| Saturday, March 22nd, 2008 |
| 3:02 am |
Kathy Sierra might have a point  |
| Friday, March 14th, 2008 |
| 2:37 am |
Academia I look at many smart brown undergrads and often ponder why I am plodding through the PhD program. So I have decided to list the pros and cons of the life as a grad student and afterwards. Pros:- A PhD program is a the only time in ones life when one is paid to do what one wants to do.
- In academia one is constantly bombarded with ideas. Some innovative, some bizzare, some useful, some useless. But that's the fun and enriching part of life.
- If you are a slacker, being a tenured professor is probably the best way to slack. ( I am not sure how many people do that though)
- One has a lot of creative freedom. You can't bring peace to the world, but you can do something on that scale in your field.
- As a professor, you not only teach the future generations but also affect their view of the world.
Cons:- You don't make as much money being a professor. I think that's because the current structure of academia is risk averse. A professor doesn't have to be risk averse and hence can make more money. In some fields you can make money starting companies or selling technologies. So, it's totally up to you.
- Academia is just another business where people are willing to be objective. It's neither clean nor egalitarian.
- Academia also makes it much more difficult to be non-orthodox than industry. Luckily, brown is liberal. There is room for every reasonable argument.
After looking at both industry and academia, I think the difference is not much at the end. If you are good, you will prosper anywhere. In academia one is likely to trade-off intellect with wealth, but it need not be that way. |
| Saturday, March 1st, 2008 |
| 4:35 am |
VMWare Fusion I have a VMWare Fusion license. If you need it, please leave a comment. Earliest comment gets it. PS: It's perfectly legit. |
| Friday, February 22nd, 2008 |
| 5:55 am |
Office 2007 UI
Just back from a presentation on the new Office 2007 UI by one of the managers of the office 2007 team. In short, MS is not interested in making a better UI as long as the customer is paying for it. Summarizing my thoughts:
- Office UI is more or less in the dinosaur age of UI design.
- Design philosophy: We will design the UI in a way we see fit. If it's really inconvenient for the user, we will think about it.
- On Support for users with disabilities: We designed the UI and got "certified" by the usability companies that the new interface is faster to use than the old one. It looks like no thought was given to designing better interfaces for users with disabilities.
- On alternative UI's for office: We don't think people need alternative UI's. All users have the same need. Which is ironic given that their usability data shows that only 15% of the features are common to all the users. Many people in the audience repeatedly echoed the same opinion but the question was just elided.
- Data Analysis: They automatically collect a lot of usability data from office 2003. But, sometimes they seem to relate correlation with causation.
Anyhow, I think MS should do a better job at the UI design of a product that's used by 400M(800M unofficial) users worldwide.
|
| 5:34 am |
On Knowing
The "feeling of knowing something" gives more pleasure than "actually knowing something".
- In the context of Presentations, Speakers and Audience
|
| Thursday, February 21st, 2008 |
| 10:31 pm |
I'm a Photo Journalist Yay! I am photo journalist now. Herez my first kill. |
| Sunday, February 10th, 2008 |
| 2:19 am |
Compiler Technology When I was an undergrad I thought a course on compilers was a waste of time. But, I only realise now how powerful the idea of "transforming one language to another while performing some smart optimisations" is. Here are 3 examples from very different fields that I came across in the last few days: GWT : "Google web toolkit" is simply a java to javascript compiler. While compiling java to javascript it takes care of all the things that a smart web developer would take care of. Things like Image Bundle, Shrinking the size of JS code, Maintaining web standards, Cross browser support, generating HTML and WAP code etc.. come for free without any performance overhead while running the application. Supple: Supple is in short a research prototype of a GUI builder that takes an abstract specification of a UI and generates UI's for any target device. But, while compiling the abstract specification it chooses a decision-theoretic optimisation model(AI stuff) to choose a UI design that fits the device constraints and user behaviour (from traces). This is cool at so many levels because the generated UI's are not only intuitive but also the optimum solution (assuming that the constrains are realistic). Supple++ is an enhancement over supple which dynamically generates a UI while taking the visual and motor skills of the user into account. For example, it can dynamically adapt the UI to the tremor in the user's hand. This means that the UI can automatically adapt itself to the a person with parkinson's or to a person on a bus. Bluespec: Bluespec is a modified haskell that generates HDLs. It's better than existing HDLs because it is a high level language and during compilation it generates RTL with no compromise on speed, area or latency. Moreover, the generated hardware is provably correct, which is extremely important because only 2% of a chip is usually tested given the extremely competitive environment. PS: The quality of a transformation depends on the expressive power of the input and output languages. If there is a mismatch in expressiveness, then the generated code is hard to debug. I think GWT generating bad code is one such example, but that's an engineering problem. PPS: Writing a ruby to JS compiler may be a good idea :p. |
| Wednesday, January 16th, 2008 |
| 8:48 am |
My Next Laptop Something that's like Macbook Air and cheaper. |
| Wednesday, December 19th, 2007 |
| 9:40 am |
Powerscrap and Orkut worm
Looks like there is a worm that effects powerscrap enabled orkut accounts. I am getting a lot of scraps that are spam.
|
| Monday, December 3rd, 2007 |
| 10:02 am |
Why do I
- lose the ability to think independently when someone is helping me?
|
| 9:54 am |
EQ? In one of my courses, I did well in all the assignments. But the last one was supposed extremely tough. Even after working on it since morning I do not get it. I am probably the only guy around who did not get it yet. Instead of working hard on solving the problem, my brain has turned off. It refuses to think. Does this show that my EQ is very less? |
| Tuesday, November 27th, 2007 |
| 3:41 am |
Experience is
... the reality that seeps through the gaps of non-conformance.
|
| Friday, November 2nd, 2007 |
| 10:42 pm |
Wake up Wake up sun. You lost a long time ago. |
| Saturday, October 27th, 2007 |
| 8:58 am |
Finger Food: Trail Mix I have a voracious appetite and I need something to munch on(lets call it finger food) while I am working. An ideal finger food I like should have the following properties: 1) It should be healthy so that you don't put on weight. 2) It should be filling so that you don't feel hungry every few minutes. 3) It should be dry so that it doesn't interrupt me while I'm working. 4) It should appeal to my taste buds. 5) Something that you can eat over a long period of time without getting fed up. Soya nuggets satisfy the criterion 1,2 and 3 only. But since a few days I have been trying Trail mix (aka Student Food) and I find it a perfect finger food. Firstly, the omega-3 fatty acids (unsaturated fats) in them are good for one's health. Secondly, they are filling because the nuts need quite a bit of gastric juice to burn. They are dry, tasty and above all very easy to prepare. Moreover, you can try variations of a trail mix if you are fed up with the regular one. Have it, but remember that too much of anything is harmful and trail mix is no expception. |
| Wednesday, September 19th, 2007 |
| 7:26 am |
Randy Pausch Randy Pausch an ex-student of my CG prof and a big name in the field has given his last lecture on "How to achieve your childhood dreams?" today. The webcast was one of the best I have watched in recent past and is quite inspiring and touching. As he points out, it was in short a talk on "how to lead a life". PS: I wish I had the link to post. Powered by ScribeFire. |
| Tuesday, September 18th, 2007 |
| 9:48 pm |
(Some) Profs in Brown literally "ignite your passion". Powered by ScribeFire. |
| Monday, August 6th, 2007 |
| 8:45 am |
Phonies Most of the world I see now is full of phonies. Is there a place for people who really crave for quality? |
| Monday, July 30th, 2007 |
| 6:43 am |
|
| Tuesday, July 17th, 2007 |
| 8:01 pm |
Lyfe Upgrade Going to Providence, RI in august. Got a fellowship at Brown for MS/PhD in Computer Science. Most likely, I will be working in the field of programming languages and software visualisation with Prof. Shriram Krishnamurthi and Prof. Steve Reiss. If everything goes well, I might continue working on NPL. PS: It would be great to know any LJ'ers in and around boston. Please leave a message. Powered by ScribeFire. |
| Saturday, July 7th, 2007 |
| 6:27 pm |
I need a better CPU The only time I wish to have a better processor on my laptop (during day-to-day work) is while compiling lot of code. Powered by ScribeFire. |
![[info]](http://p-stat.livejournal.com/img/userinfo.gif){kind=small}
mansu's journal