The Skyrim soundtrack has created waves in the gaming music industry. This is probably the 20th time I’m listening to the main theme, and it still gives me goosebumps. Check it out here:
Here are some amazing covers of the song.
And of course, we cannot not have a metal cover.
Steve Jobs passed away this last week. A very unfortunate and unexpected event for all technies and people use have enjoyed using the Apple products. I am personally not an Apple fanboy, but there are two thing that I admire Steve Jobs for. First, his talk at the Stanford Commencement. This has inspired me a great deal in the past, and help get past some dark days.
The second are his world famous keynotes. Inspite of not owning a single Apple product, I had come to look forward to what Steve will bring on stage. Would there be a new phone? A new subscription plan with T-mobile? A cheaper classier iPad? Some new cool way of interacting with mobile devices? I had even begin to complain that the three month wait was too long.
Thanks, Steve — for making us look towards the future in optimism and hope.
The impact of this loss could be best summarized with the posts on Hacker News a couple of days ago. Screenshot (take from here) below:
I just got off a very interesting discussion today on how teaching introductory computer science must be improved. I must say that I was very excited to be part of this discussion where people (Rich and Eric) were very keen on changing things to improve the overall situation. Here are some of the points which were raised:
That being said, my personal take on this is the following:
A more interesting language must be taught to these students. Eric and Rich are currently targeting Python as a possible replacement for Java. MIT is also heading in this direction. I personally believe that a better language for this purpose is Javascript. There are multiple features which make it interesting.
A few other things students must learn in an introductory programming class is the importance of working together. No real CS project is done alone. Working with other people is definitely the more fun case). Students must learn how they can form groups and work out projects together.
Defocussing from the standard examination grades is probably the most important thing to do. The role of education is to make us aware socially and technologically, not to post a 4.0 GPA on the front door. At the end of the day that’s just a number. What is more important that you are able to creatively think, formulate ideas and build robust systems which are useful to the community. Unfortunately, most courses are the opposite. They love to have that dark and gloomy examination at the end of the quarter to ruin all the fun. This is true in India (where I took courses for 20 years) and in UCI (much much lesser here, but still relevant). Rich had a wonderful point when he said that only entrance exams must be conducted, and not qualifying ones. The entrance exam has a better feedback opportunity. If one fails it, he can retake a pre-req course which he didn’t completely understand. A final exam is just a nail in the coffin — there is no way to mend things after it. UCI makes it worse by declaring that students cannot retake a course (once they have a grade for it).
K.K. Downing, the guitarist of the heavy metal band Judas Priest recently announced his retirement. Came as a gentle surprise given that he definitely has some riffs still in store.
Priest has been one of the most influential bands for both to the heavy metal scene and me personally. They have truly made a difference with all their anger and leather.
Here is K.K. with a recent amazing solo:
Are social networks useful? There seems to be no clear answer. There seem to be both disadvantages and advantages when I ponder about it. So I feel its time to tinker with this. As a first experiment, I unfollowed all the famous personalities from my twitter list. These are people who actively engage in building modern day software, or are world famous entrepreneurs or philosophers. Though they have good tweets, I want to see if it really makes any difference.
My guess is that it doesn’t. My life is going to go on without any changes. As if it never happened. A ton of these things are simply mind-engagers. Things a human mind can easily fix its attention on. Its similar to those crappy TV serials. Similar to those countless blog articles who give you advice on how to be better programmer/entrepreneur/scientist/<your-profession>/<your-hobby>.
To see how many of my followers are actually drowning in noise. I am going to ask them to send me a “hi” reply tweet. Needless to say, a person following me without the social noise syndrome will see my request to the “say hi” tweet, whereas those who do face it might not see my tweet itself.
On the usefulness of social networks, here is something that I found fascinating today: http://jonathanstark.com/card/. Really looking forward to see how this turns out. Wonder if the NY Bronx and Bombay Chawls can be connected in a such a way that they can help each other out.
On a slightly different note, this reminded me a lot of the population implosion McLuhan talked about in Understanding Media, and I decided to call this article Social Implosion. We are definitely facing social implosion on most of our networking sites. There is just so much jabber — very little of it actually useful.
The news is raging about Soc Net websites requiring people to put their real names. Danah Boyd hates the idea, citing how minority people mask their real identities. Whereas, the Atlantic thinks its a revolution. I hope the developers of future social media websites realize that both viewpoints are right in their own ways. What we need is flexibility for people to switch back and forth between the two extreme points. Just like in real life no one either hides in a cave or stays on the street their whole lives, future services must empower people with tools to completely hide their data whenever they need it, and expose private information to only those eyes to whom its deemed necessary.
As far as using private data for advertising is concerned, well, I don’t like it any more than you do. But someone’s gotta pay those bills.
Some lovely tools have been released demonstrating the power of JS and HTML5. The first one is Fabrice Bellard’s Linux Emulator: http://bellard.org/jslinux/. The second being a JS based MP3 player http://jsmad.org/.
I just stumbled upon popcorn js which very similar to a class project I worked on with a couple of friends: http://popcornjs.org/ Popcorn lets you create a time synchronized multimedia presentation. The demo on their front page explains it all.
This is a right time for all platforms to seriously consider HTML5 as THE platform for cross platform applications. I really wish iOS and Android embrace html as their de-facto SDK language rather than Objective-C or Java (both being great languages in their respective ways).
Just had an ideabolt and decided that I *had* to write down about it.
Many different websites exists. Each serving a specific task. Usually, as users we see that one service alone does not suffice. People go to different websites, accumulate knowledge and then draw insights from their mental model to execute their ideas. Usually, when a person makes a hop from website A to website B, there is a break in context. For example, lets say I want to watch a movie right now. So I look up Showtimes on Google to see what is playing around me. There are usually about 4-5 new movies in the theaters. Not satisfied with the reviews appended to the listing. I go to rotten tomatoes for reviews. But now, the RT website requires that I type in the names of the movies individually to get to their ratings. Lots of work!
With kry.io, things could be different. kry.io consists of two parts: The first is a database of absolute knowledge (things that are same to everyone: movies, books or real world events). The second is a collection of key-value pairs of data describing what a user did with respect to these absolute knowledge pieces. So coming back to our example of movies, lets say the new movies listed on Google’s page were:
I would click on the first 4 movies to see some limited reviews and showtimes. For each movie I click, Google appends a key-value into kry.io like:
When the user goes to the rotten tomatoes website, RT can query kry.io for movies which have the flag changed, and show a set of quick-links to these movies.
kry.io will remember Google’s flags for only a short duration of time (like 20-30 mins). It’s a global transient user-context memory. An interesting experiment would be to see if users would use the kry.io service, so they can be tracked within different media (WWW <—> iphone/Android apps).
The first one being the prosecution of Italian Seismologists for not warning citizens correctly:
http://www.geekosystem.com/italy-scientists-earthquake/
This is unmerited. Anyone who has ever taken a Science/Engineering course will know that its clogged with errors and failures. The whole point is to reduce errors with effort and experience. Japan definitely has some great weather monitoring technology, but they couldn’t do anything to prepare themselves for what happened recently. The world is too complicated for technology to be 100% correct about it. Any mature person at an impactful position must understand this.
The second one is more disturbing:
http://www.bbc.co.uk/news/business-13546177
All I can say is that this seems more reminiscent of the dark ages than the 21st century. People must not stand for this kind of acts. This is the reason I’m writing this. Its very good to see media reporting such cases happening in corners of the world where we wouldn’t otherwise look. Hope they keep up such reporting and eventually put pressure on the governments to take measures against such narrow-minded decisions.
People have been trying to give structure to completely unstructured unstructured data since the 70s. Efforts have largely been to study images and videos, create textual attributes (who is in the photo and what are they doing?). This has been the holy grail of computer vision. There have been some interesting advancements in the areas of surveillance and monitoring. But in the area of personal media management, my personal pictures are having a hard time using these techniques as they were created in a much more dynamic environment. The task is simple: Organize my pictures. I’ll throw in all my photographs taken by my Nikon D-60 camera, just put them in the order of (who, what, etc etc..). The difficult part is that the bridge between pixels in my photos, and the attributes that form an organized collection is broken.
The funny thing is that most of my interactions with people happen online. I call/email/text people before meeting them. Look up concerts on ticketmaster/livenation before buying tickets. Friends send invitations on Facebook for attending an event. There *must* be a ton of data on the internet which I have access to, which can describe more gracefully what my pictures are trying to capture. The right way of going about organizing personal media is to find the right associations between media and already structured information that’s out there. Its hard to understand context from the photos itself. But different pieces of that context exist in the other sources. A personal media organizer has to look at these sources and connect them together.
More on how this will be done later…
