Mojo support #1434
Replies: 1 comment 1 reply
-
Honestly - I have no idea what to make of Mojo at this point. A well funded startup has announced a new programming language that is "like Python but faster and typed". It's not open sourced (but they promise it will be! Super pinky swear!), and there's no way to try it out beyond a sandbox. But it's totally going to solve all your problems! At best, I'm sceptical. Projects that are funded by a single VC-backed company have a nasty habit of disappearing once the company behind them runs out of cash (witness the recent mess surrounding Gatsby and Terraform, to pick 2 examples). Unless Mojo is actually open sourced, and able to develop a significant developer and maintainer community outside of Modular as a company - and fast - it's hard to see how it will survive. However, assuming Mojo materializes in a meaningful way - what will BeeWare's relationship with Mojo be? It's literally impossible to tell at this point in time. They appear to be targeting the Python community with their messaging - but they've also been pretty explicit that Mojo isn't a language superset, so the migration or compatibility path for Python tools into the Mojo ecosystem is completely opaque. If it turns out there's a way for Mojo to use BeeWare tooling, or to be compatible with BeeWare tooling, then I'm not fundamentally opposed to supporting that use case. However, as with most suggestions regarding "edge use cases", my preference is for any support to start life as an externally managed plugin, rather than as direct features in the core tool. If it turns out that a BeeWare tool can't support a required mode of extension, adding plugin hooks to allow extension is more likely than merging explicit support for a third-party tool. |
Beta Was this translation helpful? Give feedback.
-
Hi. I'm a big fan of type safe code that runs reasonable fast, that's why I always try to pick something native and would pick TS over JS if I need JS. That's also why I never consider Python for larger projects, although I think Python has by far the best syntax.
However, things are going to change with Mojo. Here is a good introduction video which is short and to the point: https://www.youtube.com/watch?v=V4gGJ7XXlC0
For a long while I'm also looking for a tech stack that is able to target all platforms without any WebView or Canvas trickery. I want a DOM on web targets, I want true native UI on mobile and desktop apps. And by all means, I don't want things like React, otherwise I would probably stick with React Native.
BeeWare seems to check all marks if it would run on a more performant interpreter that has better performance and type safety. Mojo seem to bring that opportunity. There will be interpreters for desktops, although I don't know if there ever will be any for mobile.
I know there is probably a long way to go. After Mojo is not even publicly released yet. APIs will change, stuff will be unstable or break. There is a long road ahead.
I'm opening this discussion to ask for first impressions by the team behind BeeWare. It's certainly to early to make a decision on that, but I hope it will still bring some interest to you, I hope you are as excited as I am about Mojo and what it can bring to the table.
Beta Was this translation helpful? Give feedback.
All reactions