#US - Utah
1 messages · Page 21 of 1
Thanks
i was lucky only broken stuff on mine have been a dead emmc.
@real cedar you make it down?
We are almost safe
On Bates canyon road
We will make it just fine
My smart watch reports very abnormal calorie and health stat numbers 😂
@prisma saddle
to the bike or truck yet?
Cars a mile away that will be easy compared to the rest of carrying someone else
Chugga chugga chugga.. can't stop a freight train 😂
I'm glad you are (nearly) safe, I'm sorry your day took a turn to the bad side of "adventurous"
We are in the car
And thx
oh good. he need the hospital? or just sore?
glad you got there. your going to be dead tomorrow. @real cedar
sorry that all happened.
and like i said when you decide to go back up. if it can be a sat. i’ll go up with ya.
the wife gave me “permission”. lol.
Hospital for a check up tomorrow tonight yes icing and going to bed
Honestly the brush is extremely overgrown atm. It might be best if we can get an escort up
Don't worry about it. It was a fun experience ... Sum total I think the weight added to my shoulders was my buddy + his bag + plus my tool bag + all the water and stuff. Got down though
I'm home
I will likely just be sore haha
im going to try to look into that this week get some sort of contact that can offer that.
most of the land looks riotinto
Yeah the road would be ideal. Hell id bike up it if they let me. I can bike in the dark
any word on your buddy?
Won't know till 1 pm today
thank you for letting us know ❤️
@obsidian storm @prisma saddle in frequency we should put frequency slot 51 not 915
That might be why people keep selecting 20
ya i saw that.. was a bit confused myself. will take care of it
Thx. Idk if that was my mistake or just a Mistake in the many updates we have had
who knows. 🙂
im on it
that whole quick setup needs to be changed around.. ill see what i can do. its very confusing. might just give the menu paths depending on the phone OS being used.
Go ahead and give it a whack
going to be teaching from it on wed night.. so will have something done by then
I'm seeing a new node with the name utahmesh.net pointing to a different discord
on 20 or 51?
It’s just a Discord group focused more on the mesh in Utah instead of a channel for Utah in a larger Discord (this one). It’s not very active.
Do we have repeaters on 51? Getting trace routes with no hops one way and two unknown repeaters the other way.
we shouldn't.. but that does not mean someone has not set one up.
Yeah copy the routers, I’m talking repeaters
yup.. i think out of date routers can do this or have a hand in it too if i recall. could be wrong.
What channels do you have configured?
A private and LongFast 51
Sometimes that'll show up if a node doesn't have a shared channel with you
Hey HAMS
I've got an amazing idea. I know @reef ridge is putting up FPR but as a great option do y'all think you can contact some folks for OGDEN radio towers and put a BBS node up there. It has clear LOS to mt Harrison
http://ogdenarc.org/hardware - MO.html
And would spare me having to climb this
I was actually thinking about this one, I was going to go to one of the OARC meetings and ask about it.
I thought Lewis Peak might be a good spot too.
I would just go place it on that tower that Batman was using but make sure you're lightning arresting is top notch. In the meantime, yeah I would highly recommend going to one of those meetings and seeing that they'll allow us to run a nebra up there just like lake mountain
Doesn’t have a shared channel, is on slot 51, and is setup as a role to retransmit.
Oh so it won't reply via DM?
Not in my single dm test.
Was on a node with only an encrypted network, saved a contact from the default longfast 51, sent the dm, no work. Think the error note was something about no shared key.
Got ya. Is that on purpose. Who owns BBB
Was hoping it lit up everyone on the default or through dm.
Thought that beautiful bbs was @obsidian storm , but @manic fractal claimed one too. 🤷🏽♀️
No, I meant something supplemental specifically for central Ogden. FPR is still getting the fox.
Mount ogden might be better than Benny Lommy.
Ben lommy is a novelty
My ogden is a good idea
But that makes FPR completely redundant
I don't think so. I couldn't hit FPR from the middle of Ogden. I had to go further east to make contact.
I have one running up north. There was one down south too maybe one RaxAux’s node? I thought we were running the same one (meshing-around) and the cool thing about it is that different instances running in the same mesh can sync up and transfer messages back and forth through the mesh.
What I mean is those routers would be far too closely spaced. I'd take down FPR in favor of mt ogden
If we got mt ogden
That's what I mean
That wouldn't leave a dead zone in the South Ogden/Layton area?
What program do you guys use to simulate radio propagation?
I don't control any bbs at all
No it wouldnt
Mt ogden has excellent LOS south of the base
With fpr i could 1 hop bountiful starting at 400n and I15 in mariott slaterville. Probably similar results plus 20’ up a tower to move the coverage towards the mountain.
Compute summits visible from any location worldwide
Yeah, but real world traceing hones it in.
No BBS on lake mountain yet?
I'm not generally worried about number of hops to a router in the valley. Like the routers main job is to send things as far across the network as possible not to provide perfect coverage.. better coverage in the cities can be accomplished by well positioned solar battery powered clients
They are sort of your large air routes so to speak. Collect all the jazz in the bottom. Route it far south north east and West of where one is to the closest router.. send it back down into the noise
Agree to disagree i suppose. Router is high up and prioritized to kill dem hops.
Well, the problem with thinking about it from the perspective of minimizing hops is that it's a mesh Network and you can't really control that as well as you think
Sometimes it will just take hops through the city and not take the router
That's kind of how a mesh Network works. It's not an infrastructure mode Network so to speak it's very ad hoc
Does the router help? Absolutely but guarantee you minimum number of hops absolutely not
The developers have actually said this quite a few times. In fact, I remember having a conversation with them in general about that
🤷🏼
I'm trying to think of a way to explain why it doesn't guarantee you that in a better way. I wish I could have a good way to do it for the docs, but let's say you have a router that has itself first in the contention window. Well if you miss its rebroadcast then you're left with either the router lates in the area or the client rebroadcast to receive the message. Does that make sense? So it's entirely possible you will just miss a router's message even if you have LOS. And that's the case where you'll see it. Hop around and do other things. Since every other client in the area is going to be rebroadcasting anyways just after the router.
I'm not saying the routers don't help but I have very good data on rufio's node and basically even though he has direct line of sight to aur, he doesn't always go through aur.. 😂 much to my frustration
This actually opens the conversation that I'm genuinely interested in seeing. If every router had synchronized time, would the network be more reliable? Would having reliable timekeeping and timestamps across the entire network make it better able to coordinate contention windows? I really wouldn't know this isn't the part of the code I'm familiar with
Because I know there's RTC devices available for the majority of our compatible radios and there's GPS as well. To me it begs the question, is it just timer-based or does it actually use the date and time as part of the calculation? I don't know
Don't all nodes send out a time stamp and the receiving node adjusts accordingly so the average time is agreed upon across the mesh?
With a preference for gps time, of course.
I'm pretty sure it's something along these lines, but I've never really looked into it as deeply until well now
I dont think it cares about time. Doesnt it compare the packet id?
It has to care about time, how would it calculate contention windows otherwise?
Discover the Meshtastic Mesh Broadcast Algorithm: an effective routing protocol designed for off-grid communication using LoRa technology.
Yeah they would have to be some means of time keeping even if it was just a timer and not accurate date and time
Not it reading that link.
Oh okay so it does care about time
It's using csma so it literally has to
And that's for the unreliable hop. The reliable one I'm sure builds on that
Not to mention being able to ascertain how much longer to wait in the contention window when the channel goes idle and everything wants to rebroadcast
I haven't looked into it very far, but if it is useful I can be a relay point.
D801 also had one once upon a time. unsure if it's still active
So relative time. Not exact time.
It is but we need to do some freaking Network maintenance there before I even put it on a permanent Port
No, it's unclear as to which the docs there don't actually say which
I'm guessing we would have to read the code in order to ascertain the answer to what I was looking for
I'm guessing relative would probably be the easiest implementation so I would lean towards yes, I just don't know for sure
Maybe I'll read the relevant code later. Right now I need to go make a ground beef bowl. I'm fucking hungry 🤣
Seems to be averaged relative time, but it's weighted for a preference for GPS so the timestamp on messages doesn't drift to far from the actual time.
Yeah I'm seeing that. So as long as there's some GPS discipline, devices or GPS disciplined routers on the network, the network should be able to maintain relatively low drift
This is making me want to go install GPS on one of the routers, although I'm really not feeling up for a climb right now 😂
Now I'm thinking I should have gotten a GPS module for FPR.
I guess if we have enough nodes with GPS it shouldn't matter too much.
Yeah it shouldn't matter much also there's a non zero chance we are gonna lose FPR to lightning again so don't go too overboard 🤣
Late to the game I know, but I brought an extra antenna mount to use as a step to reach the top of the pole on NP. It was aluminum and bent a little but worked fine. A steel one would be ideal. Tiny and light weight. A strap to tether yourself to the pole and lean against would be even better.
https://www.amazon.com/UngSung-Stainless-Cellular-Accepted-Connector/dp/B08DKLLQ54
What a good idea! I was looking at lineman gear, but that's similar and way cheaper. 😂
Don't tell me that, I'm loading it up with lightning arrestors, hopes, and dreams. 😭
Also noticed, turn on “unmessable” through the app, node wont populate the dm list. So ~cant dm an unmessagable mode.
So what I would suggest is south facing solar panel but west facing box and vertical antenna below the maximum height of the tower. Let lightning strike the tower not the antenna haha
Also yes @ember torrent is right you Must set the femtofox to be messageable if your going to run a BBS on it. Get with @umbral ice on BBS configs. Hopefully with any luck both your 1 watt nodes should see one another. 120 miles from each other
I was going to mount the antenna on the bottom of the case to maximize water resistance and hopefully make it less tantalizing to lightning.
That would do more harm than good unless you have a true dipole and even then you'll be pointing the upward lobes down which won't help its range as much you.. want it straight up and down
In fact, it doesn't matter which orientation up or downward that you put it in for lightning protection because lightning will still induce across the wire since it's parallel anyways
What's essential is that you have a lightning arrester on it and you're properly grounded. As in, I would stretch a ground wire along the entire length of the tower down to the ground or wherever anything else is grounded.
The antenna would still be vertical, it's just be on the bottom of the case instead of the top. Doesn't the donut come from the center of the antenna and it shouldn't matter if it's pointed up or down?
There's a reason why you see all antennas pointed straight up and down with the feet on the bottom. It's because of the ground plane. Now you're going to make the ground plane opposite to the reference plane which is going to point the lobes downwards and thus would destroy any chance of fpr seeing Mount Harrison
I wish I had a good graphic for this
A dipole looks like this, no?
An ideal dipole does
But reality is the donut is shifted upwards instead of downwards and not uniform
It's a good approximation but it's also a misrepresentation. There is a beam width in every dipole but because the ground reference plane is below, there is an upward shift. Simulated versus real is always different
Okay, the top it is then. Solar panel near the top, surge protector on the power line from the panel, enclosure further down on the post with an arrested antenna on the top.
OK.
You're alive.
What was your trail you took to POTM?
I want to go add a gpio labs filter to it.
the trail I took last time was miserable.
I haven't actually read up on the BBS part of things. This is the one everyone is using, correct?
https://github.com/TheCommsChannel/TC2-BBS-mesh
oh thought it was meshing around.
i’m prob wrong.
I'll use whichever fits into our system best.
#1197577977781821541 message
think that’s it.
I feel like after yesterday me and you have very different ideas on what is miserable
😂
haha.
I’ve tried out two different BBS systems. There’s TC2’s bot that is lighter on the number of functions but seems more robust and SpudGunMan’s bot that has more fun, gee-whiz features but it doesn’t seem to be as reliable but that could entirely be my bad implementation and config. There was a little discussion on which bot to run on the router nodes a while back.
I have the systemd job running the spudgun bot restart every 9 hours
TC2 is more simple but I had fewer problems and higher reliability.
Meshing-around has more “fun” functions but often doesn’t do what I expect it to do. I’m not a coder though and someone else could probably really tune the config file or .py files to make it stand up and dance.
I bet that would solve a lot. I know that I’ll manually reboot my Pi and things run swimmingly at first but then it slowly loses functions over time. Never had that problem with TC2’s bot but it is much less complex and feature dense.
I may just switch if it's that much more stable
You're not wrong. Your version of miserable is my version of "this is akin to running a marathon"
with someone on your back.
🤣 yeah. Well look on the way down I just kept happy thoughts like winter coming soon or how absolutely bangin hot the chick is that I've been talking to and so on. Gotta have something to keep the spirits up while dragging someone down a mtn LOL
In my defense I was already 3 beers deep and you get drunk way faster at altitude
the chick is hot without beer goggles right? :).
Uhh yes but it's a tricky situation. See she's kind of the inlaw to my best friend so I gotta be careful in this triangle 🤣😭
Bash probably knows who my best friend is or at the very least can make a very educated guess LOL
ahh ya. can’t mess that up.
Nobody ever accused me of good decision making so here I am LMAO
Would be fucking hilarious if it worked out all the way though. We would go from being accused and asked if we are brother to actual brothers 😂🤣. We laughed about it over beers once. It's not like he doesn't already know what's up
same girl i was talking to you about months ago?
It was super light-weight and was the first one that I installed and tried. It was easier for a n00b to understand and use. Then I was lured in by the siren song of more functions in meshing-around. There’s definitely some cool stuff in meshing-around but a lot of it isn’t necessary and I don’t have a use case for a lot of its functions (when am I going to need a volcano warning?). The only reason I still have it up is because I’ve been trying to iron out all of the bugs and make it work. Also, it seemed like that was the way people were heading. We could easily run parallel bot networks though and people could just use the one they prefer or can reach. Vote with your pocketbook as it were. They both have comparable messaging systems.
Haha… I have 4 brothers and a friend from high school was like a 5th to a few of us for a while. Then he eventually married my cousin, so actually part of the family now.
ROFL nah. That one was out of the picture once I realized current one was available .. went from back burner to pursuit so fast it wasn't fair
Hah well you know it's how things used to roll. You met people through family and friends
You know it would be cool if that bot and meshing around could actually interface with each other on the BBS level
Good info, thanks! I think I'd prefer something more stable that does its job without headaches, so I guess I'll go with TC2-BBS. Especially because I want it up ASAP and don't want to spend too much time ironing out kinks. 😅
You would be incompatible with what we're building then
We already chose meshing around BBS and that's on Harrison and DC 801
We would have to go rebuild some nodes if you chose that for fpr
It's not like it's not doable. It's just annoying 😂
I'm pretty sure that would mean that malice has to make another trip
Nope, like I said, I want minimal headache, and that doesn't sound like minimal headache!
So Meshing around it is.
😂 fair enough. I was willing to go figure it out if you were adamant but that works for me too
But like I said, I really wish the both of them could actually interoperate. It would be nice
I mean how many different ways can you make a BBS right
Famous last words
It would be definitely be cool to get some collaboration on that front, the BBS thing is neat. I didn't know about it until I got here, the concept runs deep.
uhhh oh I said the magic words
Unfortunately not. Meshtastic hardware generally doesn't have a precise enough clock, and cannot be guaranteed to have a GPS to make up for that lack.
So it uses relative time across the board. I got a weird question then. Why do nodes with GPS have much more regular telemetry updates
as in
air quality and stuff
Just trying to understand what we are seeing is all haha
It doesn't even use relative time. It just uses random delays, weighted by received SNR. Nodes will learn the (approximate) clock time from other nodes that do have an RTC or GPS, but that info isn't used in routing at all.
Re why you're seeing more frequent telemetry updates for GPS nodes - I have no idea sorry. As far as I know there's nothing in the code that should have this effect.
GPS nodes can have more frequent position updates though, if they have smart location enabled.
maybe its just an artitfact haha no worries and huh.. you know I wonder if that could be a port num someone makes. Exchanging accurate time across a network. Wonder if that would be possible. I guess youd have to know the delay of each hop to compensate in an NTP like manner
The mesh isn't a suitable transport for millisecond-accurate time sync unfortunately. It's too variable. NTP logic doesn't work over it.
Best you can do is "the time will be within a few seconds of X"
For most purposes, the latter is actually good enough though 🙂
fair enough. So even when a node is gps disciplined the time is only useful for time stamps on the actual device? like in the logs?
Also on timestamped packets. But yes, that's broadly correct.
Other nodes can also learn the approximate time by observing the timestamp in packets sent from a GPS node.
Ahh so the more gps nodes there are the closer everything gets within a few seconds of X?
Nope. It's not that complex lol - it's just a simple "ooh, that's a timestamp, imma use that". Even having just one GPS node on the mesh is enough.
shiiit then we all good here. Plenty of nodes have gps haha
It's the rebroadcasting delays that will ultimately determine how close any given node gets to the actual time.
(and any delays caused by chutil congestion)
And the more routers the greater the delay for the average client?
Routers are one reason for delays. They aren't the only one.
But yes, every hop adds some delay, every time a packet is left in the TX queue because the node is still receiving, every time a RL delays a packet due to overheard rebroadcast, every SNR-based delay that gets added, etc etc.
And because none of that is consistently measurable or predictable - it varies a lot in a real-world meshtastic network - you can't control for it, and therefore as a result you can't accurately set node B's clock by communicating with node A.
If you're wanting more reliability, I am fairly close to pushing a PR for this: https://github.com/meshtastic/firmware/compare/develop...erayd:meshtastic-firmware:wip-replay?expand=1
Should get the likelihood that a missed packet will not be noticed & fixed to under 1%.
SWEET.. @obsidian storm would love this
it literally might fix is main complaint
consistent message delivery
That is the whole point of this feature. It's basically a really lightweight zero-hop proactive store & forward, designed to ensure transport reliability without adding excessive overhead. Typical on-air overhead is 18 bits per packet plus any replays that are needed.
I figure given that (other than DMs) the only thing a client knows is "delivered to the mesh", the onus is on high site operators to ensure that this is synonymous with "delivered to the entire mesh".
Will be looking for testers at some point this week, if you guys are keen.
Expect bugs
It's a nearly 1300-line feature, so...
I can definitely see that is the goal reading the code changes in the PR. I freaking love this.. uhmm I think some of us might be able to test. Bash.. littleton you all down?
i can put the firmware on jovialis when released
and see if it has more reliable message keeping
Note it is NOT ready in its current state yet. Don't try to run it until there's an actual PR.
(expecting that this week sometime)
We probably wouldnt be down to put it on our routers except maybe lake mountain just because LOL its a lot of work to climb a 9-11 thousand foot peak to go update firmware haha
Yeah don't do that lol. Test it on stuff you can easily get at.
Especially given it's on top of 2.7, which doesn't even have a beta release yet.
i’m up for it!
I have one router late I can easily get to but Id need to upgrade via bluetooth so yeah whenever there is a PR and artifact for the build id be pretty game to convert Aurora Ceresia Router Late to that
to help test
Its sitting on the hill near the weber state campus
about half a mile from my place
Yeah F51 would def benefit from this we will test anything you need 😂
I can put it on lake mountain and we can test quite a lot with that.
Considering we have remote access to that really tall router
Plus C6C is the router late that can see it directly. That would be amazing. Lake mtn and it would be able to guarantee message delivery north and south
I'm expecting some increase in ch util as a result
At least early days
There will definitely be some increase in chutil. What modem preset are you guys running currently?
And what's a typical high-site chutil?
Currently LongFast. With the expectation we move to short fast before winter
I'll pull the long-term average channel utilization from C6C right now
16.1%
100+ active nodes
That is surprisingly reasonable chutil for LF
Well placed routers
Except 1 😂
I do need to implement some preset-based limits I think. At the mo the tunables are all set fairly aggressively for testing on SF.
You know, honestly, it would be very interesting to see what that would do to a long, fast Network. I'm guessing it wouldn't be as reliable because it's expecting a shorter window when it has way more window available to it on long fast
You don't want it dumping e.g. 50 packets of replay all at once. To be fair, you should never end up in a situation where that would happen... but if it did, it would certainly tie things up a bit.
It shouldn't impact reliability due to that. Biggest issue I can see is if there are a ton of replays queued up, but not enough time to actually TX them.
Normal operation should see only the occasional packet get replayed, which is fine.
How to set new long fast ch utilization records with one firmware update 😂
Yeah I really don't want it doing that lol
One thing I have wondered about is gating it to faster modes only. Would help ensure it doesn't get unleashed on default meshes that are already struggling, plus provide people an incentive to move to a faster mode.
To be fair, when I was using RA once I spiked our utilization to almost 72% cuz I could see lake mtn directly all of sudden walking up a hill and things went really bad really fast
What is RA?
Aah yup. That can be fairly chatty.
It’s not like meshing-around is impossibly difficult. It’s highly configurable. I’m sure you could make it more stable by turning off some of the more esoteric functions like checking in/out nodes. I don’t use that or a bunch of other things. It just depends on what you want your bot to support.
It'd be really cool if you could send commands remotely to the computer ota, like SSH over Meshtastic or something.
IP over meshtastic exists. It truly sucks.
I figured as much, why bother with LTE modules if you could, right?
If you wanted to do something like that, IMO your best bet would actually be to abuse the serial module. But yeah... MT really isn't the right tech to be using for that. Not enough bandwidth, plus high latency.
To put it in perspective... on LONG_FAST, your entire mesh is sharing what is effectively the same throughput as 2-3% of a dialup modem link. (ignoring the obvious terrain complications)
Look into WiFi HaLow
https://www.seeedstudio.com/blog/2025/03/06/two-new-wi-fi-halow-modules-making-long-range-low-power-wi-fi-accessible/
It's also in 902-928 MHz,
The Military contacted me...
I reverse-engineered a $20,000 military IP mesh radio using just $106.23 in open-source parts. This isn't a gimmick—it's a fully functional MANET router capable of push-to-talk, GPS sharing, ATAK integration, and more. In this video, I’ll show you how I built it, tested it, and deployed it in the real world—no...
Yeah, I was thinking about some janky workaround using the serial port to send text commands.
Not saying that this is a good idea, but if the aim was to perform common functions, perhaps with input variables, a script name with an input variable or 2 could perform quite a lot (depending what the script is) with very little needing to be sent
Yes, if I had a predefined script to do something specific that could work, but I was thinking about being able to reconfigure the BBS on the fly.
Really cool! MPCIE module in a Nebra would be great if we could get something up on Mnt Ogden with shack power, like you've got on Lake.
Heh. You're thinking a long the same lines. I've already got some test hardware for this exact thing
Tbf IP sucks in general
But it's the best we got for now
I wonder how much you could do with those HaLow wifi chips. Some type of local intranet with solar relays on the mountains?
What devices did you go with? I want to give HaLow a try
The Wio-WM6180 Wi-Fi HaLow Mini-PCIe Module,powered by the Quectel FGH100M-H, is a long-range, low-power Wi-Fi HaLow module. Operating in the 902-928 MHz frequency band, it offers excellent penetration and coverage, making it ideal for smart home and IoT applications. Featuring the industry-standard Mini-PCIe form factor, it's ideal for industri...
need more bananas?
Phaha no amount of potassium will do anything to this. Just gonna hit some spots with the deep tissue massager
Crack that lactic acid out of there 😆
Buddies all good. Minor roll. Will heal very quickly
for solar and battery powered devices, do I want to enable power saving mode in the settings?
as in, I do wish to save power, but I don't want to lose messages. I'm not sure if it wakes up when there's lora activity
no.
just spec it so that's not an issue.
ah I see
shortly I'll learn if my panel is up to it. going to put my rakwireless on my roof
good to hear.
which panel you using? or how many watts?
eh link dead.
6W is more than enough. I've had a RAK node with ~3W panel and it stayed basically fully charged even over winter. For comparison, the WisMesh Repeater Mini panel is <1W.
nice
If you haven't bought the panel already, I would get the soshine instead. It is widely used in this community and you've got a ton of good choices for 3D printed mounts.
wtf why'd I get a notification that a random node, Ollie Tabooger, is advertising my key. does that just mean longfast's AQ== key?
yup. see that often. even saying the node im connected to has done it.
I just ignore it unless its talking about private channels
looks good look at that direct LOS nice!
is it weird that I can get traceroutes from several nodes but sending messages to the longfast channel error out at max retransmission reached?
what is your hop count settings and where are you located again?
also new antenna for the echo is fantastic compared to stock, dunno why I waited
I'm in sandy, just did traceroute to millmesh and got 2 hops out, 4 hops back
plz work!
the hop limit? looks like set to 6 which... doesn't sound right. shouldn't it be 4? node is currently meshtastic 472b, havent thought of a good name for it yet, I've had it for months
I saw it and saw a reply. and sent one as well
crap I don't see anyone's reply. but interesting you saw it!
you have any routers on the list yet?
I saw and sent reply to your roof test as well, but mine says it didn't get acknowledged
what routers do you have on the node list?
hmm I have a bunch of nodes on it, can I filter by router?
I have kr1p, and solar:3
and that's all right now
LPR directly with poor signal, NPR 2 hops. AUR, POTM and lake mountain are in there too but haven't seen for awhile
POTM and Lake Mountian are essential for my area, I think. Lehi
my node is t-echo inside motorcycle seat, outside. not the greatest height or LOS but I thought it was more common to have failed traceroutes with working chat, not the other way around. I'm testing the viability of putting a node on the bike semi permanently. The results so far seem odd
ok wait till you ahve npr / potm on there then try sending messages
will do! I just got someone in sandy
ya im making one for the roof of my truck. will see how that goes.
if this works I"ll be so happy. the freq 20 network was super unreliable for me. couldn't get up to SLC
was it from "472b"?
nope 93dc
yup i see that reply.
ya im not seeing yours.
ok so lehi, ya do you have lake yet? if not hold out for that one
saw the clearfiled one? @mighty pike
if it's full name is "clearfield" not yet
nah it is @manic fractal the psuedo one
once you see lake / potm your list will grow more and should get some real traction at that point
I sent one from TTN / a test tooele one im making did you see anything frrom that?
ah thanks
nothing yet
but not surprising if I don't have lake
ya. wait it out for sure. .
fwiw I also don't see your message from TTN
kk where were you at again? NPR is not super awsome right now so could be acting up. do you get any trace to potm?
let me grab my roof node tablet ill send some tests from there
looks like no, not sure if it its a being under the seat thing or if it's a nearby obstructions thing.
could be both. try putting it in tree or on top of something
astonishing... looks like the traceroutes that complete tend to go from my little t echo through big cottonwood to wasatch crest 2, from there to AUR and then reply is either directly received or a couple hops
ha ha ya .the direct back is funny
i love it when it does that.. potm direct back to my t1000e that is inside is always fun
so probably yeah a height problem, but it's really funny that the nearby geometry happens to just barely let it travel 12+ miles almost perfectly straight through a valley to what is probably someone's roof node
I wish traceroutes would always succeed. I have incredibly difficulty getting them back, like, 5% success rate I'd say, unless the node is super close
not sure what I can do about height on the mobile bike node... but this gives me some confirmation that a roof node at home could be hugely beneficial. since it's in roughly the same direction, but without the whole 12 miles of valley
but for all I know roughly isn't enough and some oddly well placed tree will muck it up 
What nodes are you trying to traceroute? Also, what hardware specifically are you using? Have you tested your cabling and antenna?
I would guess you should be getting better than 5% back based on your antenna placement,.
saw yours in herriman. you get any replies ?
@prisma saddle oh cool got an aknowledged message out for once. really threading the needle apparently. I just went outside to see what kind of LOS i have with big cottonwood and it's not pretty. theres like 2 thick brick walls that are spaced just right to leave a thin gap that is somewhat aligned with the canyon.
yeah 3 replies including what I'm assuming is yours from herriman. but there's no way I can rely on it like this, as funny as I might find it
yup my roof node.
Interesting, this is the window that WC2 sees in the valley.
yeah that seems to align with this, if just barely in red
I did a whole bunch of scouting and planning when I placed WC2. WC1 has better coverage in the canyon and should have hopped through NPR. WC2 is a bandaid intended to reach my friend's roof node near daybreak and Lowe if it ever gets placed. I didn't plan for it to hop through AUR, but it ended up with LoS and usually hops through it.
and that just so happens to cross just enough with where im testing mobile node, amazing odds.
Lucky you. 🙂
I wonder why they don't. Sometimes I can DM a node consistently and perfectly fine but a traceroute will fail.
Maybe it's just the packet size?
a rakwireless wisblock with a larger antenna, and a tbeam wireless with the same antenna, and a seeedstudio t1000e.
the rooftop one is a fairly new install for me. previously I was on freq 20, now on 51 and waiting for nodes to populate
ye I don't quite get it. maybe packet size?
I hope we can move to shortfast
What specific antenna, adapters, cables are you using?
there's something wrong with your hardware
I have always heard this second hand, I'm not going to bother devs for details; traceroutes get deprioritized once channel utilization crosses some threshold. At one point, I got 80-90% of traceroutes back when I ran them for 12 hours from my roof node to one across the valley. I haven't been running them lately because I didn't have anything new I needed to test, and our utilization is slowly creeping up.
Do you have any concerns with some troubleshooting?
I could come by tomorrow for an hour and we can get it figured out.
Sorry. I mean @mighty pike
or you could take it down from the roof and we can meet somewhere and fix it there
and then you can take it back home.
If you can bring your antennas and cables to the hackerspace on Thurs., I'll bring my nanovna. Maybe bash is thinking the same.
But yeah, based on your pictures, your roof node should be performing better than it is.
got message 5 from LAKE!
I think your hardware is borked
We should troubleshoot what is wrong.
I would test your cables. I've had brand new cables that were bad.
it also says ack'd for my two messages to lake
Can you give me a nearby location in DM and I'll go drive by and see if I can talk to lake from there?
sure
Here's what I would do.
- Test the location with known good hardware.
- Try connecting the antenna directly to the RAK.
- Swap antennas
- Swap raks
test 1 would be most instructive. and trying my antenna on known good hardware
If you've got a multimeter, test the continuity on all your cables and adapters.
Leave your ladder up on your roof 😄
We might want to take the known good hardware up there
aaa
oh also how'd you send a message originating from LAKE remotely?
He can SSH to it. They have internet at the Lake Mountian shack.
oo cool
the awesomeness of the nebra.
got a hold of rob! aka kr1p. he should be joining the channel sometime soon. @obsidian storm @real cedar
Yay!!!!
Over here rob @ivory canyon
awesome. I know he was in the other discord at one point
Had to wait the 10 minute penalty before it would allow me to send anything in the channel... New user thing, I guess.
Hazaah!
OK great. First question. What's your preferred form of contact? 😄 You can DM me if you don't want it public (this is a very public discord)
I'm good with any method you can reach me at.
Facebook, email, APRS, phone, discord, something? 😄
I have a very slim online presence...
I'm going to note it down in our private router documentation.
Basically a way to contact people who put up routers if we're having issues with the router
Sounds reasonable.
I sent you a dm, don't post those contact details here
My radio call sign probably gives more info about Mr than most anything else.
yup that’s how i found ya 🙂
You a ham?
heh, yes.
Something is going on with LPR. It's very spotty
It won't complete trace routes, it won't send message acks and it doesn't report firmware versions.
Not great for a router
What's the exact hardware? I'm concerned it's the old chip that has problems
My favorite scripting language (if it can be called as such)!
and won't work with many of the newer meshtastic devices
It's not responding to my pings here on the local network either. I think its hung up.
Its a T-beam.
yeah, which version?
SX1278, is what I'm concerned about
The older t-beams have that chip
and it's hot garbage
Not sure. I will check once I get comms back with it.
How would you feel about upgrading it to a better meshtastic device?
I would love to give you the hardware if you don't already have it.
I know it took a successful firmware upgrade about 3 weeks ago.
What us the best, most robust hardware these days?
I don't mind purchasing it.
There's a few options. The RAK devices are pretty great.
If you have access to 110v, there's some outdoor raspberry pi based devices (nebra) that are really good,
and SEEED has been putting out some really good gear too
I have many raspberry pi boards kicking around here. Maybe I'll look at the Nebra.
Right now, almost all of our routers are built on this gear:
https://store.rokland.com/products/rak-wireless-wisblock-meshtastic-starter-kit
I'll climb up and take it all down this weekend and plan an overhaul.
If you don't want to buy anything, I have about 30 of these kicking around:
https://www.seeedstudio.com/XIAO-nRF52840-Wio-SX1262-Kit-for-Meshtastic-p-6400.html
and I'd love to help upgrade it
(don't forget to DM me some contact details, maybe a phone number?)
I want something that has a functional network stack.
Nebras can be had on eBay for 40 bucks. Great enclosure, RasPi CM3, PoE.
So this is the nebra:
https://www.ebay.com/itm/406218059576
You need to buy/build another board to convert that helium miner into meshtastic (about $20-$30)
but after that, they're great.
Here's a short blog about it: https://meshtastic.org/blog/a-major-solution-to-a-miner-problem/
Looks fun!/
The great part about the nebra is we can get a 1W radio onto them
I run 1/2 Watt on my node.
which filter?
With the nebra boards, we're using an old old cermaic filter
with the RAKs, we're using a gpiolabs filter
What's the goal with the network stack?
For now it's just a lumped element filter but I had planned to replace it with a very selective cavity filter.
I probably still will at some point. But first I need to get the node a bit more stable.
Do you have a good description of where the antenna is located? How high up is it?
The reason for the network stack is the ability to access it quickly and reliably. Nothing worked as fast and reliable as the network method when I was initially setting it up.
We have some folks in westpoint, not too far from you, who can't connect. We're hoping to do some Line Of Sight calculations to see if it's even possible for them or if we'll need to get something back up on the top of the mountains.
I think its a 9 dBi gain omnidirectional about 30 feet up.
We had one on the top of Francis Peak and it was really good until it got fried by lightning
That can happen! 😬
heh. yeah. it happens
Thankfully it's only ever been that node. the other ones have been ok.
I have a node running in the ham shack on Lake MTN and it was getting direct connections to you until yours got all weird.
which is cool because it means 3 hop messages from provo to layton 😄
Haha, yeah. Mine's in lala land lately. 😋
Yeah. It's not very high either, which is pretty important for lora
So the thought is, let's get yours working better, we'll get FPR back up, and then maybe we can get it up much higher or put it into client
Is it on top of a house or a dedicated tower?
It's on a mast on the roof. Nothing special.
It might work well as router_late too 😄
it really comes down to testing once it's stable
I think I have it set to router-late currently.
@leaden crow has some good scripts that we can run for 24hrs and then change the mode type and run it again to compare what it does to the mesh and the util
yes, router_late is great because then it doesn't conflict with the back bone routers
Sounds good. I'll work with it this weekend and see if I can get it working.
The current guidance is:
Ground Level->5000ft = client
5,000->7,000 = router_late
7000+ = router
but regional sections means we should break
since there isn't anything really in the 5000ft+ in that area, yours is great as router_late
as the infrastructure gets built up, maybe then we'll pull it back
Sounds good.
I'd love to see pictures of your setup !
I'll take some pics whan I overhaul it.
No.
it is, by far, one of the biggest contributors to the traffic in the ogden area
If you're not using it, maybe turn that off
It looked interesting and potentially useful but as the infrastructure grows, I could see it not being as useful.
I think the community is looking to move to BBS for that type of a feature
Rgr.
a few BBSes would be ideal for reliable/async delivery
(with replication)
if we had a network, it's easy for people to get messages to a node near them and people on the other side of the mesh/hop-limited can still read them
I'll turn it off when I get it set back up.
what i love about your node is it is connecting to my dad in grantsville. but since its. not fully stable its been hit and miss on his messages and soon. i think if things get fixed working. it will be better for him there. he is the tallest thing there so your hits it nicely.
good timing on bringing it to 51. getting it working will be even more awesome.
That's cool! I had no idea it was hitting Grantsville!
ya we popped it up 2 weekends ago. and yours was what it saw first.
before i even had it on the roof.
so i knew it was going to be golden.
so until we get more placed there like stansberry. yours is the key to tooele and grants.
I might have some time Friday to work on it. I'm trying to take Friday off from work so I can get some more riding in before the cold weather sets in. But afterwards I could probably make time for it. Guess it depends on how long of a ride I do! 😋
nice what you riding?
i am going to be in tooele tomorrow training they ham club on meshtastic. so this will be good news for them too.
i like bash am happy to help with parts etc too. lots of resources here :).
I just bought a superbike... I know, pretty crazy! But I've been wanting one for years!
oh nice! that had to be crazy fast.
Yeah. Its at the top of the heap!
I'm sure some bikes can beat it. But not many. Luckily I don't ride it that way. I have a wife and kids that I need to come home to!
haha for sure. def understand that. cool stuff.
Alright, I'm off to bed now. Catch up later...
Thanks again for jumping on. between this one and the utah mesh discord. lots enjoy the mesh and all.
👍
That was me on Sunday up inspiration point learning how to do 2m SSB. Brought the T-echo along. Would be a great spot to place a repeater node but it's heavily visited.
We tried to contact you on the mesh
Maybe all of us pinging your device caused it to crash 😂
Sorry about that
i got 2 traces out of it this morning. and was able to ping my grantsville node once.. hah. node knew we were talking about it last night? dunno.
That was great, I was able to make contact with the rest of the mesh directly from my house with you up there.
If you want to hike up there, I have the perfect node
Lol, the birdhouse node. It's a good idea, seems a lot of motorists love to go up there.
I also like the bat house approach
Less folks willing to put there face up next to it
Slap it on some unsuspecting tree facing the valley and away from the trail, no one will ever know.
Too much shade I bet under a tree
People put nodes under trees all the time, I'm sure you just have to orient the solar panels optimally.
No that's not true at all. If it's a pine tree and it has a sizable trunk then part of the tree is going to be in sunlight most of the day
You just have to be under the point in which the canopy shades the tree
There's a balance to strike though because you want to be high enough that snow isn't going to just pile into the birdhouse
😆
Huh neat
That birdhouse node is 👌
This ebay seller accepted $35 for nebra nodes!
Huh, I thought everyone did FM/digital voice on 2m
Make any contacts?
Sorry, just realized he said offer
😛
I'm moving everything around at home, server wise. I'm going to take down mesh.thebadplace.org for the moment. If I bring it back later I'll let you all know. If they had their docker issues worked out it'd be sooner.
do you need it hosted elsewhere? or need it up in the mean time?
It's a nice-to-have, there are a number of folks running it at home. I'll have it back soon.
How many of those have you bought? 😆
I should probably get a couple more at that price.
hmm ive got 3.. need to actually get them setup.
once i get my roof mount moved.. its going there..
other might be mil's house.. and 3rd.. TBD..
I got my automatic image builder for the nebras to install meshtasticd and meshtastic cli on first boot / network connection and enable SPI.
So now all I have to do to set one up is tweak a hostname, flash the card (the builder can now do this too), pop it into a nebra or other compatible pi and it just starts working.
nice. that is pretty cool. is this a script yo willing to share we could pop it on the site or so. quick setup of a nebra.
Yeah, I want to make a couple more tweaks. It's a rust app, and it should be able to run on Windows as well, but I haven't tested that.
I might have worked through the docker issues. Someone want to test it for me?
loads, but no nodes, unless it it has to learn them all again
it will, if it works
I want the cleanup script and other things working too, but I have to figure out how first
all tabs work. just missing data
I'm going to break it a bit later, I want the database exposed where I can see it
but it's hopeful at least
I got it working the way I wanted. But, the nodes and conversations tabs won’t populate outside my network. I need to sort that out.
What is max hops set to?
I saw your messages in the valley, but not in NSL
7
I think. I need to check myself. I might have lowered it. But I got Fox fixed up
You should see a bunch of his tests in the public channel
I've been seeing them
Looks like my hops were set to 5
Our network is small enough that I severely doubt that if every single node was set to the maximum hop limit... That our channel utilization would really go up all that much
The only thing I think would happen is we would start seeing a lot of duplicate messages
😂
Might explain why my roof node saw my mobile's message 3 times lol
Well actually most of the time that comes from the hidden node problem
So because we have strategically placed routers in every location, we do have the problem that not every client can see all the routers on the network
Joao clients in other pockets won't necessarily know that that message has reached its destination
There's actually something erayd is working on. That's going to be really awesome for this and guarantee message delivery without duplication it looks like
It would be really good if we could test that on our router lates whenever it's ready. 😆
Yep saw them. What was his problem?
Bad cable
Ah, well it's good you found it and it didn't ruin his transceiver. No continuity, or was the center shorted to the braid? I've seen both on Amazon/Aliexpress cables.
I didn't even test. I just swapped it out with another one he had and it worked
Nice, easy fix
yeah, quick an easy!
I enjoy seeing the data it puts up. Thanks for hosting it.
Thanks @manic fractal . Remember everyone, if you want to see your data on there you have to set “okay to mqtt” 👍
It’s working, some of the menus won’t show data outside my network. I’ll sort that out tonight, not sure why that’s the case. The data is there, I can see it.
Fixed, should be all good now 
The update is using the /api path for the other tabs now. I had blocked that because /api/config shows server secrets 🤷♂️
Now only the config path is blocked
Re-flashed my T-beam and made a couple of minor changes to its configuration.
It now has the latest stable beta version of meshtastic firmware.
I also disabled the store and forward stuff.
Feel free to run some tests on it/verify its behaving properly.
Also, I think its putting out closer to 1 Watt now.
One other thing... My T-beam is hardware version 1.2 so I think this is the good/more stable version.
Thanks for the update. Hopefully a reflash may have shaken out any bugs.
I think that is an SX1276. My T-beam 1.2 is.
https://store.rokland.com/collections/lilygo-t-beam-series/products/lilygo-ttgo-meshtastic-t-beam-v1-1-esp32-lora-915-mhz-wireless-module-wifi-gps-neo-6m-with-oled-display-soldered-for-arduino-q349
The SX1276 cannot talk to new LR1110 devices like the T1000E trackers
I also installed an external WiFi antenna so that it connects to my AP more reliably. Yeah, hopefully this later version of firmware will make it more stable.
If it is an SX1276. I will give you one of these to replace it:
https://www.seeedstudio.com/Wio-SX1262-with-XIAO-ESP32S3-p-5982.html
This thumb-sized LoRa dev kit, featuring the XIAO ESP32S3 and Wio-SX1262 LoRa module, is an ideal starter kit for Meshtastic and LoRa/LoRaWAN projects. Supporting WiFi, BLE, and LoRa, it includes a built-in power management chip and can be extended via IIC, UART, and other GPIO interfaces, with compatibility for Arduino development.
It replaces the module that is currently on my T-beam?
No, it would replace the whole tbeam. It has an ESP micro and the better transceiver (SX1262)
I need something with a network stack though.
It does, it has the same ESP microcontroller, will do wifi
But it looks like one would need to spin a breakout board for the filters, connectors, etc.
You don't need a filter. The T-beam doesn't have one. You would use ipex pigtails for your antennas. I can give you those too.
I guess there must be dev boards that could provide some of that as well.
If you want a filter, I will make you a cable with one, like this:
#1197577977781821541 message
What sorts of improvements does this newer module provide other than TX/RX performance?
The XAIO are tiny, but they have everything you need. You don't need a display on a remove node.
The SX1262 and SX1276 are both low-power long-range wireless modules. The SX1262 is an ideal choice for low power applications that require long-range wireless data transmission. It supports a maximum data rate of 300 kbps over distances of up to 15 km. It also features a low power consumption of 6 mA, a high sensitivi
I've got an inline SMA filter that works reasonably well. I plan to replace it with a proper cavity filter at some point.
Like the GPIO labs filter?
I have an Acasom on one of my nodes. It's definitely nice in noisy environments.
Yeah, the display doesn't even get seen by anyone. Not needed at all except for initial config when Bluetooth code is shown on the display.
That Rokland description is actually not great. You get slightly higher broadcast power with the SX1262 (22 dBm vs 20 dBm), really the huge downside with the SX1276 is that it won't talk to LR11110 devices, the Seeed trackers are one of the most common devices now.
I see. Well, I guess I need to plan for an upgrade soon. 😋
LMK if you want that Xiao. I bought a couple of them when they were first released just to steal the transceiver and make rpi zero sized hats with them. I only ended up making one of those, so I still have one I'm not going to use for anything.
Thanks! I'll let you know.
Also for everybody in your device module configuration if you want to configure a time zone the time zone for our area is
MST7MDT,M3.2.0,M11.1.0
By the way, what is your node named?
I have a couple nodes. You'll find my main node that's always on me is name to c6u Umbriel.. my other node is c6j Aurora jovialis
My two router nodes are Aurora Ceresia C6C and Aurora uranialis router AUR
My nodes are named dsr1, dsr2, wc1, wc2, dst1-3...
Speaking of routers, could you please put your device into client mode
The router I have on the hill can literally see all of Davis and all of Layton just fine
In fact, I have a router late up here higher than anybody in Layton could have even if they had a really tall mast 😂😂😂 .. It's both a blessing and a curse because whenever something happens or whenever channel utilization peaks I usually see it first up here.
I live up close to the mountains here near the base. Very close to the base
Mine fills in some areas that don't get coverage, I've been told. At least until they get the one up on the mountain back online. I moved it from router to router_late.
Mine ties in the Tooele valley, which I wasn't aware of until yesterday.
I got your test earlier, so it's definitely improving connectivity up North to SL.
Well, if you read the documentation on the messtastic official docs they say that a client does the exact same thing so you would still be doing that. You wouldn't be hurting anybody
But what you would be actually doing is benefiting people and that the routers that actually can do the bulk of the heavy lifting would then do it and your client would forward the traffic to them
Right now I'm forced to hop through you even though I don't need to do it
Especially on one of my nodes where I have direct line of sight to Lake Mountain. But when you're in router late mode then I have to hop through you because that's what it chooses to do most of the time
Does that make more sense
The difference isn't that either three of those modes won't rebroadcast it's when they rebroadcast. Client router late and router all rebroadcast messages. It's just router is first in the contention window. Router late has a different contention window and client generally waits for both of those to handle their traffic.
Router late is a little bit more complex. We actually have a contributor here that can talk a lot about it. @broken turret
But anyways those three all do the same thing just at different times
@KR1P
What amp are you running? On my traceroute, I got a return hop that went from your node to Lake Mountain. That's pretty huge.
It's a block amp that is putting out about a Watt.
I get the same massive jump and I don't have an amplified node but it is on a station G2. Which is a 1w device
I think both him and I have one watt devices in our really good mounting spots
Right as I'm ready to test bash is like hang on. I froze my pi after I just upgraded my firmware and did a device reset
The timing of that is BEAUTIFUL 🤣
Not an external device, but I would still describe it as amplified since it has an integrated PA. I wish Unit Engineering would build a hat. I think their PA, LNA and design is probably better than the Ebyte. I just don't want a crappy MCU if it's going to be a high power node. Would much rather have meshtasticd
You know I really wish they would build a station G2 but base it on the nrf-52
I would freaking love that
This is incorrect. ROUTER goes first, and preempts everything else. Then CLIENT and ROUTER_LATE will have a go, using the same timing rules. If ROUTER_LATE hears anything else rebroadcasting, then it shuts up and defers its own rebroadcast to an even later window, after everything else.
If CLIENT hears anything else rebroadcasting, then that client shuts up, discards the packet, and will never sends it at all.
For solar nodes yes, but I bet if they built a hat it would perform better than the ebyte.
Okay yeah that's a way better explanation and that's preferred where he's actually located. Considering there are much higher elevated devices with much more line of sight
So his device deferring the rebroadcast to those two other devices would be very ideal and still fill in gaps for devices that just can't talk to those other two devices
Which I find to be very unlikely considering where one of them is located and how much it can see
😆
@real cedar Would a blog post on this be useful? I feel like it's something you end up having to explain to people a lot in here, and that way you would have an authoritative resource to point people to.
Dude I would love a blog post. Oh my God you would be a lifesaver
Just having something that explains all of that so that I can just be like. Hey here's this URL. Go read it would be amazing
Yup. CLIENT is good to help devices that can hear well, but can't TX strongly enough to be heard (e.g. the T1000-E in the middle of CBD office towers).
If you need the device to be guaranteed to relay traffic in all circumstances, but don't want it to preempt, then ROUTER_LATE is the correct choice. Not intended for rooftop nodes etc though, because it adds to chutil load.
There is a new role coming (CLIENT_BASE) that is intended for rooftop nodes, and situations where you only want guaranteed forwarding for packets to / from specific nominated other nodes.
I will write you one, then. "Demystifying ROUTER_LATE" is now on my todo list.
And that's what I'm trying to point it out here. And I do appreciate you making the client base role. That's freaking awesome
His node is in a very good spot. It's on a rooftop very high up in Layton parkway, but the C6C node and aur node are at least 2,000 ft higher than both of those and have a very clear vantage point with equally good rx/TX.
When we do get that new role I would love it if he went to client base because I have a very strong signal from him.. It would be a very ideal role for his device and it would probably really help the reliability in the Davis county area
Granted that's on my device that has a 10db antenna on it 🤣
But I still get 3 DB down to Layton parkway with many many things in the way. And that's pretty fucking awesome which means he has a very solid amplifier on it
Haha
That would be awesome and honestly maybe if you can talk about client base.. and some of those things coming out or maybe don't.. probably wait until that's tested 🤣
Its been changed to, "Layton Parkway Node" and its just a client now.
I can talk about CLIENT_BASE too. Will probably just mention it in passing though; that one really deserves its own blog post.
I had hoped that being a router would improve connectivity for this area but if client mode is better then that's what it'll be configured to.
@ivory canyon We'll do some traceroute testing and see if that helped or harmed the network
is it super difficult to switch it back?
@leaden crow do you think you could get that script running?
sure, I'll run it overnight
cool, then we can switch it back to router after it's done and then compare
Sadly, it's a mesh Network and not an infrastructure mode Network. So as much as router might seem very lucrative most of the time, it actually harms more than it does. Good. We're finding out that minimizing routers as much as possible and just having really strategic ones works a ton better with a lot of clients in the valley. The only suck ass thing is we have to climb mountains to go put it up there
😂
Its easy enough to change settings. I can access it over the network from anywhere.
At the end of the day though, don't feel bad. Your device and client mode is still helping every single other person
Client is the kitchen sink 99% mode that is just a workhorse
👍
Let me know when you want to test router mode on it.
For now, I'll assume client mode is best for my node but I would be curious to see what your findings are.
Yeah that will actually be really good data to collect because we have neighbor info being collected on @vital hemlock server. In fact, if he could give you all the data to connect your node up in that way so that you could collect your neighbor info and send it to his server.. since your node is network connected... That would be awesome. It makes us collect a map but it also gives us a good amount of data as to who's connecting through who.
And then we could actually get some data to see. Would you being router late versus a client help or detriment the network
That would be an awesome case study to just put on the website as a example of hey. We have data on why this may or may not be a good idea and you know what. I could end up being totally wrong and we could just all get proved wrong and it turns out your device as a RL really does help vs being a client and if we get that data I'm all for it.. 😆
Its all about testing and experimenting. 😋
Let me know what to set up and how to get the data to where it needs to go. I have configured a MQTT server before and used it to tie geographically separated meshes together in the past.
Not the same thing, I know. But some of the same methods being used...
Anyway, I can set up some custom configurations to test stuff. 😋
@real cedar Anything else you want to see added to this post? https://github.com/meshtastic/meshtastic/pull/2030
Good morning all. We set up a new solar node at the north end of Utah Valley. It has a clear and unobstructed view of the entire valley. The node is "MEC UV North Base" and is a RAK module and has a 5dBi antenna. (one of these: https://www.etsy.com/listing/1795573484/peakmesh-magnet-climber-solar-meshtastic)
It is currently set up as client but we're wondering if it should be set up as a router or router_late. Any opinions? The reason is that there aren't many nodes in Utah county yet.
View from the node location.
Not at all you covered it extremely thoroughly.. I appreciate that
Please do not set it up as a router or a router late. A client is just fine
Are you on 51?
There is a router on lake mountain if your down south
It should see lake mountain there based on the picture. @obsidian storm runs it. I would coordinate any issues with him
@acoustic mason
I am not seeing it on fs20 or 51. What freq slot are you setup on?
It's on 51. I have not seen Lake Mountain rounter on my node list for at least three weeks. Is it still alive?
It's on 51 but I think I set it up to only broadcast node info every few hours.
Ah ok. Ya i have not seen it yet. what is your hop count at also?
and general area? is this like alpine/AF ?
Hop count is set to 7. It's east of Thanksgiving point about 1.5 miles. SE of Adobe about 1 mile.
Ah i see. its on this social climb building?
so prob too low for Potm, might see it. but ya should be lake for sure.
Yes. I and a few other Meshstastic users work for ADI in that building.
I run the engierring QA department at we're on the third floor.
Very cool. lets have you try sending a message on longfast.
when you have some time
I'll do that once I get into the office. It's set up as unmessagable at the moment and I don't use if to send messages either. It's really meant as a relay/repeater for the valley.
ok cool, if messaging is left on, it could be used to send test messages for signal tests as well. But ya. cool eager to see how it works
OK, I'm headed into the office in a bit. I had a remote meeting this morning so I worked abit from home.
Could you DM me a location of the node?
Or post it here
LAKE mountain is very active, we were just playing with someone in Lehi over it yesterday
This node is a perfect candidate for client_base
Weird, i dont see it yet.
hmm Must be my amazing roof node placement.. 🙁
Im seeing the test pings in sandy but I'm in a bad tx zone
saw the message but not in node list yet.
ah there it is...
ug.. i really need to get on my roof this placement is not doing well
and swap to the nebra
https://mesh.thebadplace.org/chat
https://meshview.freq51.net
MEC shows in both sites
Is there a way we can get that above the building?
I'd love to see the signal go east too. East it'll be blocked by the building
Very cool. That's great. I think it's a good placement. Although, it can't see to the north direclty.
I don't have access to the roof, unfortunately.
put t here? lol
It's blocked to the north, It can see E, W, S.
or ya. on top.. darn.
I wish I could! There aren't balconys that I have access to besides this one. There is a perfect one on the SW corner on the 5th floor but that is property of another company.
Heh. I wonder how your landlord would care if we put a long antenna up from your balcony
You're in a good place
other than lake, thats prob the best placed in UC
Not good enough for router_late long-term, but it might be a good candidate for a short-term router_late while we work on building the infra
I think this is a consequence of using a private channel for telemetry. The mesh will function correctly, but they either don't appear in node lists or they appear at the bottom with all N/A values.
│ 131 │ POTM Router │ !77272d1f │ POTM │ RAK4631 │ x1TctwkNJnyjm7F2XEobRoRYYZl2SwHLv+FCqnP9fFk= │ ROUTER │ N/A │ N/A │ N/A │ N/A │ N/A │ N/A │ N/A │ N/A │ 0 │ N/A │ N/A
ya i dont get much chatter from them on normal nodes that arent part of the tele channel. like does not hear from it in days
Normally we would want a router_late above 5000ft+
https://freq51.net/advanced-configuration/router-deployment
But where we don't have one right now....
Let me know how you need it configured and I'll set it up. Right now, it's set as client.
How would you feel about putting up a long pole up from your balcony to get above the roof?
I'd feel great. There's no way that building management would allow that. We've had a hard time dealing with them already. I'd love to have access to the roof but I don't have that either.
That's a no-go.
Dang. Ok, what firmware are you running?
v2.77
Sick. Ok, I'll run some tests for about 24 hours in the current mode, then tomorrow let's switch to router_late and we'll compare to see if it makes it better or worse
OK, sounds good.
When you do tests what exactly are you testing for? What factors are you considering?
Lol, I finally got a ping from LAKE with the Nebra Hat
That SNR though.
Signal Bad lol
Did you use the filter on your Nebra hat or go through the IPEX?
I'm using the filter + sma connector right now
I cant get a node info back from 54ef2f2e
meshview wont add it until it gets a node info packet
not sure why it wont reply with one 🤔
someone else on android select it on the list and try an "Exchange user info" for me?
Done
Wait, is it on mqtt already?
What do you mean?
I see it on your 51 meshview, it's replying on the Freq51 channel
that's why I dont see it on LongFast meshview
Soo... Mine is longfast too
Because you're a hacker, I get it. 😉
Yes.....
Do we still need mine, or should I bring it down?
It's good to have both
So we can see if messages are making it to different places
I dont mind leaving it up, but I think I would prefer having both channels too. It'll take some hacking on this docker to make it work
Once i get things settlled in grantsville/tooele. maybe ill setup a meshview there. so we can see whats hitting that..
If thats needed. for troubleshooting that is or w/e.
once pops gets network to his barn..
I may be taking tomorrow off to do a long bike ride from Layton to Orem. If I can't switch it tomorrow, I'll let you know when I can so you can do your tests when it's set up router_late. Most likely, I could come in over the weekend or at the latest, switch it on Monday.
No rush
Cool.
@acoustic mason Your node is showing as 1 hop away from me and I'm in Layton! However, a trace shows 2 or 3 hops. I guess that means sometimes we hear each other direct?!
Or maybe that means one node in between...
Probably the lake Mountain node.
Probably 1 between, there's a filter on the app that lets you see direct nodes only
Ah, OK. Its probably a hop through Lake Mtn then. Still pretty good!
what do you see in the traceroute?
here is mine lol @acoustic mason
How strong are those magnets? I may go with one of those for my unused chimney if it'll hold up well
I can't speak to his magnets, but I know several people have had luck with these:
https://www.amazon.com/dp/B0CYZM164V/?coliid=I27PO4Q6CN7M&th=1
yup thats whats on my car node.. pics coming soon. they are strong magnets.
You show up as 2 hops for me. 1 hop means that it is going through one additional node. zero hops is direct.
Very strong. I have a hard time removing it. LOL There's no way its' coming off on its own.
Perfect. Think that's my next mesh purchase
It is a lime well and permanently mounted to a radio building up there with a power source directly wired into the building. So yes, it is very much alive
Wait someone made a 3d printed car roof node.? Link to files
Thanks
full power of ASA for the win
I'll probably use nylon cuz it's automotive and paint it my cars color. Additionally I will not be using magnets. I will modify it to have two screws that screw into my rear antenna mount. Magnets don't hold well in high heat on dark paint because they demag from personal experience haha but I'm glad some have had a good experience. That and I routinely enjoy exceeding 100 mph on open stretches. Mags aren't gonna like that 😆
But thanks this will be great
I'll use the gizont whip not a rigid antenna as well
Should be able to make the car chassis it's ground plane
Which I mean LOL talk about the kingpin of ground planes
hmmmm now your making me rethink it lol
I dont go 100.. though
I've got personal engineering experience to back up that they will demagnetize over time in the heat and while they may not permanently demagnetize or demagnetize to the degree where you don't think it's magnetic, it could go low enough to the point where it just flies off the car and I don't want to be responsible for that with my insurance
Screws are cheap and I have an antenna mount so I might as well use it
What I really want to know because I can't tell from the printables page is can I fit two 18650s in there
I'm on mobile at work so I don't really have anything where I can pull in the 3D drawing right now
nope..
Oh jeez okay I'll make my own then. I don't trust the silver flat cell type lithium batteries in direct sunlight. Those bad boys have burned me before
😂
Someone uses orca slicer and didn't simp to Bambu.. ABSOLUTELY BASED
pfft.. never!
A man of taste and culture
I was thinking of just adding a little extra volume to this one and slapping a 21700 in it, I don't like the pillows either.
Orca all the way. creality ftw.
I've had one burst into flames before on my kitchen table. 😅
Honestly yeah I would say the same thing. The bamboo is cool for like 30 seconds until you realized. Oh it's really just a k1c and a slightly thicker bed and a couple tweaks here and there.. with a little bit better firmware out of the box which doesn't matter. Anyways cuz if you're any kind of serious 3D printer you're going to go off and use klipper3D. One of my biggest gripes isn't actually with the printer, but rather it's community. A lot of people buy that printer and then they just sit there and print plastic toys forever
But you don't do anything super useful with it
🤣
If I had $1,700 kicking around, I'd just build a Voron. One day...
lol ya all my halloween/christmas house outdoor decor is 3d printed.
B-B-B-bub-buh-based
Yes
Vorons are more amazing than any off the shelf printer with all it's possible upgrades COULD EVER be
It's my plan to build one as well but I'm budgeting for a really awesome build
I'm going to go turtle box ams
Automatic cutting, purge and retract
4 filament AMS ideally
Dual print head with dual extruders
I've got a heavily modified ender 3 with a microswiss running klipper
And I will be designing it's chamber and components to handle 500 C build plates and 150 C chamber temps.. which is about as high as I believe you can safely design it to handle
I want to be able to anneal in chamber with gcode
I've got a k1c with a micro swiss flow tech hotend, upgraded direct drive hardened gear extruder, and a diamond nozzle with a chamber heater in the printer
I've learned what I needed to learn on it. It's been a phenomenal printer. No regrets choosing it over Bambu but it's taught be exactly what I wish to have out of a voron 2.4
For those who have 1500 dollars and just want the kit.
https://www.fabreeko.com/products/ldo-voron-v2-4-kit?srsltid=AfmBOorEtIoyFMtMPi5bBpU8N0AOU_SP2-95Zg-J_QcH0ilFZYDsWTLl
Here's the kit
It will absolutely decimate and destroy. Bambu in any 3d printing challenge 😂
I was looking at the Tridents
yup same.
Nearly the same printer
https://docs.ldomotors.com/voron/voron-trident
ender 3v2 with DD, dual z, mircosiwss. and fulament bed (waste of money). .
silicon bed level as well
Personally, one of the things I don't like about it is the bed leveling sensor. A long time ago Eddie sensors weren't all that good, but now Eddie sensors are fantastic and there's no excuse not to use them, so don't use ancient technology like the clicky sensor
Modern Eddie sensors can do a much better job and they can do it extremely quickly
Fancy with the v2 here 😆
The cartographer is one of the best Eddy sensors out there
Lake MTN is pretty OP in getting traffic from Utah County up to Layton.
It used the adxl chip
that's the power of a really high 1W node
@ivory canyon
Successful traces: 30
Failed traces: 18
62% Success
Outbound: -4.07dB mean SNR from 28 recorded values
Inbound: -8.36dB mean SNR from 28 recorded values
Well :/ that's not great
Actually that's pretty good for the number of hops he is away from me.
Route traced towards destination:
!eba4d26b --> !8f7bd9af (-16.0dB) --> !73e074fc (-12.75dB) --> !e0d01bcc (-11.75dB)
Route traced back to us:
!e0d01bcc --> !452b53bf (5.25dB) --> !e745fa4d (-9.75dB) --> !8f7bd9af (-17.0dB) --> !09946e36 (-3.0dB) --> !eba4d26b (-11.5dB)
What are your percentages for aur and POTM when you do the tests on them?
Just for something to compare to
I'm not using NPR cuz you know NPR is borked
😆
I'd have to run them. When NPR was working well, I had 90% successes. Usually my connections to AUR and POTM are not direct, so it won't be as high as 90%.
Ya know this could be a good way to evaluate a node before turning it into infra node.. what's a good baseline success rate for us to consider being good?
Cuz idk 62% seems low to me but I also have a 93% success rate out of 50 traces to AUR from my house
I think it's great for testing equipment or settings changes, but you need a before / after to the specific node. I think the variability is too high to have a general baseline.
Yeah, and I guess things can interrupt the trace route can't it
Like if the firmware throttles?
Especially if some required hop is high utilization. I think the most useful thing is the mean SNR values. Like if you get a significant improvement with new antenna or filter. I'd have to dig to find it, but it was a great improvment when I added the Acasom to dsr2.
Can you run traces to Aurora J now?
Sure
This will help us know if having it as a router helps getting to Aurora Jovialis (!a2ebc068) better.
Because right now, It's coming straight from LAKE
I'm curious if LAKE will go straight to Aurora Jovialis too
and skip LPR
and if we turn it back on, what'll happen.
those are the two common routes I've seen
when looking at your traceroutes
ah there we go.
Do we have anyone around ogden with a non-roof mounted device that can be left outside for a day but on the ground? (non-tracker because LPR can't talk to those)
after Aurora J, I'm hoping we can use ^ that device for our last test.
then switch LPR to router_late and then retest
Aurora J is a client
Just so everyone's aware
I'd love to see traces to Aurora C though as it's the northern most and highest router late. I know I can hop directly from it to AUR and Lake when I run trace routes from it on my phone when in Bluetooth range
I'm curious how it's doing on its high perch ... fyi the control tower of Hill AFB can literally see it at eye level with binoculars... That's how high it is
Lol
@leaden crow do you have the script you use to run the trace routes. I'd love to expand on it if it's python and be able to make nearly formatted neighbor charts and trace charts with it
It's bash, but it's a 33 line script. It would be trivial to rewrite it in Python.
https://github.com/dagronslayer/meshtastic-bash-scripts/blob/main/mtrl.sh
Awesome. Yeah I'll probably rewrite it soon.. could use a tester hahah
There's got to be other groups that are doing testing like this out there
To try and stress test and figure out the better configurations
ya there has to be. i know bayme.sh uses medium fast. so im sure they have some things figured out
just from what ive read.. never talked to any of them
They are mostly on Medium Slow
https://meshview.bayme.sh/nodelist
Ah right slow. umm.. it can be intermixxed? or the meshview is just able to listen to both?
meshview is just looking at mqtt
ah ok
so multiple devices can send to the same mqtt server
rgr
Intersting..
ya i like this idea https://meshview.bayme.sh/net
just to check connections to see if things are wavering weekly/etc..
Shall I change it back to router_late so you can re-run the test?
I would love to be able to do scheduled messages .. ha ha set my roof and other nodes to do it weekly. just for science..
guess i could witha nebra. crontab etc..
The best test for your role is actually tracerouting a client that hops through your node. I'm doing that right now with C6J. Lets wait 12 hours for that to finish before you switch to router_late and I'll run it again.
where are we wanting broadcast intervals for node / tele / power / gps now days for client/client mute. and eventually client base? @leaden crow @obsidian storm @real cedar
fine tuning the onboarding / config guides
currently it says " what ever you want"
and if you didnt see this #1197577977781821541 message wonder how much that helps.. for med slow / long fast. and their network
I put the accurate numbers in the documentation already
For node info and all broadcasts that are related.. clients 21600... Rl 43200.. R .. 86400
Telemetry no more often than 1800
Neighborinfo 14400 if enabled and on the nodes that need it
Ideally, it really shouldn't be whatever you like because the reality is the developers have said multiple times that increasing any of these intervals really doesn't help, particularly for node info broadcasts
I think a lot of people think just by turning it up they'll somehow magically discover everything faster and that's not true
All you'll do is congest the network for all of us, including yourself. At least that's my understanding from how it's been explained to me
right thas what it is when it was added first off. so felt we need to adjust that 🙂
Yeah we definitely need to make a point of that. In fact, I would love to find where in the documentation they describe how node info works and stuff because I would love to reference it in our documentation so they know it's not just arbitrary
I know sometimes it can feel arbitrary but a lot of this has been figured out by just looking at the source code, talking to developers and all that kind of stuff
For node info and all broadcasts that are related.. clients 21600... Rl 43200.. R .. 86400
so that would include tele/gps ?
oh i see telle
sorry mised that
Gps just use the smart algorithm to account for being faster than a reasonable interval and from what I can find, 10 minutes or greater should really be your interval and if the node starts moving to the point where it actually needs to be more often, then allow the smart broadcast algorithm to handle it
The idea is you broadcast when necessary and you minimize when unnecessary. For example, if you have a stationary node, there's no reason you should even be broadcasting GPS that often, right? But if it's moving in a car at 80 mph, I could totally understand wanting it a little bit more frequent
agreed
One of the things I really like about the rak12500 is it supports the smart algorithm really well
Non ublox gps devices are far more buggy
But that's the cost of "cheap".. it's enshittified experience 😂
Telemetry/gps no more often than 1800 is a good start then add the smart broadcast caveat/solution for gps.
30 minute updates is plenty even for weather haha
do you suggest anything diff for t1000e?
No, it's really just based on the mode
If we made a device specific recommendation all the time.. that would get very confusing for our users
true.
The meshing-around bot can be set to do announcements and messages on a schedule.
@real cedar That blog post wil be published after the CLIENT_BASE role lands in an alpha release. So imminent, but not quite yet.
Rationale is that because it talks about CLIENT_BASE a bit, it's important to get the ducks in a row on that first. CLIENT_BASE will land in the other docs at the same time.
right. thanks.
Sure. It might not be exactly what you’re looking for but it does a decent job of putting out scheduled messages and there’s some pretty good granularity on when you want it to broadcast.
cool. us i’ll have to get my nebra up. maybe next week. starting to do halloween decor. so ladder will be out.
C6J
Successful traces: 17
Failed traces: 31
35% Success
Outbound: -1.47dB mean SNR from 17 recorded values
Inbound: -.77dB mean SNR from 17 recorded values
@ivory canyon if you switch to router-late, I will run it again
I need the nicest filter we can reasonably get for a 1W nebra radio.
Any suggestions ?
I'm going to try to tackle farnsworth peak with a 1W nebra. with @solemn sedge
A 1W node up there would change the experience for SLC, Davis, Weber, and Tooele county. It would be night and day.
@leaden crow you've used the acasom, right?
If you want to go down the rabbit hole of cavity filters, search nullrouten post history. A bunch of the stuff he has researched is expensive or unavailable. The Acasom is not the best, but it is reasonably priced and available. Because it won't fit in a Nebra, I would get it with N-connectors and just mount it outside
I'll probably mount it in a plastic box.
I think @real cedar might have one around that was intended for NPR
Who's post history?
Ah, I see.
#antennas message
He has fit some cavity filters in Nebra boxes, but I think only by gutting the OG hardware. The Nebra really is a good deal just for the enclosure alone.
The cavity filters should be waterproof and OK mounted outside an enclosure. They also give you lightning protection.
"Furthermore, RF Cavity Filter provides the best protection from lightning strikes (50 kVA or double of typical lightning arrestors), and if designed correctly they come for free with a proper RF Cavity Filter design."
I couldn't tell ya what the best filter is. I'd go with what the community info and consensus says out there
I feel bad 🤣 I unplugged it yesterday and only plugged it back in around 3 am
My bad
That SNR is strong though
Very strong