anyone use datalevin for implementing ecs 👀 or is there a better alternative

question about backups + DB dumps: is there any downside to invoking the functions defined in datalevin.main from a deps.edn alias instead of the standalone CLI? I know I need to add a few :extra-deps to load that namespace, but I figured this would be an effective way of invoking the copy and dump commands in a more platform-independent manner that ensures the backup always reflects the version of Datalevin in deps.edn instead of whatever version of dtlv I happen to have installed. Any pitfalls to watch out for with this approach?

DB version upgrade question: I have some values that are encoded with non-default tagged literals in my DB (this may not be a great idea in general but for now I have to live with that decision). When trying to read those back in with datalevin.main/load I get a parse error. For this use case, would it be advisable to make datalevin.main/data-readers dynamic so users can rebind it like *data-readers* in other contexts? Or should I just do some post-processing on the decanted DB after using datalevin.main/dump to clean up this data before I import it to another Datalevin DB?