Thanks @jorainer.bsky.social , that is the culmination of twenty wonderful years at @ipbhalle.bsky.social !

matchms is a wonderful fit in the microservice architecture of the MassBank software rewrite. We even have a REST interface to query MassBank programmatically.

Data has a life-cycle, and justifies that it is maintained (like software), rather than just dumped when the paper comes out.
P.S.: I think mastodon allows longer messages ;-)

2) Instead of downloading data from somewhere, massaging/fixing stuff, and doing your own analysis, get the word back upstream what didn't work with the original data ! Have the tool / data people fix things upstream, so that the next person has less pain than you did, converging on interoperability

Prepare your data for submission right away! No need to publish same time, but if your analysis is based on the same information that'll be published later, you ensure that it works for one person (you!), and certainly that'll be interoperable for others as well.

Hi, two ideas: 1) if it is about your own experimental data, be your first re-user ! Instead of doing the experiment, organising and massaging (format wise!) your data so you can analyse it, and eventually publish the data in a repository like MetaboLights, Metabolomics Workbench, ...