Gregory's ServerOriginally, my goal with Pinafore was to build what I thought was the "ideal" Mastodon UI from a technical perspective. Fast, accessible, offline-first, single-column, true virtual list, etc. Since then, the Mastodon frontend has caught up in a lot of ways (e.g. single column), so these things are less of a competitive advantage.
|
4 comments
 @Gargron I maybe overstated my case. A lot of it is not Mastodon's fault but just how hard it is to write offline-first apps in general. When I worked on PouchDB, we had a whole sync protocol between the client and server that dealt with versioning, conflicts, etc. You just called `db.sync()` and you were done. I don't expect Mastodon to add this, but it's just an example of the kind of thing I miss from the PouchDB world.  |
I also saw it as a way to push the limits of what the web could do. However, I've become a bit chastened and disillusioned about this. For instance:
- Pinafore doesn't have nice animated transitions like native apps do. The web still sucks at this.
- Offline support is almost impossibly hard given Mastodon's API. The main benefit of IDB right now is low memory usage.
- PWAs haven't come far enough. Pinafore isn't listed in app stores, so it "doesn't exist" for most users.