Yes, Jon, there are many things to improve in Marvin – and in its competitors, too. We may not agree on the particulars – which missing or misbehaving features should have priority – but I think most of us would agree there are many things to improve in Marvin. The dozens (hundreds?) of user requests listed in
Marvin's GitHub speak volumes.
The trouble is, given the rate of Marvin's development from the last few years, these things will get fixed/improved in 10 years, 20 years, or never. (I use certain other pieces of software in which some bugs have been unresolved since the late 1990s, so I'm not exaggerating.) This is not meant to criticize Kris – we fully understand if he now has other priorities in life than developing Marvin at the frantic rate of the Marvin 2 era (especially when the financial rewards aren't what they should be). But it's a statement of fact that these improvements are unlikely to occur anytime soon if Marvin's development is to continue at the rate from the last few years (minus the Marvin 3 intermezzo).