Academic software has a bad reputation for being poorly maintained. We discuss our experiences maintaining our own bioinformatics tools and the systemic issues that influence the quality of academic software. Topics * Lessons we've learned from maintaining our 8+ year old tools: Ginkgo, Assemblytics, SplitThreader, Ribbon. * Discussing career implications and systemic incentives of maintaining academic software. * How we build bioinformatics web apps today versus when we first started out. * Supporting users when you're not even in that field anymore. * Lots of embarrassing stories of bad software engineering in bioinformatics tools. Join the discussion in the comments on YouTube. Questions from the episode: * For academics, do any other metrics get considered for your tools other than citations? * How do you cite tools you used that you didn't mention in the methods, e.g. samtools, parallel, IGV. * Do any programs match up volunteer experienced software engineers with scientists writing software?