Not using the bad names Meshtastic is using could be something good also for not experienced users

[serialrf433]

Currently Columba is using for presets names from Meshtastic.
But it is known that those names are just a terrible choice in Meshtastic. Based on personal talks with even technical experienced people, i got for example told things like this:
LongFast have more range and higher speed then ShortSlow because it is named 'Long' and 'Fast'. Both would be better then 'Short' and 'Slow'.
This is just 😮‍💨

Columba is a project targeting the masses. But making the same naming mistakes Meshtastic did is probably not a good idea. Teaching people the wrong thing because they already know some bad name decision is probably also for Columba not a good choice.
Greetings to people naming things still the wrong way by using Pound, miles and all the other wrong names for things.

I also do not think there should be something like a name for LoRa settings. Many people are putting in general discussions examples with other things in their life. So lets put here such a thing. For example 20°C&&55% humidity&&1000hPa = George.
Why should it be named George? Why should people talk about 'George'? This is stupid. Just name the data standing behind this name.
Now about the 3 LoRa values we are talking here:
https://github.com/torlando-tech/columba/blob/main/app/src/main/java/com/lxmf/messenger/data/model/RNodeRegionalPreset.kt#L140
LongFast == SF11&&BW250kHz&&CR5
Then use this values to talk about this. Do not name it George, LongFast, HyperSuper or whatever.
The names for the LoRa settings are already defined by the company who produces the LoRa chipsets we are using. Semtech.
For example go to https://www.semtech.com/products/wireless-rf/lora-plus/lr2021 , press on on Documents and read the PDF file named LR2021 Datasheet. There are all the names for the different LoRa settings.

And now from the usability point of view. Of course it is to expect that people opening up Columba would have never heard of a spreadingFactor. They want probably to have something that is just working. Something they can use to communicate with other people. So lets make this possible without using names.
What is required for people to make it possible to communicate with each other? The LoRa settings have to be the same.
How to reach this goal? By putting the world into LoRa regions and counting the already existing Reticulum LoRa device settings.
Happily we have now public RNode LoRa settings: https://rmap.world/
For example scrape the data from the USA, and show as a possible preset something like for example:
100 people in the USA use SF7&&BW250000Hz&&CR5.
50 people in the USA use SF12&&BW125000Hz&&CR5
and so on.

By doing so, you are removing stupid names. Instead of the names there are now statistical probabilities. If you are in the USA and want the highest chance to communicate with other people in this country, you choose the USA-preset 100 other people are using because this is the highest number == highest statistical probability to meet on this setting someone else.
Yes, of course this 'preset' gets updated in Columba from time to time when the data from https://rmap.world/ changes. When from the example '50 people in the USA use SF12&&BW125000Hz&&CR5' the half of them move over to '100 people in the USA use SF7&&BW250000Hz&&CR5' after for example 6 months, then Columba updates it. The preset is then
'125 people in the USA use SF7&&BW250000Hz&&CR5'
'25 people in the USA use SF12&&BW125000Hz&&CR5'

I vote to also add the frequency in the same statistical dropdown option. The Meshtastic names do not include frequency. Lets expect people have as less knowledge about what LoRa frequencies are used the same like what spreadingFactor is used.
What Columba should expect from their users to know is the name of the country they are at 😆

Have i missed something important in the idea i explain here?

[scanhead]

I'm only bringing it up because the meshcore lora users seem to have a better "out of box" experience, and that seems to be what you are going for? I have watched the lora settings debate for meshtastic and meshcore quite closely. Meshcore is in the process of standardising to one general setting because they have found it the most effective for the medium. I don't recommend copying anything that meshtastic is doing. meshcore lora settings are directly learned from meshtastics mistakes. MeshCore (by Ripple Radios) standard LoRa settings are designed to be efficient for low-data mesh communication. While settings can be manually adjusted via the web flasher console or mobile app, the community has moved toward "narrow" presets for better performance and reduced interference. Meshcore recommends letting the users decide the best local frequency because of local interference and they do quite detailed site surveys before choosing a frequency for their area. Agree that names need to be based on specific config but frequency should be pooled from the maps data as you suggested. Standard "Narrow" Presets (Recommended 2025/2026) Many regions have transitioned to a *62.5 kHz bandwidth* to improve signal-to-noise ratios and fit between existing ISM band interference. Region Frequency Bandwidth (BW) Spreading Factor (SF) Coding Rate (CR) *UK / Europe* 869.618 MHz 62.5 kHz 8 8 (4/8) *USA / Canada* 910.525 MHz 62.5 kHz 7 5 (4/5) *NZ / Australia* 915.8 MHz 62.5 kHz 7 or 8 Region dependent

…

On Sat, 28 Feb 2026 at 08:07, serialrf433 ***@***.***> wrote: Currently Columba is using for presets names from Meshtastic. But it is known that those names are just a terrible choice in Meshtastic. Based on personal talks with even technical experienced people, i got for example told things like this: LongFast have more range and higher speed then ShortSlow because it is named 'Long' and 'Fast'. Both would be better then 'Short' and 'Slow'. This is just 😮‍💨 Columba is a project targeting the masses. But making the same naming mistakes Meshtastic did is probably not a good idea. Teaching people the wrong thing because they already know some bad name decision is probably also for Columba not a good choice. Greetings to people naming things still the wrong way by using Pound <https://en.wikipedia.org/wiki/Pound_(mass)>, miles <https://en.wikipedia.org/wiki/United_States_customary_units> and all the other wrong names for things. I also do not think there should be something like a name for LoRa settings. Many people are putting in general discussions examples with other things in their life. So lets put here such a thing. For example 20°C&&55% humidity&&1000hPa = George. Why should it be named George? Why should people talk about 'George'? This is stupid. Just name the data standing behind this name. Now about the 3 LoRa values we are talking here: https://github.com/torlando-tech/columba/blob/main/app/src/main/java/com/lxmf/messenger/data/model/RNodeRegionalPreset.kt#L140 LongFast == SF11&&BW250kHz&&CR5 Then use this values to talk about this. Do not name it George, LongFast, HyperSuper or whatever. The names for the LoRa settings are already defined by the company who produces the LoRa chipsets we are using. Semtech. For example go to https://www.semtech.com/products/wireless-rf/lora-plus/lr2021 , press on on Documents and read the PDF file named LR2021 Datasheet. There are all the names for the different LoRa settings. And now from the usability point of view. Of course it is to expect that people opening up Columba would have never heard of a spreadingFactor. They want probably to have something that is just working. Something they can use to communicate with other people. So lets make this possible without using names. What is required for people to make it possible to communicate with each other? The LoRa settings have to be the same. How to reach this goal? By putting the world into LoRa regions and counting the already existing Reticulum LoRa device settings. Happily we have now public RNode LoRa settings: https://rmap.world/ For example scrape the data from the USA, and show as a possible preset something like for example: 100 people in the USA use SF7&&BW250000Hz&&CR5. 50 people in the USA use SF12&&BW125000Hz&&CR5 and so on. By doing so, you are removing stupid names. Instead of the names there are now statistical probabilities. If you are in the USA and want the highest chance to communicate with other people in this country, you choose the USA-preset 100 other people are using because this is the highest number == highest statistical probability to meet on this setting someone else. Yes, of course this 'preset' gets updated in Columba from time to time when the data from https://rmap.world/ changes. When from the example '50 people in the USA use SF12&&BW125000Hz&&CR5' the half of them move over to '100 people in the USA use SF7&&BW250000Hz&&CR5' after for example 6 months, then Columba updates it. The preset is then '125 people in the USA use SF7&&BW250000Hz&&CR5' '25 people in the USA use SF12&&BW125000Hz&&CR5' I vote to also add the frequency in the same statistical dropdown option. The Meshtastic names do not include frequency. Lets expect people have as less knowledge about what LoRa frequencies are used the same like what spreadingFactor is used. What Columba should expect from their users to know is the name of the country they are at 😆 Have i missed something important in the idea i explain here? — Reply to this email directly, view it on GitHub <#573>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHB66IU7BACRG2L5OTIZBTL4OCIXZAVCNFSM6AAAAACWCCLSB6VHI2DSMVQWIX3LMV43ERDJONRXK43TNFXW4OZZGU2DMOJXG4> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[serialrf433]

No, it is more about not using names instead of recommending anyone looking into what that partly closed source project Meshcore (by Ripple Radios) does.

Also telling the people to use 62.5 kHz when maybe everyone else around using RNode is using something else is not the way to go.

To use your words: Your 'out of the box' experience would be terrible if Columba is setting up some preset for you when everyone else around you in the region is using something else.

The Columba users are not the one making the move from LongFast to ShortSlow and trying to convince everyone else around them to do the same. The Columba users are the one joining the masses.

Columba already have a LoRa config implementation telling you what settings are more famous in specific regions.
Now adding the numbers from the https://rmap.world/ to know how famous a setting is in a specific country combined with removing the Meshtastic names is what i try to explain here.

[torlando-tech]

There is an old saying that has long resonated with me that goes something like, "to give something a name gives you power over it." I don't know who the original author is; a quick web search resulted in https://www.goodreads.com/quotes/7777162-once-you-can-name-something-you-re-conscious-of-it-you

I also do not think there should be something like a name for LoRa settings. Many people are putting in general discussions examples with other things in their life. So lets put here such a thing. For example 20°C&&55% humidity&&1000hPa = George.
Why should it be named George? Why should people talk about 'George'? This is stupid. Just name the data standing behind this name.

I am not sure if you are intentionally straw-manning with your "George" example, but either way you seem to place little value in names. That's okay. Like it or not, the general population (and perhaps even you, without realizing it) rely on names to easily and concisely communicate more complex ideas. I would argue this is one of the key foundational pillars of human language. I do not look in the pharmacy for 2-(4-(1-Methylpropyl)phenyl)propanoic acid, I look for Ibuprofen. When getting directions to a location, I do not look up GPS coordinates, I look up an address. When navigating to github, I use their domain name, not their IP address. (Even this example - I am using an acronym IP instead of Internet Protocol - another kind of name!)

Many people are putting in general discussions examples with other things in their life

Even you say "people" and not "homo sapiens sapiens." Hopefully you understand my point; naming things is hugely important for effective and efficient communication.

Even LXMF has a feature for "Display Name" because remembering and recognizing a destination hash is more difficult for most people than a simple display name.

So, aside from the argument about whether things should have names at all, let us move to the quality of the name.

=====

Note that I am not at all a radio expert, rather, I am a beginner. Please take everything after this statement with that context.

The Meshtastic naming scheme has two words, one to imply range, and the other to imply speed. Of course this will be an imperfect reduction of information, just like Ibuprofen is an imperfect reduction of 2-(4-(1-Methylpropyl)phenyl)propanoic acid, but it will be easier for people to remember and communicate about, and those who want to learn more can use it as an entry point to do so.

https://unsigned.io/understanding-lora-parameters/

Generally speaking, choosing a narrower bandwidth will result in a slower transfer rate, but improved range. Choosing a wider bandwidth will result in an increased data rate, but a shorter range.

The LoRa spreading factor is the parameter that controls how spread out in time each data bit is. If a higher spreading factor is selected, each payload data bit will be spread out over more symbols.

The coding rate describes the ratio of actual data to error-correcting data added. In choosing the LoRa coding rate, it is important to consider whether it's necessary to permanently employ a high coding rate, with subsequent loss of data rate, or whether it is more efficient over-all to allow for the occasional dropped packet due to interference.

Let's use the above calculator to understand why such names came to be in the first place. I am learning as I write this; I am not a radio expert, nor am I terribly familiar with the LoRa parameters.

We can see from the post above that higher bandwidth means higher data rate (colloquially called "speed") at the cost of lower range. Conversely, lower bandwidth means longer range but lower data rate.

Higher spreading factor means each data bit takes longer to transmit, but is more resistant to noise, because each data bit is spread across more LoRa symbols. Conversely, a low spreading factor spreads the data over less time, but is more susceptible to noise.

Finally, coding rate is somewhat similar; higher coding rate means more resistant to interference, at the cost of repeating some of the data you send for redundancy. Notably, this is the one parameter of the three that is not required to be identical between peers for communication to succeed.

So now, looking at the presets, generally speaking they are listed from shorter range and faster date rate to longer range and slower data rate. Going down the list from Short Turbo to Long Slow, the bandwidth starts high and gets lower, the spreading factor starts low and gets higher, and the coding rate starts low and gets higher. All of these names seem to be more or less reflective of the parameters they represent.

[torlando-tech]

Regarding your suggestion about rmap.world: I agree that sharing and crowd sourcing local RNode settings is the way to go. I only disagree on the data source. Rmap.world is a great project, but to pull this data into columba requires an HTTP connection. I did briefly discuss with the rmap.world dev the possibility of fetching lora data via reticulum. They put a nomadnet page up that hosts the list of nodes, but it is on some older version of the site or something, it's been a few months and I can't remember the specifics. It seemed like there were higher priorities for them at the moment, which is fine.

Since RNS 1.1 we have a native solution for this problem: interface discovery. I love the rmap.world interface and the project is great and should absolutely continue to do what it's doing, but the core weakness I see is that it relies on user submitted data, which can go stale as people turn their nodes off and never update rmap, and as mentioned, more or less requires pulling data via http. For the purposes of Columba being an easy way to discover RNode settings of their nearest geographical neighbors, I would instead lean on interface discovery. Perhaps I will actually encourage users in Columba itself to share their RNode settings via discovery with an opt-in option.

[serialrf433]

Rmap.world is a great project, but to pull this data into columba requires an HTTP connection.

I seem to not have been clear enough on that point.
The change in the example from:
100 people in the USA use SF7&&BW250000Hz&&CR5.
50 people in the USA use SF12&&BW125000Hz&&CR5

to

'125 people in the USA use SF7&&BW250000Hz&&CR5'
'25 people in the USA use SF12&&BW125000Hz&&CR5'

is not done by rnode.map HTTP connection. This is done by a Columba version upgrade.
The numbers how many people are using are a fixed value. Its just for orientation what is the biggest RF user group around you you would probably have the highest chance by joining with the same RF settings.

Yes, i know about the RNS 1.1 interface discovery. I also love the functionality to copy over the LoRa setting from some other RNode you have found based on the interface discovery functionality. This is great implementation!
I am talking here about the initial setup for people opening up Columba for the first time and having probably no experience otherwise. The most simple LoRa settings decision for such new users is probably choosing what is the most used settings in their region. Telling such users choosing one thing named LongFast is recommended (current implementation) and ShortSlow for example not the recommended setting is even more complex. And people start to think again about the preset-names and what they could maybe be telling them.

[serialrf433]

So, aside from the argument about whether things should have names at all, let us move to the quality of the name.

I really think putting a name into it is wrong because of the endless variable possible.
@torlando-tech

I am not sure if you are intentionally straw-manning with your "George" example

No. I tried to find a different abstraction using a name for something where it does not make sense to do so. I wrote it this way to try to make my point even more clear.

Higher spreading factor means each data bit takes longer to transmit, but is more resistant to noise, because each data bit is spread across more LoRa symbols.

Yes. In a RF chamber.

In a practical case i can tell you, that it is a common thing, to have often more success at a RF transmission when reducing the SF to a lower number.
Let me try to use here again a real world example (this is not straw-manning ! ). Think of the RF world to be somewhat like a road with cars. The cars are sometimes driving at fully random times, sometimes at a specific time for example once in the minute. The result is a chaos situation you have to be part of.
You and your friend, you are standing on both sides of the street. You want to tell your friend something and if a car drive by, its cutting your voice for a moment.
If you speak really slow(SF12), your friend can maybe have success thanks to the low speed to still understand what you told because not that much have been cut off.
If you speak really fast (SF5), you are transmitting much faster. If a loud car drive by, it is maybe removing 80% of what you have told. Impossible for your friend to understand you. But there is also a different option possible. Because you was speaking really fast, you was able to say 100% of what you liked to say in between of two cars. So there was nothing that have distracted your voice thanks to only using just the shortest possible time to transmit your voice to your friend.

And now i hope this makes is more clear: Put a name into such a thing. Try to explain this in a single word(name) to a new user who have no idea how RF is working and what they are doing when starting the first time with this new LoRa device.

And this is just a tiny, tiny example of a huge RF world with endless variations. This was just about the SF.

Bandwidth:
Take a look at things like this: https://github.com/orgs/meshtastic/discussions/448#discussioncomment-15561006

We can see from the post above that higher bandwidth means higher data rate (colloquially called "speed") at the cost of lower range. Conversely, lower bandwidth means longer range but lower data rate.

Yes. In a RF chamber.

To use again the street-example. Now you and your friend are in a car driving on the street. The street have lanes. Take a look at the gqrx SDR screenshot. You see the LM1 and LM2 numbers. If you make the cuts even further apart to leave out all of the 'busy lane' in the middle of the picture (yellow thick line) and use only the free lane (blue), then you know what lane on the street is not clogged up and can still be used for transmission/driving.

Now again: What name do you want to add to this where you have to take a SDR and check what part of the RF spectrum is still free to use?
You get it? Names are fully wrong.

Same like names for temperature+humidity+air-pressure combinations in my George-example.
Those are just different things that have different functionality. There is no name possible and putting in some name is just bad. If it would have made sense to group fully different RF things together, why is Semtech as a LoRa manufacturer not doing so? Because it is not possible, it would be terrible to do so and much more.

[serialrf433]

I do not look in the pharmacy for 2-(4-(1-Methylpropyl)phenyl)propanoic acid, I look for Ibuprofen.

Ibuprofen is a https://en.wikipedia.org/wiki/International_nonproprietary_name . It does not have any values people could understand wrong in it. It is not named 'fast' and also not 'slow'. If you say the one or the other 'name', it does not contain any different information.

Even you say "people" and not "homo sapiens sapiens."

It is homo sapiens. Not homo sapiens sapiens. https://en.wikipedia.org/wiki/Human
And people is something different then human. See in here https://en.wikipedia.org/wiki/Person

I do not look up GPS coordinates, I look up an address

There are areas in the world that have the same street name few times in one city. This is confusing. It is not exactly clear to the user looking up things. The people have to waste time asking things like 'The address next to the supermarket?'.
Also an address is often confusing because one huge place have one address. When you want to meet at such a address, you could waste your time searching each other when you are there.
Simple solution to all those issues: Using GNSS coordinates. You exactly know where you have to stand.

There is an old saying that has long resonated with me that goes something like, "to give something a name gives you power over it."

This is historic and bad. People are giving names to things they do not understand. This is a known issue for more then 3000 years. The stupidity with fantasy inventions like https://en.wikipedia.org/wiki/God and all the names people gave those fantasy figures is a good example for this.
To name one example: There is electricity, heat and sound waves. Not things like https://en.wikipedia.org/wiki/List_of_thunder_deities

I hope to change the mind of some people not to just name things because of wanting to put names on things that have already clear, technical values.

So, aside from the argument about whether things should have names at all, let us move to the quality of the name.

I hope there wont be any additional names. For example SpreadingFactor is SpreadingFactor (SF). This is the name of it. It does not need a different name. This setting already have its name. Just use it. Do not make the same mistake Meshtastic did with adding confusing names to things that have already clear and exact names.