[codesections]

> They explicitly call out the summer [and winter] holidays effect above the first graph

Given that this is a measure of "Desktop clients active in the past 28 days", that means that there are people tech-savy enough to be regular users of Firefox on a desktop – but who go an entire month without using Firefox because it's Christmas. And enough of those people to show up on this graph.

That blows my mind a bit. I don't doubt it, but wow.

[zamadatix]

Remember FF has been bundled in installers and OEMs have preinstalled it. Out of 200,000,000 users there are probably more users that simply aren't tech-savvy enough to have changed to a different browser than are still on it because they are tech-savvy.