Open source, but to what extent?

You've installed Linux on your laptop, may have an Android phone, use a Raspberry Pi for home automation and Arduino to do hardware experiments. All of these are 'open source' as used in the common sense of the word - you're happy while you do it and maybe also enthusiastically tell your friends about it. But are they really?

This talk is supposed to drill down into day-to-day technology components that we use and apply the "open source" definition at each level of abstraction that you know or are aware of until it either breaks down or stands to the test. By the end of the talk an average person should walk away with the understanding or appreciation of where we are in terms of what open source means for a piece of software or hardware and how it's a spectrum instead of something binary.

Common examples: you may use Linux but as soon as you have a NVIDIA card / want to use fonts and third party codecs for media [which are technically patented], you may lose some 'open source' points. Your Android phone may be running a custom Linux kernel customised by the manufacturer/OEM who may have introduced additional hardware support by adding drivers not in the version of Linux that's available on Git. The Android OS may be open source for OEMs, but a large portion of app functionality also relies on proprietary Google Play services APIs.

[not providing citations because this is an extremely last minute submission, but you can verify the facts if you need to]

Even the Arduino codebase and IDE is open source [and the code that you write with it to do something is your choice to be open sourced], but the chip on which the code executes, isn't.

Additionally we have "open source" LLMs today - but are they really, if you do not publish the training or the fine tuning set and the recipe to train it?

Does all this means we start calling things open source closed? No. I think it's important for us to develop a nuance of 'until when is something open source' and foster a discussion around this topic. Being able to take away binary yes/no to something being open source and showing it for spectrum that it is will enable people to make better decisions about open source and appreciate the nuances it takes for something to be open to the very core.

To throw light on the fact that it's important to understand the extent to which something is open sourced. Sometimes, especially in hardware, things are not that way.

Spreading awareness about nuances in calling something open source, ideology vs practicality - the fact that in real world open source is a spectrum. The software that we might be running at higher levels is open sourced and well developed but shedding light on the fact that it's still running on a foundation of proprietary components or 'blobs' at the low levels.

Note: I have applied for the main track because I personally feel everyone will benefit from this talk but if organisers feel that this talk is better suited for the policy or the hardware devrooms, please feel free to move the talk into a suitable track.