I have one last question regarding backports: we tried the starter frontend excessively over the past weeks. So let assume we have 2.000 taxons and 50.000 products. So we loaded everything up into the staging system and provided the staging with a caching system (Redis) as we have excess in ram and sas SSD, but no NVME, so solid cache was not really an option. What we noticed: • removing /t/ and /products/ was no measurable performance issue as long as resources with few items (categories, pages, system pages) were tried before products • if you apply caching (which I think we all do) the current construction of the routes seem to be based on historic performance issues that are largely solved • A shorter URL does provide seo benefits It might be a reasonable default for the starter to provide all resources starting from / and not from /t/ for taxons or /products/ and carefully document it. I do know that Shopify and Airbnb make excessive use of folders such as /collections/ or /appartments/ but I do not think this is a default but an outlier. I’d like to know what you think working on the field and if this has ever been measured by somebody. We are considering throwing the route model of the starter away as we feel it’s a relic for most stores and analyzing also other pages using solidus but what we see on the frontend they seem to agree. I’d strongly suggest to make this a 2.0 as it’s breaking for the SEO of present stores and would require massive amounts of 301s, but I could see this being combined with bringing a strong redirection solution into the core or at least make redirector in a better way and bundle it with the schema for the starter. If that’s of interest we can provide some data from the staging system and a demo with a fully fledged catalogue and you try it out yourself end of next week.