I’m looking to commission a developer from the community to create a native Sailfish OS application for me. I’m using an Xperia 10 V with Sailfish OS, and I’m very happy with the OS, but I’ve encountered a specific app-related issue that I’m hoping to solve with your help.
The Problem
I need to monitor a Battery Management System (BMS) for my e-bike, which is usually done via a dedicated Android app called AntBMS. Unfortunately, this Android app does not work correctly under the Android AppSupport layer on Sailfish OS due to its need for a specialized Bluetooth connection to the hardware. The connection to the BMS is critical for me.
My Request
I would like to commission a developer to write a native Sailfish OS application that can connect to and monitor my BMS via Bluetooth. The new app should replicate the basic functionality of the original Android app, specifically the ability to:
Connect to the BMS via Bluetooth.
Display key battery parameters (e.g., voltage, temperature, state of charge).
Perform basic control functions if possible.
I am aware that this is a complex project involving specialized Bluetooth protocols and would require significant development time.
I am ready to pay a fair price for the development of this application.
If you are an experienced Sailfish OS developer and this project sounds interesting to you, please reply to this thread or send me a private message to discuss it further.
in my opinion, your path forward here has three distinct steps:
get your battery working with antbms over bluetoothet/uart with open drivers/code/protocols/etc on a standard environment (say, on a linux laptop or a raspi)
getting that reverse-engineered solution to work on SFOS on the Xperia 10 V
make a quick little GUI app that wraps it up
step1, the reverse engineering some closed device protocol to FOSS land, is normally the hard part, but it seems like it might be done already, based on a quick google search that turned up a few repos like this:
you should see if you can get that working for your specific device.
step2 is probably either easy or nearly impossible, depending on if you need support from a custom linux kernel, or specific bluetooth hardware or a specific version of bluetooth driver or something.
note that step1+2 require having your device on hand.
step3 is easy and just about anyone can do it, after step1+2. it is fairly common for non-developers to only picture this last step, however, since the gui is all they interact with, and so i worry that maybe step3 is the only one you really have in mind when you talk about paying someone to develop an ‘app’, rather than paying someone to reverse engineer a proprietary device they don’t have and then get it to work on sailfish.
does this work over classic bluetooth or ble? i suggest using android to gather bt logs, which can be analysed in wireshark. of its simple ble comms, my library qble should be able to help, its what powere amazfish and some other bt apps on sailfish.
For this and feature posts like these which want to pay “fair price” without specifying, or like the last guy who wanted to pay 50$. Let’s do little estimation:
lets take quite low wage: 13€ per hour
now making app is quite time consuming depending on info available and how difficult it is. Lets take the decibici for example. I can’t really remember how long it took but lets say 50h
that’s 650€.
And hobbyist, since they already got day job and the last thing they want is more job, will want higher wage at least 20€/h
that’s 1000€
Now that should put in perspective how much developer time costs.
It also should remind you to be nice to developers who make apps for free and spend HUNDREDS of hours
pluss getting the developer the bike/device, it can be as easy as biking across town (if you are lucky) or sending it by mail or just pay for a used one.
correct, but while it is widely accepted here to pay that for cleaning and twice as much for a car mechanic (all black money of course)
lets try to request the amount when the neighbour has an issue with his whatever (pc related) ..
IMHO we have conflict of interests here: On the one side: No one should work for free, because no one wants this. I also would not work for free.
On the other side, No one would buy an app for 1000 $, when you are private person and dont earn money with this.
I am very thankful for @Mister_Magister for creating decibici app, which is used by exactly 2 persons. If i wanted to pay @Mister_Magister fair - it would be more than 1000 $, which no normal user could pay.
So yes we should be thankful for the developers who write apps for free. But we have no possibility to pay them fair…
I know nothing of coding but if you have spare time maybe try to make the app with AI. I know not everyone support this approach but it might be possible. I have read at least some experiences from this in this forum
For anyone who tries this, please work with your chosen AI until you understand the code, it builds cleanly, and the app works as intended. (Other developers may be willing to assist, but will very likely be annoyed if they find themselves talking with an AI-by-proxy rather than a learning human being.)
No offence, but Mr Magister who is experienced, tells you how much it takes him.
And your best idea is: AI can do that.
Download SDK, download an example app and make it buildable and install it on your phone.
I bet it takes you at least 10 hrs.
Just to build someone elses work
Well as I understood Mr Magister probably won’t do this application at this point of time without OP digging deep from pocket. If OP can’t find any developer do the job with price he/she is willing to pay and has spare free time I just suggested to try AI route. As I said, it might work which also means it might not. Never said that AI can definitely do it. Others have done apps using AI so in some cases it has been possible. But no delusion here that it would still need your time and won’t happen in one evening. And as jojomen said, it would be good anyhow to learn understand the code. AI just makes the learning curve for coding much easier for total noobs like myself. Barrier to start lowers significantly. If you are willing to learn on the way, don’t really see the problem. You can ask stupid questions even without AI.
Let’s be honest its time consuming yes, but its quite simple, well, from my pov. Thats why i did the decibici app, to have introduction into bluetooth so If i were to do it second time it would be much shorter. But point being, the app is simple to make, and as much as i despise to the core entire idea of using LLM to write code, you could literally just copy paste decibici and probably come up with something sensible
