Existing players used to logging in with their character name and moo password must signup for a website account.
- Thia 9s
a Mench 8m Doing a bit of everything.
- Fay 39m
- Raven 4s I lost myself, in the dark charade.
- Sulfurado 33m
- Sadegh 2s
- Balreskan 46s
- Rillem 28m Make it personal.
- PinkFox 2h
- Pladdicus 42s uh
- vantablack 4m
- Sivartas 48s
- zxq 57s Tools: https://ansicolortool.neocities.org
- Hivemind 1s
- Ameliorative 6s
And 20 more hiding and/or disguised
Connect to Sindome @ moo.sindome.org:5555 or just Play Now

[Jan '26] Improvements & Bug Fixes
Rolling thread of updates and bug fixes for January 2026

NODE COLOR TAG TB

Fixed a TB that would occur when a color tag was formatted improperly. Now it will give a helpful IC message about the issue.

Train Board

You can find train boards at certain stations to see levs.

Kinda cool. +1

MAX NODE NETWORK DEVICE TAG COUNT

I've updated the max node network device tags per node from 2 to 4.

TRAINS AND SHUTTLES GRID ENABLED

The Mag Lev trains, Freesky shuttles, and Freesky/Kashflo shuttles are now all grid enabled.

Go forth and FOIC.

Train Board

Felt bad that certain other stations didn't get any.

Added some more train boards.

MAG LEV UPDATES (AND SPACE SHUTTLE TRANSPORTS TOO)

While we were working on getting mag levs grid enabled, we realized there were some issues with the system. The mag-levs and shuttles to freesky/kashflo run on the same system, using the same train manager. The code worked in such a way that each train ran its code one after the other, waiting for the previous to finish. This was 'working' but it meant that as we added new trains/shuttles the whole system actually got slower, as the system would wait for the entire train to move stations before the next train would move. This accounted for about 34 (sometimes 64) seconds of suspends / waiting per train.

This was not intentional.

The updated system uses departure times and runs concurrently, so each train/shuttle moves independently and not sequentially.

Since trains can move at the same time, this means that two trains might want to be in the same station at the same time. Just like in a subway, this means the train that gets there second will wait in the tunnel until the station becomes free. Usually this will take 15-45 seconds.

However, all in all, a lev ride should be faster.

Best case, for a mag-lev it should take 12.71 minutes for a round trip. Worst case, if it runs into the other lev a bunch already in a station it wants… maybe 14-15 minutes.

A few general things to note:
- Trains will wait at stations with doors open for ~60-75 seconds
- Space shuttles will wait for 120-135 seconds or so.
- The E6 and E7 will not be on the same 'timing' as they were before.
- Trains will likely have to wait in tunnels once or twice during a round trip, depending on timing
- There will now be more predictability in when a train will be at a station, and when it will leave

The lag-levs™ are still 'slower' than other transportation, and that is intentional, but they should be running faster overall. If they end up being TOO fast, and this impacts taxi businesses for PCs, I will make them wait at stations longer.

(Edited by Slither at 10:04 am on 1/3/2026)

MAG LEV GRID UPDATES

Due to the changes above, the information available to the Grid about trains and space vehicles has been tweaked slightly.

Injury Reminder System

Added an automatic injury reminder system that periodically notifies players about their wounds until they are fully healed. This should help remind players that their characters are hurting after recent combat or injuries and help prompt them to RP their injuries better.

Players automatically receive reminders about their injuries based on severity, and the reminders rotate through all injured body parts (head, chest, left hand, etc.), with each of the 23 body parts having a unique message. More severe injuries = more frequent reminders, and the system stops when the player is fully healed.

The reminder frequency is based on the severity of the injuries (i.e, how little health you have) and can be adjusted if needed, but as with all my changes, not for a good month of play testing.

As always, please @bug if you get any random tracebacks, but I think I have everything covered so far.

(Edited by Mench at 5:16 pm on 1/3/2026)

MARKETS & CLOTHING & APPRAISE

If you are in the market, and you appraise a piece of clothing, you will now see what the clothing is worn on and the clothing's messages (wear, worn, tease, remove, etc).

This should make browsing the market for clothing much more productive.

KNOCKED 7 MEGS OFF THE DB SIZE

I was looking over flyers/newspapers/resumes and saw that even though these are written to disk and lazy loaded, we never clear the description property, which is what holds the content when you 'read page from paper'.

I've added clearing this prop to the auto pruner and made it so if the prop is sparse it will reload the page the object is on from disk and save it back to the description prop. This saved us 7 megs in text being stored on MOO that was rarely used.

GRID TERM: G OPTIONS REMOVE-G-PREFIX

Usage (with a term): g options remove-g-prefix true

Outcome: Grid output will no longer have a [g] in front of it.

This will make the lives of anyone copy and pasting grid output much better.

WEBCLIENT LOCAL NODE EDITING

The webclient, and any client that supports local editing VMOO style (popping out a new window with content and then allowing you to save it), can now do node editing in a much better fashion.

If you are on your term and do 'g options local-editor true' and then 'g nodeman edit nodename' your webclient will open a popup window with an editor. The editor will be populated with your existing node content. You can then make changes and hit save and it will save the node, assuming you are still logged in/term isn't disabled/you still have access.

This should make editing nodes much easier and enjoyable.

- This feature was designed for the webclient, and desktop Firefox. It should work fine in Chrome. Any other browser or Mud Client is untested but may work!
- You may need to allow popups from sindome.org for it to work properly.
- You can have multiple nodes open for editing at a time as well, which I'm sure many node editors will appreciate.
- There are shortcuts (like command/control + s). You can view with the shortcut button.
- There is a vim mode for those that like vim.
- There is tab/editor navigation via shortcuts
- The webclient has an option for 'Editor Type' which you can change to either IDE (preferred) or Individual Windows (if you don't like the IDE style pop up).
- There is a scratch pad, but you CAN NOT save, even if it gives you the option. Admin only. But feel free to use it in the window if you want.
- We use this editor on the admin-side, so it's fairly well tuned.

WARRANTY:

This is a quality of life feature. The admin staff are not going to debug why your mud client or browser if this feature is not working. Do not ask us to. You can ask on game-help and may get help, but admin aren't IT tech support. Not all browsers will work. Mobile devices, who knows. This is designed for desktop browsers, preferably Firefox.

WEB CLIENT - WARNING WHEN CLOSING WINDOW

If you try to close the webclient tab or window in FireFox or other browsers and OS's that support it, you will now get a 'Are you sure you want to leave the page' warning, if you are currently CONNECTED to the game. If you @quit it should let you close normally. If it's not working, it's your browser/os combo and I can't do anything about it, sorry.

(Edited by Slither at 11:58 am on 1/7/2026)

WEB CLIENT - NODE EDITOR LINE LIMIT

I've added a line limit to the web client node editor. If you edit a node, you'll be restricted to 200 lines. It gets a liiiiitle weird with how it deals with it when you get close to 200. Cursor jumps around a bit. Pastes won't go through, sometimes it erases the extra lines, sometimes a few extra, so be careful around it. Not much I can do about it, and hopefully it doesn't come up much.

This should hopefully limit issues where someone sends more text than the receiving verb is expecting and that text gets treated as commands you are entering.

(Edited by Slither at 12:09 pm on 1/7/2026)

GPT NODE SERVICE UPDATE

I've updated our node service that we use for connecting to the OpenAI GPT api's. It was using the deprecated completion API, now it is using requests. I've also added some additional config that can be passed in from the MOO to allow more control over the model and things of that nature.

GPT TYPO SCANNER

I've written some fun code that will scan through rooms and their descriptions (not items in the room or things integrated with the room) and send each paragraph of the rooms description to a GPT model, with specific instructions to flag and fix grammar and typos.

After much mucking about, I have it working fairly well. It isn't perfect, because we're using a nano model as we want it to return quickly, so it sometimes messes up the output and the verb fails, but it's working 90-95% of the time.

Each time it finds an issue, it returns what the issue is, the original text, and the correct text, and presents it for the admin running the verb to review. If they think the fix makes sense they can select to apply it.

We have a lot of rooms and it takes 5-10 seconds for GPT to respond, so it'll be slow going. However, it's finding stuff everyone has obviously missed for a while, like missing words, misspellings, and grammatical issues.

I've run it on 35 rooms so far.

It's fixed 54 issues.

A few examples:

Sinn Street - Parking Lot: Corrected verb, hyphenation, and adjective form to improve grammar and consistency.

Knife Street: Added hyphenation to 'Burnt-out' and 'neon-lit', and inserted a comma before 'which' for nonrestrictive clause. Removed the trailing period after the closing quote to correctly punctuate the sentence.

Sinn Street and North Fallout Road: Removed an unnecessary comma before 'and begging for change' to fix a comma splice; sentence punctuation corrected.

South Tamiya Avenue and Blitz Road: Comma added after 'knives' to separate clauses and improve readability. Subject-verb pronoun agreement: 'Each figure' is singular; use 'its' instead of 'their'.

Death Row: Removed the unnecessary comma after 'gateway' and corrected subject-verb agreement from 'have been stripped' to 'has been stripped' for the singular noun 'outline'.

It will take me and Mench a while to run through more of the rooms, but doing so should dramatically reduce our @typo count, and save the builders a lot of time and effort.

We aren't using GPT to generate room descriptions or anything like that. Those are hand written by our beautiful builders, but typos and grammar issues– that's something I'm all for using LLMs to fix for us and let the builders spend their time making cool shit instead of working through boring ass typos.

(Edited by Slither at 2:28 pm on 1/9/2026)

TYPO SCANNER CONTINUED

I've rewritten this to be a background process that we can kick off. It runs, and collects the possible corrections, which can then be reviewed by an admin after the run is complete, that way admin are not sitting around waiting 10 seconds between rooms as we await the response from the GPT API.

I've also refined some of the words it should ignore, and tweaked the prompt a bit. For context, we have 120 or so typos in our @typos system.

Running the scanner on bars + most red sector streets (excluding those with color in the description cause we can't process those effectively yet) has already netted us 120 fixed typos/grammatical fixes.

This updated scanner should make this a lot easier, and we can abstract it to handle other objects aside from rooms in the future.

Outside The Crossroads

Fixed a bug with one of the exits, which caused the 'door' to create with a random object as the .destination property (We saw both a flak jacket and a plushie).

The effects of this bug meant that anyone who looked in that direction would get a traceback. It also meant that anyone who tried shooting in that room would see the weapon fire, see the combat messages for it as if they are doing damage (/victim seeing as if they're taking damage) without actually doing or taking any damage, and their clothes being damaged from gunshots. It also affected the sounds of gunfire, which would not make it into the room on the other side of the broken exit, and staff being flooded with some tracebacks.

TYPO SCANNER DATA

So, the typo scanner has almost finished running on 8844 IC rooms. We're at 8269/8845 and I've continued to refine the scanner as it has been going on.

Currently, we've found 2076 different rooms have issues, for a total of 2236 issues. This number is not 100% correct, as when the scanner recommends multiple changes to the same sentence, it counts as one issue.

I'd say the majority are grammatical errors and not actual typos (missing or too many commas, missing semi-colons) but there are also typos like missing words, misspelled words, and things of that nature.

There are some false positives, where we get dinged for something stylistic, that we actually intended, and the rare case where a word is corrected to a different spelling when we don't want that, or capitalized because it is technically a proper noun, but we don't want that, so not all suggestions are getting approved but the vast majority are.

The completionists on staff are quite excited to get through all of them and have our room descs be super grammatically correct :)

Fixing these has already begun via reviewing the suggested changes and accepting or declining them.

1100 TYPOS FIXED

We have fixed 1100 typos and grammatical errors thus far. 2085 to go.

Wheel

I found a wheel the other day. Decided to make it into an actual thing.

If you know what it is, you'll know where it is.

"Dj, Spin that shit." ~Yo Momma, Probably.

(Edited by Mench at 7:36 am on 1/20/2026)

TYPO UPDATE: 1868 TYPOS NOW FIXED

We've fixed an additional 868 typos. We now have 1682 entries remaining to review.

Ringtones

Added some new Ringtones to the Ringtone app for the Progias.

@trust-autoclear

The trust auto-clear feature allows you to automatically clear your @trusted list every day at 11:59 PM. This is useful if you want to ensure that your trust settings reset daily for security or roleplay reasons.

COMMANDS:
@trust-autoclear - Check current status
@trust-autoclear enable - Enable daily auto-clearing
@trust-autoclear disable - Disable daily auto-clearing
@trusted - View your current trust settings

WHEN IT RUNS:
The auto-clear happens every day at 11:59 PM game time.
If you're online when it happens, you'll receive a notification.

IMPORTANT NOTES:
- This is an opt-in feature - it won't happen unless you enable it
- Once enabled, it will clear ALL your trust settings daily
- You can toggle it on/off at any time
- The clearing persists even if you're offline


Edit by Slither:
- There is no guarantee this will run at exactly 11:59. It will run /around/ that time.

(Edited by Slither at 9:40 pm on 1/28/2026)

Afterthoughts

Going forward, records will not include the @title of the person making the record; instead, they will use the bit name, so instead of 'scary joe' it will just be 'joe'.

Thank.

SPEAKING QUEUE FOR TOWN HALL

Updated the town hall speaking queue so that you can't leave it as the current speaker, which should help us from accidentally skipping people, which has happened in past years.