And Sage has been (and is!) huge undertaking, which developments of the project surprised you most? Which parts of it you like most and which least?
In the name of many, many users thank you for starting it!
> which developments of the project surprised you most?
It was dramatically more difficult to get real funding than I expected. I remember my first NSF grant asking for funding directly for Sage, where I put in what I thought was a humble request for about 10% of what Magma operates with... and they cut everything except one single 50% postdoc. I definitely thought the math community would support development of a community owned platform for computational mathematics much more strongly, and was extremely surprised when that was not the case. Europe has been much more supportive (with OpenDreamKit), but I'm in the US.
That said, I'm also very pleasantly surprised that actual contributions to Sage have continued and grown in many areas of mathematics, and the project is successful and healthy today and supporting research level mathematics in many areas.