I don't think your work is entirely useless here, under the assumption that the tool is deprecated but nobody has discontinued using it yet. Even deprecated too

I don't think your work is entirely useless here, under the assumption that the tool is deprecated but nobody has discontinued using it yet. Even deprecated tools need some maintenance from time to time, and in a big organization there's lots of legacy stuff that's kept around for a long time because replacing it is a lower priority than other roadmap items. For a useless/meaningless project I did once, I wrote the read and announce portions of a userspace BLE stack against the BlueZ sockets in Linux because it was quicker and easier than figuring out how to use the DBus interface. It also turned out to be about 10 times faster latency wise. We needed it to do a proof of concept for something which later didn't pan out in the market. The code still lives in our repository and gets compiled into our products but nobody plugs the Bluetooth dongle that triggers its activation into any fielded product.