ANDREW OLIVER

Let's be clear, I don't think .NET is dead yet. It will continue to be the Windows API of the present so long as people write only-Windows apps, it will be around.

Agree that marketwise (I even mention that in the article), I'm strictly speaking technically.

The main story is linked. This is my (the author)'s blog. A lot of time I put extra content in my postings. Also Disqus sometimes doesn't work so comments can be left on my blog if it is broken. I agree there aren't a lot of extras (except a line at the the bottom) on today's posts, but often there are :-)

you hope? ;-)

This is better as an indictment of fanboy marketing than MongoDB. Is the document db for reporting? No.

There is also a fundamental difference between a single developer project and a multi-developer project.

I generally post an excerpt of my Infoworld articles on my blog often with additional content such as pictures, etc. The excerpt links to the rest of the story at Infoworld. However often the pictures and other content are only on my site. (i.e. Infoworld wouldn't publish a picture of Gavin King as shrek even if Gavin approves)

I actually acknowledged this overall in the article. Given how horrible their android app is, it may even be the right decision. However, it is a pretty temporary thing. Do you REALLY see coding in native compiled Objective-C just for one mobile device in the face of an acceptable alternative?

I agree completely with regards to DSLs. I made this very point in the article. There is a difference between functional programming being good for specific cases and "let's rewrite our corporate presence site in Erlang because it is beautiful" (to take an extreme case). Good software architecture takes a number of things in mind, not the least of which are economics, complexity and field of use.

excellent. Glad to hear it!

I agree that Rails is not great to be honest. I make mention that I'm not a big Rails fan. In general I haven't found anything much better than JSON for the front end and "whatever you like" on the back end that emits JSON. Node.js appeals to me for the hope of a unified language. Because dealing with the costs of a project are a key part of my job responsibilities, I'm more dubious of Scala and Erlang. Most business apps are not rocket science, so really I need a language that lets me utilize junior developers for the bulk of the project and senior developers only for the heavy lifting.

I think you hit the nail on the head. I don't think these are logical or quantifiable arguments. There are certainly things that make me individually more productive with Ruby than with other languages (such as Java), but I can't say these make a huge difference on the project outcome when it is a more than single developer project. In fairness I think it is a "growth" thing. Now that Ruby is used and proposed on larger, more serious projects -- it needs to make a larger and more serious argument. When it was used on more 1 man projects "you don't have some horribly cumbersome packaging system and can 'edit in place' if needed" -- is a decent argument.