Fine effort indeed. But do not forget, he had to do all that heavy lifting just to fit into google's corporate policies. To me JavaScript makes very little sense because I like RoR mostly for the first 'R'.

Putting Ruby aside, I don't see much of "rapid" or "pragmatic" (most hyped words) in Rails. I am coming from ASP.NET and I don't see "Rails" part as a major step forward. The need to manually implement most of user logins/accounts/sessions in Rails alone killed all the time savings I had provided by ActiveRecord.

Frankly, I don't see why everybody is so hyper on MVC. I has been around for ages, old&stinky MFC in 1991 had doc/views but nobody calls MFC cool :-) Is that because more and more of web development is done by people without CS degrees who had never heard of MVC before?

My son, let me ease your pain:

1) gem install login_generator

2) script/generate login

3) before_filter :login_required, :except => [:login, :signup]

4) session[key] = value

Every once in a while someone manages to create an especially great implementation of some technology or methodology that really shows its power. That's what RoR has done.

s/MVC/OO/g; s/MVC/FP/g; s/MVC/XP/g; etc.

"Is that because more and more of web development is done by people without CS degrees who had never heard of MVC before?"

Is that jab really necessary? There are very distinct advantages to getting a CS degree, for certain. But, knowing what MVC is/was and having a CS degree have very little in common.

And what would this NBE thing be? Next Big Environment? Next Big Emacs mode? Next Big Effort? Any guesses?

Next Big Email...app? Is he going to one-up PB and crush Xobni? :-)

The lack of any big improvements to Gmail makes that seem like an actual possibility.

I know! It would be the Next Big Editor! He is going to port Emacs to JavaScript and have it work in a browser. Or rather it would be the browser. Whatever. But mark my words, you heard it here for a first time: the Next Big Editor is going to be programmed in JavaScript :)

Sorry to be a party pooper, but someone beat you to that expansion on his blog comments. He later disabled comments though due to spam issues.

Oh, yeah? What a competition! But that's life, no big deal, and I was a bit satirical anyway. :)

Wow, what an effort. I'd like to think he could have taken one [Pylons | TurboGears | Django] and reworked them into something closer than Rails, but to each his own.

He seems to have preferred the 'momentum' and quality of the Rhino codebase over Jython - although he admits Jython might have seemed like the right choice at first glance. See the section about 3/5ths down.

I'd argue that there are at least two other MORE important factors: JS (IMHO) is somewhat closer to Ruby, compared to Python. In fact I suspect that his "secret 20% project" is Ruby to JS compiler (or vice versa)

Also, (most importantly) using the same language for server AND client development is big-big win. He's been dreaming of a universal "web platform" which allows you to have real development process: pieces of you app here and there, and you freely move them around, executing server or client side, whatever works best for you at the moment. Because right now you have this huge stack of things to keep bookeeping for: HTML/CSS/JS/Ruby/SQL, all sitting in their own little files with their own little syntax. At least he got last 3 merged into one solid JS framework.

Next Big [Development] Environment? Okay, I'll stop.

The wrinkle is that this project had to run on the Java Virtual Machine. The only reasonably good Python interpreter for the JVM is Jython, which is significantly out of date. He'd have to port Python to to the JVM before porting Rails to Python.