cyberp

I'm in a similar space some of the time, so I thought I'd look for good ideas. Review:

* Poor idea -- misses out on issues of motivation, and the soft sides of how people influence each other.

* Isn't great for hard side either. Most good arguments are too nuanced for soundbites.

* Indeed, it would be interesting to see the argument for Reasonwell (over github and similar) presented in Reasonwell's format, and then used as a discussion for how such a system should be designed. This would be a good first use case, and of similar complexity to most political debates. My claim is it would be difficult or impossible to do this in their format.

* The implementation doesn't work (dialog pops up when you try to post a comment, and "okay" doesn't work).

* UX is unclear (I couldn't figure out how to start a debate after minutes of trying, or how to search in useful ways, etc.).

* System isn't open source/free software. This is important for a tool for democracy.

Overall, good ideas were scarce.

Yup. Both Berkeley and Stanford -- Coursera's first schools -- are fleeing for the hills.

I stopped booking with Expedia when there was a problem with an incorrect flight time. Delta said they couldn't do anything and I'd have to resolve with Expedia. Expedia said they couldn't do anything, and I'd have to resolve with Delta. I had a clearly nonsensical flight plan (I was scheduled to be at the destination for 1 hour), and no way to resolve it. I gave up on both businesses.

The iPhone popularized a lot of best-of-breed technologies. There were start-ups trying to do what the iPhone did. Apple is claiming those as Apple innovations. This is very similar to Apple popularizing the GUI invented at Xerox, and suing Microsoft.

Posted at the article, but the blog is censored ("Comment is awaiting moderation").

On a high level, Java vs Ruby/Python is a question of suitability to purpose.

Java is a great language for large teams of incompetent programmers. You will not be able to find competent programmers to code up the business logic for: Paying with a credit card. Going to the emergency room. Adjusting your 401k. Using your insurance card at the dentist. Filling a prescription.

That code is mindnumbingly boring, and has few technical challenges beyond sheer size and complexity.

With incompetent programmers, you want a B&D language -- Java forces object oriented. It forces a single object oriented paradigm. It forces a set of standard libraries. Etc. It makes it as hard as possible for programmers to shoot themselves in the foot, and it minimizes learning curve to get up to speed, making programmers as interchangeable as possible. If you've got a team of half-wits trying to write an employee management system, that's exactly what you want. The problem falls well within the limitations of the system. If you want to do something more complex -- a startup solving a problem that's never been solved before -- and you have a team of hotshot programmers, you really do want a less constricting and more flexible language.

COBOL is obsolete. Java vs. Python/Ruby vs. LISP/Scheme is a question of suitability to purpose -- primarily problem difficulty and team composition.