In R, we have assets_version() that uses the default (hard-coded) assets version but also consults the SHINYLIVE_ASSETS_VERSION envvar. We should replicate that with py-shinylive. Doing so makes it easier to move between asset versions, improves testing and decouples py-shinylive package releases from the shinylive releases (which do need to be updated with py-shiny releases).
Note that not all assets versions will be compatible with all shinylive versions.
In R, we have
assets_version()that uses the default (hard-coded) assets version but also consults theSHINYLIVE_ASSETS_VERSIONenvvar. We should replicate that with py-shinylive. Doing so makes it easier to move between asset versions, improves testing and decouples py-shinylive package releases from the shinylive releases (which do need to be updated with py-shiny releases).Note that not all assets versions will be compatible with all shinylive versions.