werg

Yes, problems abound :)

1. In fact, there is syntax highlighting for all the C-like languages, it's the mode called 'clike'. I guess that label isn't so very fortunate 2. Oh actually, the underlining has been added by me as a means to do author attribution. I should add a checkbox to turn it on and off. Before I had had the text background in various colors but I found that to be an even worse distraction. 3. The keyboard shortcuts are tricky. I'll have to delve into that a bit more (and test across platforms). 4. So obviously for now it's just security-by-obscurity. For now you can take the sourcecode and host it somewhere private. I left out explicit user account management in order to get the first version out the door. 5. Hmm. that's a very interesting thought! It should be fairly easy to build a small command-line tool to upload and download files.

Thanks for the feedback!

So I started building this as a coding exercise for an interview, I'd particularly be interested in getting feedback on my approach to Operational Transformation: http://gpickard.wordpress.com/2012/02/17/my-approach-to-oper...

It would be supercool to see something like this using websockets.

"Yahoo has no vision. It has no purpose. It's dispensable."

Facebook may have no vision. (I actually think they do. Even if it's not very clear or savory.) But facebook certainly isn't dispensable. The analogy would be much more complete if you imagined that Yahoo was the only main email provider, emails didn't get exchanged between servers and in order to get an email to yahoo mail users you had to go to Yahoo to mail them. That's about how dispensable Facebook is to a lot of people's private lives.

It's not their fault that this happened. But it is their business. So if this turns off customers, because they fumble this rather expectable occurence, then yeah - that is their fault. If you run your own business there just are no excuses. That's the flip-side to not having to put up with a lot of BS.

I'd agree with the gravity of the problems described, but I sincerely disagree with the almost-inevitable turn to optimism towards the end. "Gilding insists he is a realist". Straightfacedly Friedman hands on the assurance that given the choice to trod on as always or change, we'll suddenly wake up and give up greed, sloth, consumerism for some happiness-based economy (which is bullshit, in a catatonic collapse nobody's happy, no one except a few mavericks I guess).

How is that supposed to happen? I get it, miraculously, so we don't have to get off our precious asses now and can wait it out till the crisis sweeps us off into some kind of all-together-now rapture.

Already the ice-caps are melting off, within my lifetime world population has doubled, food-prices are rising, life for very many people in this world is starting to get more difficult, bit by bit. Yes, there is technological advancement, yet the price of energy has steadily been rising, at the same time fundamental world problems are on the rise as well. It does not suffice to assume that we invariably will find technologies in time to mitigate all these problems. We are in for trouble and very many people will feel the consequences.

How harsh they'll be is hard to tell in advance, but there's reasonable evidence that they (and arising social dynamics, i.e. war) could be lethal to the social and economic world system du jour, when given several decades to do their magic.

So it's better to think ahead and take action, as individuals and as society.

I think the point is you don't have to switch, even if you're doing some kind of realtime webapp you can still piece it together using php and redis (though I'm not sure whether there's a good comet/websocket framework for php). But it's sure going to be a hack. Also, using Akka or OTP you get a lot of stuff for free: supervisor hierarchies, lightweight processes (not everyone must be an OS thread), pattern matching. So I would agree that pitching these technologies to garden variety web and mobile apps on the value proposition that it's better at dealing with multi-cores is disingenuous. As you said, it has its benefits there. But maybe some of the proponents are just using multi-cores to illustrate a general trend, and maybe less of a singular value proposition.

If the system can successfully masquerade as an older Skype version it stands a chance. That is unless Skype has a baked-in not yet understood mechanism for pushing protocol changes to its clients. Skype would have to find discerning features, implement it on their servers and possibly even push it to their client updates. Whenever such discerning features were found it's a simple matter of arms race, i.e. a difficult but fair chance. Alternatively Skype can start blocking older versions which is rather unlikely.

Concerning the open and better issue; There definitely are open alternatives. None of them have quite the firewall-defying capabilities of Skype. Nor the user base for that matter. Building any kind of social network is fraught with chicken&egg problems and those first to reach mass have it made. Just check how one of the richest and most powerful technology companies Google is struggling to get a foot in Facebook's market.

I have done some work with RAAMs and other simple recurrent network based methods. It's terribly tedious, you can hardly represent more than five symbols or so.

I did my Bachelor's thesis on related techniques (pressing variable length strings into fixed sized vectors to deal with neural nets - URL below) - I can only say pertaining to neural networks that the decoding capabilities of NNs are severly limited. I actually developed a Jordan Network with a conventional additive sigmoid NN that could encode/decode more than 40 symbols! The technique was based on Cantor coding, but I had to set the weights by hand and could not retrain without losing performance (unstable fixed point attractor of Backpropagation through Time) - i.e. it's kind of a dirty hack.

It's also really hard to adapt parameters for sufficiently large nets (so my idea is that you need huge nets to represent language). In order to deal with this complexity limitation I've also looked into reservoir computing style networks (Echo State Networks) which use a large randomly initialized network paired with a linear learned network. ESNs may be good at modelling many kinds of temporal dynamics, but their capacity to represent relations seems rather limited as well.

So making an arbitrary length string walk into a vector may sound attractive, but either (a) don't expect to be able to decode it using conventional NNs or (b) don't expect to acheive compression, i.e. blowing up your representation may help (cf LVQ nets) and using compressing techniques such as Hinton's deep belief networks won't.

http://gpickard.wordpress.com/2008/11/11/well-heres-my-actua...