config, I can now code it with a bunch of GET/POST responses for someone viewing on a browser! For example, check out RIGHT NOW tfgmush.org:3349/who or tfgmush.org:3349/look?name=here and gasp in awe or something.
--Tue, 14 Aug 2018 17:43:05 MDT
216 (NEWS 408) - Social Networks
By Alara: The Social Network list in the +FINGER code should be fixed! For whatever reason it was giving up around Facebook or so, but all options should be back.
Additionally I added an easy Discord one in case you want to share your username for that.
I could stand to turn off a BUNCH of these though... They're so uncommon!
+HELP FINGER SOCIAL has all the details you should need.
--Thu, 14 Jun 2018 18:30:45 MDT
Previous 10 | Next 10
214 (NEWS 407) - Quick note on Ambiance
214 (NEWS 407) - Quick note on Ambiance
By Alara: I decided to also add on a fast little feature to ambiance: if you're idle for more than an hour, all these messages should halt until you're active once again.
We have players that remain logged in for days, and I imagine they'd appreciate coming back to see what's ACTUALLY been happening, rather than seeing days of ambiance messages filling client buffers.
--Thu, 07 Jun 2018 09:37:44 MDT