Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I can only repeat: Awesome!

I hope the FF guys will work with you to ensure that there will be only one Tornado in the future and not two independent forks.

It always hurts to throw away code that you put your blood and sweat in (in this case their custom network layer). But the advantages of rebasing this onto twisted are overwhelming for both sides.



It's not obvious to me that this is a win for Tornado users.

There's a cost to adding a dependency on Twisted. It's one more thing to download. There's more code to understand when something goes wrong. And so on.

What are the benefits to me as a web programmer? If I understand correctly, my application will work as it did before. If that's true, then Twisted did not add any win for me.


That's a very negative way to look at it.

The benefits you get are seamless integration with all the other twisted tools and a solid foundation under the relatively new Tornado stack.

And a bug in twisted will raise many more eyebrows than a bug in Tornado - hence they are more likely to be discovered by someone else before you.


  There's more code to understand
No it is not, he stated that overall he removed 1200+ SLOC.


And added a 150,000-line dependency.


Which has better test coverage and more users than the Tornado stack. Honestly code quality would be the least of my concerns with twisted. If there's a python project where the critical parts can be labeled "mature" then twisted is it.

I criticize twisted for usability and documentation problems. But as the underpinning for something else (with hopefully better usability) it seems perfectly adequate.




Consider applying for YC's Winter 2027 batch! Applications are open till November 2.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: