I’m going to use three fields as an example: computer science, biology and paleontology. How do keep with interesting developments in each of these fields? Which journals are the most reputable in their respective fields? How do I know that the journal is reputable? If the journal is reputable, how do I know that a particular paper isn’t a fraudulent paper that managed to slip through the cracks?
I like QuiteRSS (if you use Linux or BSD, it should be in your package manager). It’s super-easy to sort and manage tonnes of feeds, it’s very clear what is read and what’s unread, and you can set it to never delete unread items. It hasn’t been updated in ages, but it still works (mostly. Once in a while a Qt update breaks it until someone can fix the packaging).
My second favourite isAkregator. It’s almost as easy to use as QuiteRSS, but unread items aren’t as distinctive, and sometimes an unread item looks read.