| I'm getting serious "Building Dropbox is trivial with with curlftpfs, SVN and CVS"[0] vibes from this thread. Sure, all it provides is already technically possible with what exists - but what if the 10% they shave off turns out to be the crucial 10% that carve out a completely new area in programming? Maybe having a fully integrated development experience actually is the "retrospectively obvious" missing thing... That said, their exclusivity-first approach whose first step essentially is full and complete vendor lock-in makes me skeptical too. I personally don't think they'll be too successful, as the model they're proposing is both uninviting to newcomers and fraught with sustainability perils for any serious long-term project. And that's just from a business perspective, I can't overstate how important I consider open technologies to be. But I can't get rid of the "future of programming" tingling, which a few people here have alluded to as well. The "open-decentralized-interoperable" bazaar and the "monolithic-integrated" cathedral conflict is as old as time. And while zero-friction infrastructure is a nice selling point, I'm personally more interested in what's allowed by having the language, editor, and apparently the entire ecosystem be developed hand-in-hand from the first step - a lot of fancier features (smart code completion, thorough static analysis, code exploration, automatic diagramming, proper e2e testing) is hard specifically because it needs to built on top of what already exists. Having a garden with walls this high might actually bring a lot of surprises. 0: https://news.ycombinator.com/item?id=8863 |