He claims to have started his career by writing Altivec optimisations, being spotted by Gentoo developers and asked by them to apply his skills at certain opensource projects… Originally he worked on MPlayer but eventually, in 2005, he turned his attention to FFmpeg as well.
His main interests in the project were miscellaneous small patches and bits of code, Altivec optimisations (to the point that he bought POWER9-based workstation recently to keep playing with it even if for the different projects), networking protocols, capture devices and hardware-accelerated decoding. He also was a server administrator among other thing and had other exotic projects like libav and VideoLAN chocolate.
As a person he’s not a bad guy: he invited various people to stay at his place in Torino countless times (even I visited him there twice), he brought lots of good (even if Italian) chocolate to various conferences and helped other people. But he has two features, a good and a bad one: he finds people and gets them involved in various projects. And his second character feature is that he stops working on those projects himself, often starting a new and exciting project as an excuse.
For example, he found a student for creating a libswscale replacement in the course of Summer of Code program. There were many people thinking this is a good idea but Michael Niedermayer didn’t so he essentially sabotaged it at the review stage and the student left the project with a feel of disappointment in the project and disliking of Carl Eugen Hoyos (who wrote him a nice private mail).
Or later, during libav times, it was still considered a good idea to replace libswscale. As I mentioned before, Ronald had an idea of AVScale that he did not deliver before leaving for FFmpeg. So Luca decided to resurrect the idea and tried to create a design for this new library (I contributed a bit to it and wrote some prototype code as well). He got Vittorio Giovara involved as well and they worked a bit on integrating my prototype code in libav and documenting it (presumably in drunk state) but nothing serious came out of it. Later though I salvaged the design as NAScale for NihAV.
His next project, rust-av, went the same route. Initially he managed to convince different people to join to help him create a nice multimedia framework in Rust and virtually abandoned it at some point (and don’t get me started on Opus decoder). So I decided to make my own project instead (aptly named NihAV) and in some areas it offers better functionality already (for example, supporting more formats and neither of them via an external library).
So if you ever get by him in Torino you can expect him to treat you with some nice ice cream or chocolate but if he asks you if you want to participate in an interesting project just say no.
I’m guilty of spreading myself too thin at times, I agree.