#meshing-around BBS/utility BOT
1 messages Ā· Page 9 of 1
The issue being, the confusions around sudo
It seems to be making a lot of problems (likely me)
The service should be fired up on meshbot
back in the day just used pi and let everyone figure it out
so... just to clear up any confusion for me.. We we installing using the "pi" user (sudo ./install.sh)?
It really all depends, on your system. Everyone has an opinion I feel
As long as it works for most Iām happy
Iām worried broke it
so I installed as pi.... logged in as pi..... can't access the meshbot owned file/directories
You can, just depends
Everyone has an opinion here
Itās endless
I say run as root and fā it
itās just a bot
running with sudo... runs as root doesen't it?
Runs as sudo
Which is you with more access
Like UAC in windows
This was an attempt to fix any foxās which might hiccup
Again all depends
Some donāt update much or do it via other methods
A git pull wonāt experience this alone
Either will a release download if you just load it in without install
so... I understand what/why...
still not sure why my pull this morning broke my installation... maybe a permissions thing?
I think that maybe the installation instructions should give some guidance on what user account you should be using to complete the install. It will help standardize things going forward. I really do not like installing application as root
but... I can if I know I have to do that
Hard to say post install but yes my guess is permissions? The logs all archive so seeing why the bot shut down would be noteworthy
I did just churn out a ton of refactor and enhancing
it never restarted after the pull
If you got me on a oh shit moment it could had a bug
there was nothing in t elogs
Cool
Cool, if it has any other wacky holla
now its just a permissions issue around editing the files, accessing the directories... I hate to sudo su... but it is what it is
If you install, and use the logged in user and not meshbot
So if you are pi, ignore meshbot user use pi
right... didn't know at the time.... Should not have created the meshbot user
Another wild trick is just chmod +777 config.ini
I'm gonna uninstall again and re-install to test the permissions..
yes... but trying to test the installation scripts
question... It does not appear that the "w3 server or reporting service" is installing on my pi when I run the install script (no errors noted). Is there something that I should be monitoring>
Had to chown user config.ini after a pull yesterday
Correct Iām not sure it was working correctly to update and havenāt built a new box to test yet
I commented it out from install
But it should be there and ready to light up?
That whole process and engine needs a review I havenāt for months
This was due to likely the original install not setting this timely
Thanks
I will test...
Everything else is working normally after the re-install. Just got a single warning noted in the installation : " WARNING: Retrying (Retry(total=4, connect=None, read=None, redirect=None, status=None)) after connection broken by 'ProtocolError('Connection aborted.', RemoteDisconnected('Remote end closed connection without response'))': /simple/dbus-fast/"
I have been seeing this randomly on UDP more
MeshD
My logs are all WiFi issues
From meshD which isnāt WiFi
Hmmm.... I review my Wifi logs too
In my case I have a bad MQTT IP isnāt helping
I also dare say the overnight solar was intense
Pushed a release with the reporting timer back in
Solar storm?
Oh yeah I saw that. I was thinking maybe we had something headed our way also that I missed.
Kinda do the next 48? Worth keeping an eye on
āYesterdayā was huge I saw it in rF and earthquakes
Whatever happened .. combined with that huge thing in Jamaica .. woah crazy overall || and bill changing step ||
Am I right in my understanding that all of the custom messages are now using the channel defined in the config.ini?
They use whatever you set
Or should anyway
Checkin reminders might not be working, need to check that
This was always the time consuming part for me/ the logic was fast - I got not logic to it yet just plays dumb
Makes a nice chess
Alerting has a bug
Now patched
@somber condor
Works again haha I fell asleep waiting for this last night
I should stop changes after dark
Shit I LITERALLY finished installing then read this š
If it helps i let out a loud moan when I saw this
Will do again right now. I'm guessing this is part of what you fixed?
This is odd
If i do as su, will it break anything?
Ran it again and it worked fine
šš»
If you are installed in /opt/ https://github.com/SpudGunMan/meshing-around/blob/main/etc/set-permissions.sh
Give this a whirl
It will reset the owners of stuff
If checklist isnāt working delete the .db and just start a new one itās very possible the schema is bad
Ok
Itās in data/checklist.db
So⦠the python code references āscheduler Channelā, which I assume is the channel defined in config.ini?
I know it can be changed. But the default isā¦..
Thanksā¦. So do you see any value in allowing users to add channels of their own in config.ini (ie: alert, chitchat, etc.), and reference these by name in custom_scheduler.py?
You should in custom_ just be able to replace with 1,1 or whatever
so use ini default value or set one
There is no my_var
But there is no 1,1 anymoreā¦. Maybe I am misunderstanding
Gotcha
Alright that all worked.
Is there a way to check the location of a check-in over the mesh?
Havenāt thought of a good way yet to handle that concept
Aloha -- using ollama in a few ways on a femtofox and RPI..
Running Ollama has been hit and miss -- while wiki kind a ROCKS hihi
INFO | Device:1 Channel:1 SendingChannel: Average query time is: 42 seconds | WARNING | System: LLM Query: Ollama API request failed: HTTPConnectionPool(host='192.168.86.24', port=11434): Read timed out. > | INFO | Device:1 Channel:1 SendingChannel: āļø Request Error
I get this a lot of this on both configs -- I run Ollama on a pi5 15gb -- and link meshing-around -- I wonder about the network latency of this... When I use Ollama where it's installed its quick and responsive.
any thoughts..?
Just playing
It would, show up on the map reporting
It likes telling you who it is
-- this is reliable output--
Yes just run ollama
Is this only with longer questions
Wiki added hints?
urlTimeoutSeconds * 5
This is the timeout for the request
Do debug logs look ok no weird errors about multiple requests
I may be using the pi too much
What is your llmUseWikiContext =
I don't have that in the ini--
wikipedia = True ollama = True ollamamodel = gemma3:latest ollamahostname = http://192.168.dude.blah:11434 storeforward = True
I had to run it in debug and... yeah -- kind a cool and š¤
INFO | Device:1 Channel:1 ReceivedChannel: askai hilo hawaii is located where? From: wh6gxz meshtoad | DEBUG | System: Bot detected Commands:[{'cmd': 'askai', 'index': 0}] From: wh6gxz meshtoad isDM:False playing:False | DEBUG | System: Location data for 2022076085 is 19.4248704, -155.2154624 | INFO | Device:1 Channel:1 SendingChannel: Please wait, response could take 30+ seconds. Fund the SysOp's GPU budget! | DEBUG | System: LLM Query: Using Ollama API for LLM query hilo hawaii is located where? From:2022076085 | WARNING | System: LLM Query: Response exceeded 450 characters, requesting truncation | DEBUG | System: Splitting #chunks: 2, Total length: 186 | INFO | Device:1 Channel:1 Chunker1/2 SendingChannel: Hereās a shortened version: Hawaii is located in the **Pacific Ocean**,> | INFO | Device:1 Channel:1 Chunker2/2 SendingChannel: large archipelago of islands. | INFO | System: šŖ« New low battery record: 99.0% from NodeID:1839130823 ShortName:HS
OK.. really enjoy meshing-around š„ š«
I'll be back-- time for kaukau
This surprised me. Someone played a game that I didn't even know existed. Granted I'm not much a gamer but is this new?
Do I have to give them a real prize?
Haha awesome I slid it in
Itās to encourage more chatter on channels
It collects use of emote
it may be used in a larger game some day
For now just for fun š
If you think of any enhancements holla
There is also a word of the day and a BINGO word search
That is hilarious. So, someone picked at random was singled out in our local channel!?
That is wonderful, Brilliant
Fixed up some things, of course broke some working late ā° new release.
I removed the logic for checking permissions itās causing problems for some
tidied up some random issues
I was thinking about that this morning. I checked the leaderboard and it was me and I'd only messaged the bot LOL
I knew someone else had chatted a lot and itās how I got it. Saw the same thing
Iām going to (try) and only focus on fixes for a bit no enhancements
You've been rolling out enhancements at a pace that's never even been heard of before. That sounds reasonable haha
As long as you aren't fixing something then suddenly think of an improvement, and go down that rabbit hole. That's my life typically.
Haha yes that has happened
I wanted to get a bunch of the basic ideas in before switching gears to this pygame stuff
Iām still not 100% sure how to go about the bots futures
The legacy onReceive or the new on_receive
We need a meshtastic developers conference
Some AI powerpoint we can ask AI to summarize the changes or futures with API
Do think copilot is tuned enough to modularize this a bit
Its current state itās just ideas into prototype quickly
I wanted to learn how copilot worked
This is the fun project
So, next major is what and how to subscribe to proto.
Once that is ironed, I wanna make a new core with like ping and echo but it will be able to use localization files
And then convert modules to plugins
Possibly a 2.0
TBD
Iām a potato, this project is my primary job
What are they changing with onReceive?
I'm with @gentle kestrel on this journey.... I'll break it for you haha! With aloha from nurse dude 73

Thanks dude! (And to all) itās stable thanks to your efforts!
I was too late and the continental breakfast was all gone. As were any conference goers. But I found a flyer on the ground that said
TLDR.
The old 'meshtastic.receive' event provides both the packet and the interface object, so you can access routing/interface info.
The new "mesh.rx.packet" event only provides the packet (no interface).
.
I interpreted this to mean, either one or the other
(I have yet to determine if there is a good path with)
# Extract fields from protobuf
from_id = packet.from_
channel_number = packet.channel
I noticed scheduled š alerts were not going and squirrel šæļø
My problem is, if I can see multiple RF units, I need to route
Decoding flood is ā
If I do and send it down the RF
I can detect channel and assume routes, but I am waiting to see what goes down
The TLDR didnāt include the 2.8 notes šļø
I heard from one of the lobby staff they all left in a rush so š¤·āāļø
I donāt really wanna write a 1.5 core skip right to 2.0 if you will but
depends, if it works well in a lab
2.8 happening soon you think? I'm super removed from that discussion
Iām gonna try a double sub idea
Echo got enhanced
Admins now have super power
Send anywhere @drowsy rock
echo Hello world c=1 d=2
Updated docs, can also now echo motd for string exchange
If other echos are interesting hollaā
Hell yeah!
https://github.com/SpudGunMan/meshing-around/issues/252
Enable or disable automatic banning of nodes
autoBanEnabled = False
The bot will now ban any in the ban list
It will ban_hammer check violations
Such as āping 999ā
If you wanna test it
bannode now works
And the primary node ban list is a .txt which can be found in data/
You can synch via whatever
The old .ini will migrate you can remove then when they do
Helps keep the ātricksā down š¬
Refactor to the highfly for better altitude detection and also added flight source country to output š«
When you send the bot a ping, it responds with pong. What are the letters inside of the brackets? For example, I get "PONG [GW]" but I've noticed that other responses can contain other letters. Do those letters inside the brackets indicate anything?
- SNR: Signal-to-Noise Ratio (dB) ā higher is better.
- RSSI: Received Signal Strength Indicator (dBm) ā closer to 0 is stronger.
- [GW]: Message received via Gateway (internet/MQTT).
- [RF]: Message received via direct radio.
- [F]: Message received via mesh/flood route.
Do you remember offhand how to check somebody out via admin? Driving so can only operate that briefly
approvecl <id> - Approve a pending check-in (admin)
Let me know how it works out in practice
Different command needed because itās confusing etc
I havenāt touched it since it was working, the auto checkin was the last part
denycl <id> - Deny/remove a check-in (admin) it will hard dump any id
Ok. Will try. Arrived home and UN checked in manually. Accidentally checked in wrong node before I left and was getting alerts. Had to go do what i thought was an update mission, turns out, node fell 100 feet from a tree and everything shattered.. It was actually still powered on and working but antennas wire disconnected upon impact. Reconnected and works fine. Just need to rebuild.
Ahhh!
.
All the sensitive APIās now throttle
There is no clear but to restart bot right now
Setting apiThrottleValue = (in) autoBanTimeframe
This is unrelated to any ban
Just blocks by default at 20/hour
Havenāt rolled to USA feeds only the more sensitive ones
š¦System busy, try again later.
Also its global
If user runs 15 Whereami and also tries 6 latest, they wonāt get the last one
Cool down
If exceeded and autoban will pass ID
Um. something is weird.. did a git pull and went to add node to ignore( one of my favorite things to do) config totally missing most everything that was commented. settings out of order. wacky. and eas now coming through on wrong channels.. and it didn't ignore the high altitude node I just added.. what did I do?
A typo on config.ini and it went nuts?
Is config.ini a mess?
I did, update how EAS works by default but ⦠it should have only added EAS to your emergency channel
Iām open to thoughts on how to better make this flexible to needs
Before it was not something which can 1:1 well in future versions
Check config_new.ini made by update.sh which shows a config prepped for latest options
You can see other .txt for more info
Reference .template for instructions
Some day I hope to have this better but I was learning meshtastic and and and so itās kinda a evolution
That .ini file is large and messy
I havenāt had an idea of prettify yet
Been working on project X
I just worked out a bunch of junk of it and I was working in main so Iām crossing my fingers the issues you have are easy hihi
possibly, I had a space after comma in highflyingignorelist . oops. I haven't found any other typos. Before wxalertbroadcastch was controlling weather but now it seems to be eaalertbrodcastch . eveything seems to be working again. just caught me off guard.
project x?
Yes this tracks sorry for mess, I only documented it in the push with āoh hey sorryā hihi
ā
the graphic one
I have a game display server which I havenāt looked at for a bit but itās running on a zero real nice I hope
I have the 0.9 display manager I think
I was gonna drop something this evening but I find lately if I feel tired my code is really shit
So trying to not code tired
Gotcha Neat! Yeah, I understand the tired bit. You've been doing tons of work. Sorry to have bugged ya without checking documentation or logs first.. We'll go with tired.
No worries at all
While on another thread had a trillion ideas for project X
Which just became project x,y,z
stay tuned for more
I need to look at this again
Itās a dev kit meaning I havenāt test outside of my lab
It uses UDP
So the node with the display needs to be a āpi clockā running script/game_server.py
(This is network config on you)
I have no idea how well this works - this is test idea 1
āIt works for meā
Added some details to the game help
Minimum hardware (dev on) pi zero w
How's this displaying?
The high level is, the bot is playing a game (like seen before just script running in text_message) but it will send the display data packet as a port.256 data (so the user isnāt bothered in mobile)
This packet is being sent to the same device as the player was seen on (using channel PSK) via the API. so far this is basic mesh api.
Over on a system with pygame and a display is running mUDP which is listening for packets and consuming the display driver info to process and run in pygame
launch.sh display will venv
Ok, so pygame...
In theory it could be boiled down to an esp as well
I havenāt seen a lot of issues so I assume itās working outside my dev box š¦
More ideas to follow as time allows
Had a derail this week so it may be a busy week not nerd š¤
Same here ... utterly smashed busy
Realized the data files are being overworked by git
Refactored news reading
FYI, git pull && bash update.sh
Fixed bee as well to properly buzz
Make sure to git pull
Update to fix the data path
Or just a update 2x
` git pull
Updating 38bef50..1122d60
error: Your local changes to the following files would be overwritten by merge:
Please commit your changes or stash them before you merge.
error: The following untracked working tree files would be overwritten by merge:
Please move or remove them before you merge.
Aborting
`
also tried -- a few other git pull --variables... I've been in this place before- hihi
ok fixed it -- i think haha
$ sudo git reset --hard origin/main HEAD is now at 1122d60 verse command wh6gxz@hawaiimeshmap:/opt/meshing-around $ git pull Already up to date.
the git pull -- vs sudo git --
to sudo or not to--
meshing-around is getting close to 2.0 WOW what a ride it's been - glad to be part of it.
still messing with ai stuff and digging the leaderboard, readrss, wx, valert and... also ran update.sh - it's all good 
Good to hear!
Been busy as hell with this, among other things. Wanted linear rails, didn't want to spend $250. Instead I spent $60 and about 8 hours fine tuning it to actually move smoothly. š
Pi is responding via bbs but can't ssh into it. š¤
The new Sysinfo will tell ya if IP to internet is hot
I'll have to update it tomorrow if I can't get in soon. Walking to the attic will wake the wife. That's a big no go zone at 1:30am š
Haha, sorry I tried for an earlier release but I spent an extra hour testing!
This one has a surprise in it, itās mostly for āearly play testing and commentsā
I just got done setting up the printer so you still beat me
I always have issues like this on Wi-Fi lately. Smart bulbs that randomly refuse to function when other times they're fine. Turning mobile data on and off doesn't help, switching between 2.4 and 5g Wi-Fi doesn't help. Force closing the app doesn't help. Then other times they just work. Still other times it'll be only one or two bulbs acting up.
I'm likely compromised in some way, just not sure where exactly. Even tried brand new bulbs
Heh; I use as much copper as I can
Iāll take inconvenience of a wire over frustration wireless
I'm going to need to do that. Especially with cameras. I have nice PoE ones with color night vision that seem to be more reliable than my wireless ones in use currently
Totally or make different networks and plan out channels
But still one thing goes nuts and the xyl needs the toner replaced or something
Today I got a home IT call for SIP trunk issues - oooof
This world of iOT can really cause havoc with storms of traffic and such
Keep a WiFi for the junk and a WiFi for new new shit only
Most of the IOT stuff donāt need fancy WiFi anyway
Causes some of it to bork
Back in the day I was on zigbee all over the house loved it but when it failed /ā FF pors s suuf galark
Fire fire
FUCK ITS 3 AM
Yanks pi off wall
That's another thing i need to do, segregate networks. So so so much to do, all the time haha. I'm gonna go to bed and try not to wake her. š¤š»š¤šš»
Thanks for the breakdown.
Just found this in the documentation. Read the manual, I guess, huh?
Meh, thanks to you, asking itās there!
checkin_triggers = {"checkin", "check in", "check-in"}
checkout_triggers = {"checkout", "check out", "check-out"}
if any(t in message_lower for t in checkin_triggers) and not reverse_in_out
or any(t in checkout_triggers for t in checkout_triggers) and reverse_in_out:
interval = 0
actual_comment = comment
trigger = next((t for t in checkin_triggers if t in message_lower), None)
if trigger is None: # not a check-in, skip to check-out
pass
else:
after_trigger = message_lower.split(trigger, 1)[1].strip()
parts = after_trigger.split() if after_trigger else []
if parts and parts[0].isdigit():
interval = int(parts[0])
actual_comment = " ".join(parts[1:]) if len(parts) > 1 else ""
result = checkin(name, current_date, current_time, location, actual_comment)
if interval > 0:
set_checkin_interval(name, interval)
result += f" (monitoring every {interval}min)"
return result
elif any(t in message_lower for t in checkout_triggers) and not reverse_in_out
or any(t in checkin_triggers for t in checkin_triggers) and reverse_in_out:
all_flag = False
checkin_id = None
actual_comment = comment
trigger = next((t for t in checkout_triggers if t in message_lower), None)
if trigger is None: # safety net
trigger = next((t for t in checkin_triggers if t in message_lower), None)
after_trigger = message_lower.split(trigger, 1)[1].strip()
checkout_args = after_trigger.split() if after_trigger else []
if checkout_args:
first = checkout_args[0].lower()
if first == "all":
all_flag = True
actual_comment = " ".join(checkout_args[1:]) if len(checkout_args) > 1 else ""
elif first.isdigit():
checkin_id = int(first)
actual_comment = " ".join(checkout_args[1:]) if len(checkout_args) > 1 else ""
else:
actual_comment = " ".join(checkout_args)
return checkout(name, current_date, current_time, location,
actual_comment, all=all_flag, checkin_id=checkin_id)
trying this
Did it work? I tested it once
Figured it was good didnāt hammer on it
This was pure copilot so
Itās logic in unique circumstances may not work
If you donāt have any news or whatever update and say yes
whats the uninstall switch again?
Install - - nope
No space
But it should survive an update
I keep a few I beat on and itās normally very good at making sure runtime is clean
This is the way it looks without auto checkin
This is checked in
I hope the colors are ok?
I guessed they would work across all
If auto check TRUE none of this dance
reinstalling presently
Sorry
Only checked in will beacon
Is it possible to log in as meshbot and is that recommended? If so what's the password?
Itās passwordless
So by my config itās not login-able
You would need at minimal a password for it
Ok
I did confirm all checklist works as expected latest pull
you had a permissions script didn't you? mine is read only which i discovered upon trying to save
i just did chmod +777 for config and that worked
None of that fixed it. It's just not responding to anything even with a fresh install.
Unrelated to the bot I'm sure. I'll work on that now LOL. Always something.
You know, I just connected that new HDMI switch....
I bet it's causing issues
Unplugged it and I was just able to SSH for the first time in days.
Was it spewing RF?
Not sure. It was powering the pi even when the power is off, so that's not good. Bot still not doing anything. That can be a problem for another day now
All right it took me about 5 seconds to figure it out after I got some sleep. It's not launching when the pi boots. Not sure why but I launched manually and it's working fine
Seems system startup is messed up in unrelated categories too. Wifi isn't starting until i enter the gui, which explains no ssh. Being able to ssh is because i was already in gui when i removed that switch. That switch powering the pi definitely wasn't good though.
I wonder if my node or the bot means kilometers per hour. I only got to a Max of 109 but it says 153. I would have passed 95 on the way to 109 so that could be
Is this a femtofox?
I would guess the gps values are not perfect
groundSpeed'] * 1.60934
Itās what the value comes from
Just a normal pi4
The gps value?
Iām translating the packet that way on the bot
The value of ground speed is I assumed kilo?
I thought once inspection on it was
Oh I'm not sure. Gimme a sec
Position log.
altitude satsInView speed heading
552 8 5 275.36
674 8 6 279.12
440 8 4 295.43
574 6 0 347.95
461 6 10 197.41
461 6 67 352.3
469 6 86 14.17
470 6 70 5
486 7 109 19.18
476 7 0 229.11
485 7 63 198.14
482 6 76 180.16
471 6 59 183.43
468 7 0 214.31
465 7 0 220.9
489 8 0 220.9
468 9 0 303.67
Matches the 109 on my dash. Odd that the cars speedo is accurate ha. Normally they're optimistic a few percent.
Shit well thanks
Hi - I have a feature request.
Would it be possible to have weather alerts either elaborate or provide a link that elaborates beyond the header?
Is there an existing way to automate a message to a specific node using the custom_scheduler? I donāt want to blast a message to the whole channel. Iād like to DM one specific node using the custom_scheduler.
Yes, in the code examples today.
schedulerChannel, 0, schedulerInterface
If you set channel to zero
And put the node ID where that zero is. It will send via DM
I will try and add an example/comment to reflect that as well!
Good question: specifically these special alerts are the ones with this I will look into them!
If nothing else, would be good to put effective time vs issued time, and the locations/counties it applies to.
Not all alerts hold this data and then space limit decisions.
The fema alerts are much more in line with always providing but EAS uses the codes to set boundaries (it seems) but itās all local group dependent.
I feel like these special alerts are more just come read the web page notices
Its cold out and its the first time this year check on the elderly is what I have seen in the past
Itās bot āemergencyā but itās use of the alert for āspecialā
For all alerts itās tough but I recall seeing http:// I think for special I kinda recall choosing to not because of formatting issues with clients
In this case itās definitely not-not a thing
But getting them myself Iām always like whatās this for
Well / it depends again why I mention all this
Itās an alert over the EAS not IPAWS
This alert in Oregon was for almost the whole west coast code
They also issued like 5 others
No links and here is why I have issues including the impact area
Clatsop County Coast; Tillamook County Coast; Central Coast of Oregon; North Oregon Coast Range Lowlands; Central Oregon Coast Range Lowlands; North Oregon Coast Range; Central Oregon Coast Range; Lower Columbia River; Tualatin Valley; West Hills and Chehalem Mountains; Inner Portland Metro; East Portland Metro; Outer Southeast Portland Metro; West Central Willamette Valley; East Central Willamette Valley; Benton County Lowlands; Linn County Lowlands; Lane County Lowlands; Clackamas County Cascade Foothills; Cascade Foothills of Marion and Linn Counties; Lane County Cascade Foothills; South Washington Coast; Willapa and Wahkiakum Lowlands; Willapa Hills; Cowlitz County Lowlands; North Clark County Lowlands; Inner Vancouver Metro; East Clark County Lowlands; South Washington Cascade Foothills
7 EAS packets later the mesh would be mushy
I really wanted the beacon wxa one packet
Wxalert adds more
@eager zephyr the command wxalert should give
<summary>A cool and unstable airmass over northwest Oregon and southwest Washington today will result in scattered showers and thunderstorms through Wednesday evening. Stronger showers and thunderstorms will have the potential to produce locally damaging wind gusts up to 60 mph and/or a funnel cloud or brief tornado. If you see a funnel cloud or a tornado, or if a Tornado Warning or a Severe Thunderstorm Warning is issued for your area, you should go to an interior room on the lowest floor of your residence. In addition, you should take actions to protect your head, such as getting under a sturdy table. If you are outdoors and cannot get to shelter, get in a ditch or culvert and cover your head. Never go under an overpass during high winds or a tornado.</summary>
This isnāt sent in the wxa for space and time issues
If emergency less packet
Hmm. Maybe just something in the alert message then that says "do XYZ for more info"?
Itās a good question; the local EAS team decides things I think.
In a ipaws cap there is I think fields for this but an eas cap ..
Here is the EAS cap data
Look for āspecialā to find the one
You can see the data we have to work with
If <valueName>WMOidentifier</valueName> display?
<value>WWUS86 KPQR 051758</value>
Looking now.
While I'm doing so, some ideas:
Just add "dm wxalert for more info" - but then you may have a bunch of people DMing at once which is even more traffic
"react for more info" - if anyone reacts to the message, broadcast the elaboration.
<valueName>NWSheadline</valueName>
<value>STRONGER SHOWERS AND THUNDERSTORMS TODAY WILL BE CAPABLE OF PRODUCING LOCALLY DAMAGING WIND GUSTS AND/OR A BRIEF TORNADO</value>
</cap:parameter>```
Another possibility: Linking to the local forecast office's page? https://www.weather.gov/pqr/
Less ideal for "internet down" scenarios, but, something, maybe in addition to other stuff.
There's also this page
The National Weather Service is your best source for complete weather forecast and weather related information on the web!
I have a patch in should still alert if you update and patch
Special Weather Statement issued November 5 at 9:58AM PST by NWS Portland OR Headline: STRONGER SHOWERS AND THUNDERSTORMS TODAY WILL BE CAPABLE OF PRODUCING LOCALLY DAMAGING WIND GUSTS AND/OR A BRIEF TORNADO
I also just refactored after some testing to include
# If title is "Special Weather Statement" and headline exists, use headline only
if "Special Weather Statement" in title and nws_headline:
main_alert = nws_headline
else:
main_alert = title
Thank you!
Welcome thanks for the idea
Very cool. Thanks.
A good note was put in for the way the service files handle shutdown and this change was implemented
ExecStop=
KillSignal=SIGINT
Edit service file and sudo systemctl daemon-reload
To ensure better close and save of stuff like leaderboard
Added a yoloV5 tool for alert.txt
This stuff is so large Iām not inclined to add it in directly but here is an idea to play with it
A new beginning
Omg I love it!
Are there instructions how to add the bot to startup so every time the power goes out I don't have to manually restart it?
If you install and say move to /opt restart installer
Run again
Say mesh service
Yes to meshbot account
Should do it all for you
You can also do a embedded install I havenāt heard it fails itās for femtofox
It wonāt install ollama
Is really all but forces the rest for auto boot
The latest installer looks fresh
Should fix a bag of old issues
Ok i did that. Should be good then.
I needed to get to a new cuda on my bot so installing fresh
I see the issues here
Fixing them
Which issues? River flow not working again but maybe it's the site again.
Ohhh I have one. On the fresh install, leaderboard still showing planes in the top speed, doesnt seem to be it's own separate category from other nodes.
Install permissions I think corrected now
At least I fixed a lot of major issues
RiverFlow api is slow just the way it is
Increase timeout
What do you mean
Ok Iām not as embarrassed by the installer now sorry for all
That wasnāt fun times when things went bad
I hope update and install are better
I need to test on a pi now
I got rid of confusing logic now it just uses meshbot service account
Sets environment all over for ssl
Bootstraps better
I hope this is it for a long while here
This should work for pi2 cam and webcams now
You do need a GPU or Pi2cam
I want to figure out a way to separate top speed of planes from ground vehicles for the leaderboard
Oh that's clean! Just did it
just ran the latest update-- on femtofox... failed in a few ways but -- it's all good - a few magic cmd's and whoola
i'll try the cmd's out -- got malama aina to do on this aloha friday.
the rpi update went FINE. @gentle kestrel YOU ROCK!
Thanks for info!! Anything I can make better?
Updated with Tesseract
OCR
Right now it will logic to say āif OCR save imageā
it will pass the data to alert.txt
But this is where end use is open ended
License plate binding model or one for birds and wild life
I would like it to count animals
Right now it has āsenseā to save any image with a plate in it better vs old 2014 method imo
This is intriguing, question is flights working ok for you (off season seams to have vastly lowered my high flyers) do you see flight numbers ok
I have a test for this but itās all only in test so Iām making sure it works
I expect to see good flight info and source country in the notice
When applicable (which I hope is quality)
I could possibly ignore speed if high fly need to look
I don't see info for all, but i see it a bunch. I'm on 2.7.6 so I lost all of my old messages
Ok idea is in
It should track any over and under now
Hell yeah! Thanks!
I made a bot prompted to be a pirate be the meshbot
I also made a model called ARRRRRL same idea but has knowledge for radio
Pirate radio? š“āā ļø
I just named a node "Kaaaaaarl" with a Llama emoji
Newb here. Is there a way to access the node with other software without killing the bot/bbs process? (It looks like remote admin works for basic config changes.)
What type of access are you looking for? I run several bots with one radio on a raspberry pi. Each bot has it's own instance of meshtasticd and they talk to the radio instance automatically with UDP.
Just basic stuff, like python CLI or web client
Hmm, I hadn't thought of enabling remote admin on the bot instances...
Nope, only one client connection per node. Meshin around uses the connection.
Ok, thanks, I guess it's not a huge problem. Is there a simple way to kill the service without looking up pid, etc?
sudo systemctl stop mesh_bot
New enhancements in the API may bring about UDP mode
Having constant derailments due to irl this way
āCanāt focus on the botā wahhh
You can use #1421172939751166062
Thanks. I saw firefly mentioned in the docs, but haven't researched exactly what it does yet
Gives you access into the node sort of like a console
So you can chat back, it can get wild so; at the moment a dedicated bot node and a node for chatting is best
In a minimal sense it would realistically be a roof, a bot and a personal node
The roof could be a bot
Question - is the system doing something non standard with the way it sends DMs?
Weāve noticed intermittent odd things like āno channelā errors, and the bot somehow responding in a different DM thread than it was originally messages from.
Not to like, a different person
But like, in the app, it shows separately?
If UDP is around, the default 0 is used today in the interface of the API
So any traffic heard on UDP will default to 0
Hope to have a fix after latest changes in API to detect channels better
You are experiencing flood route I think
You do need to keep good key management - so if the bots public key changes
That will avalanche the DM key issues and thusly no channel like situations
Is it only one function? Or UDP related possibly
The errors I mentioned are on the client's end, not the software's.
In theory there's no reason the public key would be changing that I'm aware of.
The issue is specifically with DMs, and usually once the bot DMs you (for example if you ping in the public channel and it DMs you a response) you can reply to it in there fine.
It's weird. Sorry I don't have more detailed info because it's intermittent.
DMās do flood now
Like globally
So if the bot is sending out a DM in 2.7.13
Attached via API (ignoring UDP for now)
The DM will flood into a mesh
I have heard they will travel over any flood possible
Some patches in for high fly, built a better test for it so hard to test well (find airplane fake node)
Also some other base patches needed for later
I have a code on lab branch (git checkout lab) for testing with UDP specifically there is a debug log for channel name
DEBUG | System: Packet Received on Interface 1 Channel: 2 Name: LongFast
itās printing any found channel
It should print flood packets
Interested to hear feedback before moving further
Something weird going on with ping hop count
I'm definitely hitting it direct, and it responds way too fast for 5 hops to have happened
Also not sure what [F] means here
Flood
You have broadcast flood
Try the lab branch
See what you think
Itās debug log notice only right now
Also needs api 2.7.4 as will the bot when pressed
Perhaps one thread is Legacy Channel PSK while the other is PKI? I've noticed what you mentioned but haven't explored it.
This makes sense now: I noticed decoded BBS DM's on MQTT.
It turns out I was wrong. Remote Admin looks like it works. It lets me access and change the settings on my meshbot node, but apparently the changes don't really get saved.
"broadcast flood"?
You need to stop the API/bot one node user at a time
Yea if you have multiple UDP enabled you will find it quick
There might be a 1 hop RF issue looking at like itās UDP one hop but still gives F and SNR not 1 hop
My lab is offline for a bit
Agreed. that's what I found out by experimentation. But, I had tried remote admin to the node while the node was still connected to meshbot. It showed (on the android app) that my configuration changes took. But, they didn't really.
Now, I'm disconnecting my bot node, updating it's config directly, and then reconnecting to meshbot. Works as expected.
Lab branch has updated debug logs on for UDP stuff
Also if you try and remote admin the bot it should have alerted šØadmin packet
Can you use scheduler to schedule more than one action? multiple scheduler sections? or do I need to use custom + scheduler.py?
Custom, custom is just faster and more efficient @quartz maple
I just have to parse every idea needed otherwise and I hope good instruction in the .py
Like legos
Sure, makes sense. The docs indicate that multiple events are possible, but weren't clear (to me) if you had to do custom for that.
Docs always lagging! Sorry
.
I updated the core and lab branch
Put a PR onto API/CLI waiting for decision
UDP channel stuff
Here is an example of my latest
2025-11-11 17:09:47,555 | DEBUG | System: Received Packet on Channel:254 (banter) on Interface:3, but this channel is configured on Interface:1 2025-11-11 17:09:47,555 | DEBUG | System: Received Packet on Channel:254 on Interface:3
The idea here is to help identify flood
Lab branch will print these logs
But needs a patch in API still to work with default channels
Now in main ./update
Storm holding, dark side for me now. G4 with posible G5 on the 12th
would it be possible to add fire danger reporting? i see its varies wildly per state so i understand it could be very difficult to impliment, but it would be a neat feature i feel.
I assume FEMA and EAS alerts would hold them (seen in the past) do you have a API with Lat,Long? You know of for alerts? @mental helm
I assume that'll only give red flag warnings, I'm talking more "fire danger" like this
Ie if fire danger is high or above then have it push an alert out in the morning
Yea the fema ipaws should alert when bad?
I get them in WA
Heat, smoke and fire
I moreso want it so I can make sure to bring my pager along on days where I might be dispatched to a fire (im a rural volunteer ff)
Nono, I don't have a pocsag pager, two tone straight from my county fire dispatch
I will dig into my local FD pagers š I think they also tone out if so I will try and get a etc/ decoder for alert.txt see if itās valuable
With an rtl?
On my meshing around host, I use trunk-recorder with some rtlsdrs to grab recordings for openmhz, maybe linking into something like that would be handy
I have been out of scanning hobby for a while so donāt use all the new software. I was last in when UPman was around.
The idea I have for grabbing tones would just listen to audio feed
Would then dump the alert to .txt and it would go over mesh
The barrier being primarily is every agency and departments codes could be unique I have never personally sampled the size of this pool of data but the point is to not allow cross talk so
The icad tool comes with some examples so will start there
It will be a boilerplate for people to customize not built in
In theory it would take any audio sample web stream local or output from a rtl
icad_tone_detection? I think it's commonly used with trunk recorder so yeah. Would he Handy for sure but it's probably going to be my backup for my backup for my main source of fire pages.
Trunk recorder will dump directly to MQTT @mental helm
So if you see it there now, you can broker it into mesh now
Mesh-Bot works 1:1 with a node on MQTT
So just configure MQTT on your bot node and bam you added whatever
Or make a mesh-bot-Linux-native-simulator and point it at another whatever in MQTT
If you just want data sharing and the MQTT is fully handled in your mesh it may be the best path
I havenāt touched the code yet but reading it I expect I can take the traffic in from audio
Process with icad tool and have a end user ready to go code
Things like what message to trap and what to send is likely gonna be end user choices
I will goof in a bit and if it flows will have a product quickly
i dont think trunk-recorder decodes tones though, so the only thing trunk recorders MQTT does is pass along if a frequency is in use or not
and other call metrics like amount of transmissions, time of last transmission
There is a proof of concept for this idea in lab etc now
I didnāt have a streamer to check they all lock away
If you have one
Also be interested to hear improvement ideas
Ill try to find time to poke at it here before the end of the weekend, I have to wait for something to come in for my department over my sdr. We're really rural so apart from Monday pager tests, I don't get much activity.
Haha Iām also waiting for a page lotta chatter so I know I got the locals
I was pondering this, and all Uniden scanners will alert ā¼ļøon a tone out as well
You could set that up and use RF monitor as well to alert on rigctl
Or VOX alert
I have new ping logic in place for new 2.7.4 proto which now looks to better handle UDP as well
Not in main yet but if you would like to test for better ping and hop logic in lab
Did anyone notice issues with ping and hops?
New data in ping
Itās too cold for me to demo it in my production so
Imagine itās cool š
Should give you the relay or last hop
Search the id in your mobile to find its name if known
I've been tied up on another project for a solid week, but look forward to messing with this some more
Aloha Friday-- interesting output... N
Nov 14 16:11:31 hawaiimeshmap bash[1984]: 2025-11-14 16:11:31,469 | WARNING | System: API version error update API pip3 install --upgrade meshtastic[cli]``
Did that-- and yet it' persists-- hmm - the bot works... 73
Did you make sure to do it in the context in which the bot runs?
The error will pop on boot when trying to get channel info using new function calls
It wonāt cause a problem for operations just canāt use that function
here's a good one-- LOL -- I get that output from the femtofox as well it's 1969 brah
BTW I run the latest Meshtastic š„„
yeah I tried that- the error persists- will continue to play.
What version is shown?
Meshtastic 2.7.14.x daily--
Update the API
try:
ch_hash_table_raw = node.get_channels_with_hash()
#print(f"System: Device{i} Channel Hash Table: {ch_hash_table_raw}")
except Exception:
logger.warning(f"System: API ver
I update the meshtastic cli-
The work for this is on pause waiting for one more API patch, I didnāt think enough ahead.
where does that script go?
Itās the lines of code which give the error for update
That is the method which the new versions support to grab cached channel hashes so I can route UDP well
once I have a good handle on UDP I will keep working on some game ideas
COOL stuff BRAH
I'm getting to know this stuff and playing with Rapheal kit, sensors etc.
-- time to cook dinner hihi -- going to play with this later.

I play with UDP! wh6eps has been working on a interactive chess game on meshtastic. 
Are there config settings for the fruit machine?
not really should there be?
It can be disabled with the word of the day game which also runs with it
wordOfTheDay = False
It's fine. I just don't want it sending anything out on the public channel that I'm not aware of.
Its intention was to drive on air activity like .. get channel chat content, leaderboard stats and all
All fine and good. For now, I'm taking a more low-profile approach - just announcing the bot daily on the public channel and encouraging folks to give it a try
I was trying to use the custom_scheduler to automate using the send_sysinfo() command but only have it come to my node. I set it up like this:
schedule.every().day.at("00:05").do(lambda: send_sysinfo(0, 1234567890, 1))
where 1234567890 is my node ID. Then I get this error:
Nov 17 00:05:19 bbs bash[832]: 2025-11-17 00:05:19,530 | ERROR | System: Scheduler loop exception: setup_custom_schedules.<locals>.send_sysinfo() takes 2 positional arguments but 3 were given
Am I using it wrong? I thought if we set the channel to 0 then we could use a node ID in the middle argument to send a DM to an individual node. Where am I going wrong?
There is a function on the file locally calling the function in the bot, to keep it more flexible.
So you need to edit the custom files function or use send directly
send_message(sysinfo('', 0, interface, False), channel, 0, interface)
Vs using
def send_sysinfo(channel, interface):
The two positional on send_sysinfo, is passed to send_message
Oh, okay, thanks for explaining that.
@gentle kestrel , I seem to remember that you limited the number of days of weather forcast when we scheduled a "wx" command using the custom scheduler. Right now, I am only getting the forcast for a single day as opposed to 3 days. If I run DM the wx command to the Meshing-around bot, I am getting the full 3 days forcast. is this configurable or is it baked into the code?
(for the scheduled event)?
The basic event is 1 day to keep clutter down but custom you can
Change that days value to what you want
Thank you sirā¦
Unusual time and placement
And thenā¦. Hear NOAA is pressing news today
They are saying these alerts will soon come to and end if we donāt get money
Interesting they are sending alerts to get attention
Imo they learn extortion from up the chain āļøāš„
Anyone else notice or have wild results back from all data points lately
like the government reopening things got worse
Cloudflair asks, are you LLM?
My River flow module is still not working yet the one a town over is working fine. I didn't do anything to mine. Odd
Yea I monitor two rivers and one sometimes is out
outside of some elaborate caching scheme Iām not sure what I can do, and if the networks are dismantled
Look at the recent mess with mt Rainer
The sensors went nuts for the X time this year
And this time itās āfake newsā blaming the public for misrepresenting the data
Meanwhile the data is claimed to be false for ice
Ooof
Gaslit by your own government
Who is also trying to ban LNG delivery to homes who claim itās cold
Why are they trying to ban it?
My cousin lived in Rainier. She had LNG but there was a leak right under the house nobody knew about. It caught on fire in the middle of the night and burned her to death and obviously burned her house down. I wondered if it was related to that. Happened a few years ago..
Yikes, sorry for the loss.
And here I am worried about radon
I heard Australians have so much energy they get it free for a second
I donāt see how that energy issue is different vs LNG
The closest thing I got is the leaderboard
Tell ya the hottest temps
Are you also near Rainier? We randomly have radon here in PA, but every home I've lived in is too drafty to ever detect it.
I am near enough to see it daily
Im able to see baker clearly
I can also tell you imo earthquakes and solar are related
Once in HRCC I was laughed at for this opinion
I mention it because I think I observe things well
This data is interesting, anytime Rainer makes press this same area seem hot
So .. is the sun causing reading failures or real data
I enjoy a good joke but not when I detect a lie
Anyone see the volume of lava this week
Interesting ..
Havenāt heard a lot of hiccups lately pressed a release
Anyone try two player battle ship?
Haven't tried any games, but i did a -nope and then pull and reinstall and riverflow is working again with the old config data. I must have broken something else when I was trying to add all variants of "check in, check-in, checkin" etc.
Also went to gemma3:1b and it's still working surprisingly well on my rpi4 4gb
The airplane vs ground vehicle speed leaderboard is perfect
Awesome! I have caught a few lately myself and been excited to see them. There is a shocking amount of Lora in the air traffic, and not on defaults which is, double shocking. I can only imagine how well it must work up there haha
Meshtastic or just other lorA?
Meshtastic
Oh wow. What settings are you seeing?
AQ on different channels and seeing regular nodes up high going fast sometimes very fast. Itās hard to say itās not a code cheater but⦠I am in arguably a hot spot for experimental aircraft and radio
How do you find this traffic and can you read the messages?
Accidental mostly. donāt know if they know itās going out I did donāt think about others choosing random channels
Like itās a letās go to 73 or the group is on 12. The density is becoming such itās more likely you find random nodes
But then one day on a lab channel I was like wtf
And looked into it and was like woah fun
It appeared to be a student and I assume they may have had a node on them from my snoopin tail and such
If you are in the sky for hours itās a slick way to chat with the ground
As cell phones can be an issue
happy happy š¦ check this out for output...
first one is femtox with new beta 2.7.15, second 2.7.16.nuZ with latest meshing-around
š¤ is this chUnking setting?
DEBUG | System: Bot detected Commands:[{'cmd': 'cmd', 'index': 0}] From: wh6gxz meshtoad isDM:False playing:False DEBUG | System: Message length 285 exceeds maxBuffer200, forcing chunking. DEBUG | System: Splitting #chunks: 2, Total length: 283 INFO | Device:1 Channel:1 Chunker1/2 SendingChannel: Bot CMD?:ping, askai, bbshelp, bbslist, checkin, checkout, dx, ealert, games, hfco> INFO | Device:1 Channel:1 Chunker2/2 SendingChannel: CMD?, mwx, readnews, readrss, riverflow, rlist, satpass, sitrep, solar, sun, sysin> lines 10512-10538/10538 (END)
chunker 2/2 output shows in the logs but not in the browser
yeah that date linger only in the browser- not the apps.
Ahh like a packet didnāt make it?
The 2 second delay coming in?
Or is UDP canceling something
When it happens to me itās that like 90% of the time and lost packet the other
Have had very good rate of success on latest code, I use a tdek for my most rigid client
turning off udp broadcast didn't work š¤ I have a number of meshtastic nodes on udp. it's been working for the most part -- been playing with that udup tool hihi the latest meshtasticd alpha the browser is buggy - crashing @ times
might bump up these values if lower
`responseDelay = 2.2
delay in seconds for splits in messages to avoid message collision /throttling
splitDelay = 2.5`
oh yeah that worked... splitdelay was 0.0
any reason you use docker compose the specific containers instead of docker compose up -d?
https://github.com/SpudGunMan/meshing-around/blob/main/script/docker/README.md
other than to exclude the test and debug containers
Keeping it open for people to use?
I dev in raspberry so donāt use docker much these days
Not knowing what people want to run I just didnāt have a better idea yet
gotcha. the up -d will create the networks and run every container in the compose file. was just wondering if there was some nuance, but it seems like its more of a personal style/workflow choice. thanks!
Ok, but hypothetically if one wanted to buy you a coffee, or a jug of gravy or whatever it is that potatoes drink, how would one do that?
I think people like you @candid quarry @sand crater have given a lot to this community. By way of giving your selfās. (Let alone countless others)
i have a configuration question. since im running this on a linux laptop with a Seeed Xiao NRF52 meshkit plugged in via USB, is there any way i could use it through the network with my meshtastic app/webclient?
Yea just use TCP to connect to the node vs serial. The API client eats up the pipe however and having another web client isnāt gonna fly well. Using the #1421172939751166062 as a local console of sorts. Would be the best way to dual use the node.
Another avenue is to use UDP by segmenting the API applications into venv:meshD instances
MQTT being another but itās kinda like UDP
i didnt think UDP was available on NRF nodes
Oh dang yes some canāt, if it can MQTT it might be all itās got.
I have some rough drafts to meld a firefly like idea but it will also likely be a ā2.0ā rewrite type thing.
yeah some web-ui thing to interact with the node would be really handy in the future. maybe having all the options configurable in the UI too. would be neat to just plug the node into my unraid server, run something on docker and pass the USB through to it or something
You can do that, just using the same same node for chatting is also complicated, more nodes!! š
It works decently, but you can use Mesh-monitor's "Virtual Node Server" to pipe Meshing-around through
Web application for monitoring Meshtastic nodes over IP
I ran it that way for a while.
i hope you dont mind me vibe coding but im working on adding MCP support (just because ive been wanting to play around with it since i got it working) along with a full web-ui that lets me monitor the node and data.
i am but a dumb cranberry farmer who took C# in highschool 15+ years ago on a whim and bounced so hard off it that i had to cheat my way through to a passing grade
heres a WIP of the webui.
MCP is basically for me to hook into home-assistant so i can ask my voice assistant if its seen anything new on the mesh network
Fun, there is a bit of a web dashboard currently in the etc/ folder
yeah i saw that but i wasnt sure how to actually use it.
heres my repo if you want to poke at it and check it out, im not too sure i should try to submit a pull request as i only have a 7 day trial for cursor
the AI added to the config.ini to enable or disable everything i've added too, and updated the documentation for my features
It was also a vibe in by a contributor mostly a dashboard, ideally moving to a more interactive format but combining with more focused projects like prior chat had was kinda ābetterā
on the road today and sort of on a holiday fiesta but will for sure check it out!
no rush. If you think you can manage it after my trial is up just lemme know and i'll submit a PR. might try to poke around at it some more and add some other ideas i had in the meantime
btw cursor is, fucking amazing. i literally know the bare minimum when it comes to coding and i managed to do this by just talking to it. I know vibe coding gets a bad rap but its a game changer for people like me with tons of ideas and no way to do anything about it except for bug people who know how
like, i added in a map just by asking it to add a tab for mapping nodes. it'll even show trails for traveling nodes.
then i straight up asked it "what is it missing" and it listed off 10 things to add, so i told it to do it and bingo
Nice! Indeed I enjoy use of it as well
its vibe coded so expect issues, im rapidly adding stuff too right now, like i just had an idea for haveing a button to update the bot from the web-ui
See if you can hack a tab into the existing html
I haven't had the time yet, but I was going to change the high altitude report to use the nodeid or long name instead of the node number. Realized you can't search or filter by node number in the android app, so it's hard to quickly see which node it's alerting for.
It will do best effort, like dns to reverse but if a node is flying high they likely donāt have time to get a full node info down to you
If all you get is a position packet for example itās all you got
Put in a request for search and filter by node id and hex id I find it crazy handy on iPhone ping for example will only give the byte and you can search that (I am not reversing the relay currently)
Also the bot can Whois a full ID
Didn't even think of the whois and the reasons that the number could be beneficial. Nice
I got my webui working again if you want to poke at it and tell me if it's super broken somewhere not obvious
so while i was indulging on intoxicating hemp last night, i had an idea to blow my project up and restart, by taking your project and the official meshtastic webui project and having AI try to combine them into an unofficial webui with plugin support.
cause i just wasnt happy with how the AI designed the webui before
would be something you run in docker and either connect to a node via IP or USB
and about an hour of time later you can see what i've come up with in #web-client, unfortantly i'll have to have the AI port modules from python to typescript but it doesnt seem to be having any problem
Hey, I'm having trouble with DMing meshbot from a device running newer firmware - logs are showing that the replies are being rejected as "legacy DMs"
Dec 18 15:42:04 luckfox meshtasticd[1325]: WARN | 23:42:03 419 [Router] Rejecting legacy DM
Dec 18 15:42:04 luckfox meshtasticd[1325]: WARN | 23:42:03 419 [Router] Rejecting legacy DM
Dec 18 15:42:07 luckfox meshtasticd[1325]: WARN | 23:42:05 421 [Router] Rejecting legacy DM
Dec 18 15:42:07 luckfox meshtasticd[1325]: WARN | 23:42:05 421 [Router] Rejecting legacy DM```
Meshtastic firmware issue, the DM keys are a mess try deleting node db all over.
aloha mesh people and happy happy -- anyone have any ideas why i'm getting these alerts all of the sudden?
no changes to the .ini file. Thoughts?
I get this outpu every so many min's...
AI's answer
This alert typically issued by the national meteorological service, such as the Deutscher Wetterdienst (DWD) in Germany, to inform the public about a significant weather hazard. The term "severe frost" usually indicates that very low temperatures, potentially falling below a certain threshold
yeah it's ealert that decided to speak german hihi-- i disabled the broadcast for now-- 
hey happy happy spud! so it's here i need to modify that output? i've gone into the .ini file to make changes to no avail
The code will only process the NINA-Germany alerts if the entry is =True in the .ini
If false it ācanātā process that code
Same goes for any other uses
Been turning this on and off-
the .ini file
ignoreeasenable = False ignoreeaswords = test wxalertbroadcastch = 1 enableextralocationwx = True ealertbroadcastenabled = True ealertbroadcastch = 1 ignorefemaenable = True ignorefemawords = test,exercise mysame = 015001,015003 volcanoalertbroadcastenabled = True volcanoalertbroadcastch = 1 ignoreusgsenable = False ignoreusgswords = test enabledealerts = True
Very last line is enable DE
valert show nothing now odd- love to play with meshing-around 
I last tested all .gov API about a week ago they were ok, I think you had another eruption? š I saw lava spewing in the last few hours (also talk of a late hit of cosmic blasting last night!)
I donāt expect issues but given the holiday and such.
Yeah it was working this am- pele has been busy- fixed the NINA-Germany alerts. Mahalo BRO!
Might check the Valert in a while, I will lab up a Hawaii bot when I can. But like most things NOAA Hawaii, Iām leaning on saying this is a federal holiday till āMondayā and possibly just Hawaiian time hihi
I usally get some output... played with the .ini file a bit today
No alerts means for sure it had none / double check the debug if it didnāt say parsing error itās the upstream data
Dec 25 13:20:20 wh6gxz-fox python3[1032]: 2025-12-25 13:20:20,119 | DEBUG | System: Bot detected Commands:[{'cmd': 'valert', 'index': 0}]> Dec 25 13:20:21 wh6gxz-fox python3[1032]: 2025-12-25 13:20:21,516 | DEBUG | Location:Error fetching volcano data from USGS Dec 25 13:20:21 wh6gxz-fox python3[1032]: 2025-12-25 13:20:21,520 | INFO | Device:1 Channel:1 SendingChannel: No alerts found. Dec 25 13:20:22 wh6gxz-fox python3[1032]: 2025-12-25 13:20:22,336 | DEBUG | Location:Error fetching volcano data from USGS
Oh yeah...
like totaly dude hihi
Yea just checked same here anymore it could be to keep the data š
I saw some radar data for no fly zones in the last 24 look like fires once seen your way
Spicy time for lava flowing
In a ātwo day shutdownā
yeah roger dat bro! the last few eruptions sound like a jet plan taking off from my qth- a friend heard pele at his hale 15mi down the mountain the fountains have been spectacular. The park gets crazy busy and people plan for the the ongoing intervals to catch the flow- good times. i hope u enjoy and geek out on this day 
Woah thatās cool! 73 
VALERT is back -- MAHALO USGS? spud u da man...
seems to be outputing about q15min... it's on a private channel. going to just let roll...
Live view of the eruption in Halemaʻumaʻu, from the northwest rim of the caldera [V1cam]. This camera is a pan-tilt-zoom model and the view may change depending on activity. The V1cam is an autofocusing pan-tilt-zoom model. The camera automatically focuses its view, and USGS Hawaiian Volcano Observatory staff may change the view angle or zoom ...
valert works awesome -- lately if been outputing about q15-- any way to adjust that? 73
Q15?
sorry nurse talk every 15min... hihi
Roger. I just wrote it in blind. Git pull will grab it.
A undocumented [location] key defaults to alertDuration=20 if you add this to your location ini section. You can set a value >10 and it will check the API on that number xx:20 xx:40 xx:00 for example.
So alertDuration=15 on the 15,30,45,00
mahalo BRO!
Not sure if a new year new code didnāt look deep but I saw NOAA WX alerts fail for index error. For an alert locally.
Got a patch in to handle it.
The warning was for king tides, the chandler wobble, or the two moons? Either way the second is now stuck in cesium
shidd fard coom
https://shidfardcoom.com/
I wondered what that was and used it here to find I canāt remove it so ..
I've read this 5 times... What?
The tides locally are causing havoc in alerts daily my way - the alerts had no location data they just return on lat,long
Had never seen NOAA alerts without a āareaā in my feed before and didnāt account for it in code. Confounding tides.
Ohhh ok
@gentle kestrel I have a tool that I've been developing with a few ai dudes smarter than me -- i want to have some kind-a plug-in installer or something add value to it. as well i would be truly honered if you gave it a try -- i'm looking to invest some time into this -- it's called meshforg -- anyone reading this I need TESTERS! hihi with aloha https://github.com/Nursedude/meshforge
@heady raven I will check it out! Kinda took(on) a holiday hiamoe iki of sorts
awesome bro- take it slow -- ;=)
Hey spud do you mind if I try and integrate some stuff of yours into meshforge š
I got nothing special just some api calls
All the ideas I put forth are to be used however in learning meshtastic and enjoying the hobby and advancement of enjoyment, donāt break laws
Some of the ideas I had hoped make it into a concept of wider use, Python is so wonderful for prototyping
I still got more, just busy fall enjoying a season off, just got billed for a new year of copilot costs went up, I also saw major updates after feedback so Iām excited to jump in again when it feels groovy
I really love working and playing with meshing- around š¤ canāt wait to see what you come up with next?
What do the "78" and "B9" mean?
The last byte of the rebroadcasting node.
Search it in your mobile to guess its likely node
I will get a patch to a few items and PR in soon just got pulled away on NOAA related matters
I didnāt want to reveal the local bot DB name in the ping so itās just the byte no lookup
Ohh ok.
I seem to be running into a problem with my fresh install on my Femtofox. The bot operates fine all day long, then overnight the CPU gets taxed trying to run etc/report_generator5.py which seems to go in a endless loop with a new 4 digit PID each time.
Stop the service for now to save wasted time, you could also try deleting all logs it might have a bad one? Itās a very basic logic run
Thanks. I'll give that a try.
Here is the output from running journalctl on that offending PID. Does it indicate something I've done incorrectly?
Jan 13 06:32:37 F0X0 python3[1364]: Checking log file: /opt/meshing-around/logs/meshbot.log
Jan 13 06:32:38 F0X0 python3[1364]: Consumed 54 lines from /opt/meshing-around/logs/meshbot.log
Jan 13 06:32:38 F0X0 python3[1364]: Node1: N/A N/A N/A
Jan 13 06:32:38 F0X0 python3[1364]: Node2: N/A N/A N/A
Jan 13 06:32:42 F0X0 python3[1364]: Warning issue reading database file: [Errno 2] No such file or directory: '/opt/meshing-around/data/lemonstand_hs.pkl'
Jan 13 06:32:42 F0X0 python3[1364]: Warning issue reading database file: [Errno 2] No such file or directory: '/opt/meshing-around/data/blackjack_hs.pkl'
Jan 13 06:32:42 F0X0 python3[1364]: Warning issue reading database file: [Errno 2] No such file or directory: '/opt/meshing-around/data/videopoker_hs.pkl'
Jan 13 06:32:42 F0X0 python3[1364]: Warning issue reading database file: [Errno 2] No such file or directory: '/opt/meshing-around/data/mmind_hs.pkl'
Jan 13 06:32:42 F0X0 python3[1364]: Warning issue reading database file: [Errno 2] No such file or directory: '/opt/meshing-around/data/golfsim_hs.pkl'
Jan 13 06:32:42 F0X0 python3[1364]: Warning issue reading database file: [Errno 2] No such file or directory: '/opt/meshing-around/data/qrz.db'
Jan 13 06:32:42 F0X0 python3[1364]: MeshBot (BBS) Web Dashboard Report Generator
Jan 13 06:32:42 F0X0 python3[1364]: Main dashboard: file:///opt/meshing-around/etc/www/index.html
Jan 13 06:32:42 F0X0 python3[1364]: Existing index.html backed up to /opt/meshing-around/etc/www/index_backup_2026-01-13.html
Jan 13 06:32:42 F0X0 python3[1364]: Main dashboard written to /opt/meshing-around/etc/www/index.html
Jan 13 06:32:42 F0X0 python3[1364]: HTML reports generated for 2026-01-13 in /opt/meshing-around/etc/www
Itās messy logs but all looks ok @deep igloo
Says reports generated all the lack of high score is just boring
Should cut it to save log space actually
I know right... No high scores because I used a cheap 12yo MicroSD card... which failed me. Hence the fresh install.
Dude tell me about it if you run the update.sh it will .tar.gz them for ya I lost some killer scores and instantly added that haha
Knowledge for the future... Thanks. My nearly $700k dopewars score (poof, gone)
Authorized Raspberry Pi Reseller in the US. The largest selection of Raspberry Pi boards, kits, unique accessories. PiShop.us - your one-stop American store with fast shipping!
I love this guy - always some cool insights - he visits meshtastic land -
Raspberry Pi's back with a new AI HAT.
This time it adds on 8 GB of RAM and the Hailo 10H for $20 over the original.
Raspberry Pi provided the AI HAT+ 2 that I tested in this video. They did not pay for the video nor have any say in the video's contents. See my review sample policy here: https://github.com/geerlingguy/youtube?tab=readme-ov-fil...
hey spud i was thinking of having a meshing-around stand alone app- all your stuff in a say TUI - GTK work flow - the app would pull updates from your repo -- all files configurable etc throught the app - no need for a meshtastic web browser... kinda standalone app for meshing-around - check out meshforge repo on dev- stuff
yea, tinker away.
I was tinkering with a firefly shell and possibly pygame in web but itās a large bite so havenāt pulled the trigger yet on that path.
Mostly just more games is all I got on my short list for the moment
Ok, cool
- I'll work on something using the knowledge gained from meshforge -- I'll let you know whats up - your input is important to me. I'm just nursedude making stuff I want to play with. š¤
the tui i went for after breaking a few env was the raspi-config look it's been pretty stable - it's the launcher.py its the front door say for all th ui variables
Roger Roger, yea one thing to note is the current image for femtofox is in concrete so my ability to move a bunch is limited there (why it has a femtofox auto config)
I basically saw it work here and heard from user community it was working and just havenāt touched it .. everyone has a personal run method I feel
App frontend for the games would be good.
Run them over the mesh, play them like they're meant to be played
Yea, so far just have the tic-tac-toe out there for a light demo
I donāt have those spec5 developer hours
One firmware item that I hit a wall on was the delay of 2seconds for new packets
Iām debating a patch / to allow another game idea
Tbd, fire has irons flying all over atm
I have configured meshing-around on a FemtoFox. The MT side is sending/receiving messages. However, when a node direct messages cmd, the FemtoFox doesnāt reply. Same for other commands: whoami, etc.
Any ideas on what I havenāt configured correctly?
Do you see the message on the logs? Will ping anywhere pong?
Validate itās not a DM key issue, I really liked when firmware favorited the dm it handled the memory but may have overloaded something to crash not sure.
Itās a tricky game if itās a BUSY mesh or if people are goofin
I do not see the DMs in the /meshing-around/logs/
Ping will not pong. I have it set to respond to DM only.
The FemtoFox can exchange DM with my other devices using the Meshtastic client
It feels like Meshing is not getting the handoff from meshtasticd but I donāt know where the settings are that link them.
It wonāt boot without a link and by default will connect to localhost. Do you have other stuff connected? It is exclusive access to the hardware
I confirmed that nothing else within the stock FemtoFox UI is conflicting.
Check if system logs show running?
It was very cookie cutter on the 1.0
But if you have two things running it might be a problem
Validate its /opt/meshing-around/logs as well
If you see it running and no DM something is funky with meshtastic /femto image
There is a permissions script in opt/meshing-around/etc you can try and reset
Was this a new build?
This is the as-shipped FemtoFox. Iāve updated MT using the FemtoFox UI several times this year. This was my first time enabling Meshing.
CHMOD 755 did not help, but CHMOD 777 on config.ini allowed everything to start. Itās sending messages now.
Thanks for your quick responses! Looking forward to running this through its paces.
The documentation on GitHub is really good. I know how much effort that is. Just wanted to say thanks.
spud got the awesome sh*&T env. I've had Claude code MAX analyse the meshing-around domain -- sweet
i struggle with other issues re: meshtastic, rns, and other integration stuff
@gentle kestrel I had to run that script after a fresh install on my Femtofox
What's the best way to clear the leaderboard?
leaderboard reset i think
I did just put on a community patch for better save on all files
The will help with specifically docker images or the likes
There is another PR on deck for maps but I may wanna add to it slacking still for winter
I cant seem to get this to work
Remove - (min/interval)
From value weather line
That should do it?
@gentle kestrel thanks for making this project- quick question is there a way to check/poll for weather alerts more often? Which part of the configuration files should I look for it?
I am trying to make Meshing Around stop sending High Altitude Alert!
I added highFlyingAlert = False to the config.ini and restarted the mesh_bot service. That did not stop the alerts.
Any suggestions?
Set it to a fake channel maybe? I set my high altitude and weather alerts to their own channel, then have a scheduled message every evening with the channel information for anybody that wants the alerts.
That would work I will take a look into it further when I can
#1255754522303922217 message
See that
Its undocumented I should document it
not tested yet -- please test! this is spuds stuff wrapped in a tui and web interface -- also supposed to run without an antenna - standalone - i'm going to test it first on a pi2w behind a arden router - blah blah --

WOW!
https://github.com/Nursedude/meshing_around_meshforge
I've been doing some work on this - have a pi2w for a headless no rf implementation - warning it's not stable - my focus on stability is currently with meshforge - map dev is in alpha - trying not to break main hihi - check out the rf and psk stuff..
the tool uses your repo to stay up to date -- i have an updater in the raspi-style launcher tool - supposed to run it's own web-client - a big ask hihi
I will check it out is this the same as the prior project just more refined?
I gotta relocate my work stuff entirely so it will be a month or two still, construction etc I donāt have a desk to work on really took the winter off basically. So my lab and all devkit is basically offline till at minimal a desk and flooring show up hihi
Not going anywhere having fun with meshforge now it has maps. š¤š
Here's a weird one. I've set to a separate channel for planes and weather alerts. It worked perfectly, then randomly today the planes started coming on LongFast again. No idea why. I haven't logged into the pi since i made that change a week and a half ago. Not mad, more amused
@gentle kestrel Aloha Spud i have an idea or challange for you. recently i have been using PulsePoint cool app https://web.pulsepoint.org/ is there a way for your bot to capture this data and send out just the active alerts. it would be a really cool project
Respond for Web offers a customizable, browser-based view of PulsePoint incident feeds.
That's a cool idea!
That High Altitude Detected alert is baffling. I canāt seem to find a way to stop it.
Whatās the logs say it did
Highflying ini value?
Haven't logged into it. Been preoccupied with complexities of life
It would be my first question of did it burp? Itās kinda set on rails and all
Is there a way to have the bot monitor a channel for specific strings and then send a DM to that node with a phrase? I already have it set to only respond to "ping" and "test" through DMs but I'd like to configure it so that if it hears "ping" or "test" on our Main channel it will DM something like "Please move all testing to channel TEST." But if they're already sending "ping" or "test" on channel TEST, then it responds with the pong DM response. Is that doable in some way already or is it too much?
Itās possible but itās a public mesh
The QRZ hello feature might be good to welcome people and direct to a website for the use
If the bot is responding to ping and test with not useful data whatās the point and all
Wasted airtime to say go over here hope you get this message
Can also send scheduled beacons to say āwelcome to mesh see webā
Thanks
Howdy! What all does dataPersistence save and how much space does it need?
It saves the high scores and leaderboard uses very little
Disable for any concerns about write media
It was added for extra protection from crash and for virtual hosts
Got some bugs squished
What's enhanced mapping going to involve?
It went above the CSV output and added a DB for more advanced āhow far from the positionā you can save and recall positions of public and private nature.
Itās just a git pull away
is anyone else getting a "No Data for your Region" response when folks are sending the "rlist" command?
It was the other day but also fixed it you on latest?
I did notice cloudflair is being extra stingy with get
Confirmed working in my lab but I did visit the page to say human like a week ago
Visit this from the subnet behind the firewall https://www.repeaterbook.com/repeaters
Thanks, I will pull down the latest build...
Bingo...that worked.. Thanks @gentle kestrel
This is cool, sorry for late response been ooo kinda.
I donāt see a API link today but will keep an eye on this!
Accessed on my phone on WiFi, clicked I'm a human, still getting that error on the bbs though
Did you update to latest?
git pull
Aloha Spud it is a nice system sadly it was operational for 6 years paid for from our local hospital however the contract ended so we lost access itās a paid service so Iām not seeing an API either I believe broadcast is working on something similar with the CALLS system and āIncident Source & Project Radiant Glow ā AI-Powered Incident Intelligenceā so will see thanks for looking into it
@gentle kestrel I'm getting close to having a drop-in UDP virtual node framework. No need for all that firmware nonsense cluttering up our bots. Very untested aside from broadcast/DM functionality.
https://github.com/pdxlocations/vnode
I been seen I lot of movement your way š
I will need to test it more - I got node info but havenāt got autoresponder example working yet - I do have a fork out there š
May need to wait for the db to populate if dm.
I have an ai bot using it on our mesh.
Havenāt really tested anything outside of that
I booted it in software only just stopping the meshD on my lab - saw a node info but no text yet will investigate more in a bit.
Iām running latest alpha
Might have spilled the beans too soon. I havenāt pip installed it on anything
Listener seemed to be fine but auto responder didnāt blip I havenāt looked at code there yet
cloned on a fresh pi, ran auto-responder, didn't work. hit exchange nodeinfo on the real node, now it works.
gotta let the meshdb populate i think.
I went for a walk I noticed it ran on a cycle
Yea same, I see nodeinfo but no text / listen works fine / autoresponder isnāt dooin much
hrm.
Yea - I used up all my energy loading it up I will poke more in a while - my first guess is PKI but not sure yet. Setup is 3 nodes(Ubuntuās) in a subnet made sure meshD was stopped if I enable legacy methods itās all flowing ok
meshd shouldn't have anything to do with it. In my case, i need my linux meshd/radio running to route the UDP to lora. I've only tested with a t1000e sending messages to nebra node with vnode running on my mac (and recently a pi) all on same network.
Adds to the noise levels, could just be the soft lab, I will look more into it // meshD worked fully in UDP same setup
might be something in my deduplication logic, been meaning to dive into that. My UDP stream is pretty busy with all the meshd bots.
I just found itās a DM responder only
The issue now being keys are not moving (same issue seen with OEM)
Iāve seen 4 hops within my lab. Maybe check yours.
Itās like itās being duped out I saw this same issue on build up of a new lab
Same thing @stone lance had to reboot nodes or - force a full send 2x to DM
@cold jungle works now after all keys are around
Tossed a log note on the fork
If you run femtofox a image builder for meshbot is forked here
https://github.com/SpudGunMan/mPWRD-OS-dev-kit
if there is need I can look into building resources for image download // pricing changes this month so not sure if I can build for free
Pricing ?
GitHub Hosting for runtime and storage it was free for open source but I saw a memo about 0.2 cents per 500mb
For the moment itās DIY no nice download on my fork
AH! Github..
I also fixed that link to the proper fork š“
the "rangetest" is greyed out ... yesterday i got i running for a moment ... but i didn't get it on aggain ...
Range test in the firmware is one way but it can be hard to manage in real life situations.
If you got a pi or docker or something that can easily run Python projects give this a try
Plug a node into it and boop
You can say āpingā or ātestā on air and it will respond with data
It will help visualize what you can accomplish
No theory just raw data and meshing about
The map ? has like a range finder 2.0
Yes sounds really nice, but i don't have the time for more "installation" i just would like to get the rangetest running when i am going to work and make some ways around over the day without doing anything. As i understand it will save a csv which i could easily import on a map
no for no reason i could "activate" it on the mobile version, but i ant deactivate it aggain in the menu ... i hate stuff like this
Run it again? Iām quite sure itās not globally down
I fixed a minor issue today in fact so itās there
Fixed an issue if you run multiple versions of Python same box