Pages

Breaking the new iPhone

See the crazy video where a person rips apart his new iPhone minutes after he purchases it at Apple Store, Palo Alto.

Popular programming languages

One question was always of my interest right since I started working as a Open Source Data Analyst at SAP Labs, Palo Alto: Which is the most popular/used language in the Open Source development? Thus, I decided to do an analysis of number of projects where a particular language is used. The list below shows the top ten entries.

Language

Percentage

Java

18.6

C++

17.3

C

16.0

PHP

13.2

Perl

6.1

Python

5.0

C#

3.8

JavaScript

3.5

Visual Basic

2.1

Delphi/Kylix

2.0

Unix Shell

1.9

Assembly

1.5

PL/SQL

1.1

Others

7.8

Then I thought that this metric is not a fair metric for evaluation. First reason is that the projects are not of the same size in terms of Lines of Code and secondly there are is large variation among the number of developers for an open source project. Then I rank ordered the languages by the total number of lines of code for a particular language. The results are as follows.

Name

Ratio

C/C++

47.29%

Java

26.90%

PHP

7.63%

C#

5.17%

JavaScript

3.44%

Python

3.00%

Perl

2.22%

Pascal

1.15%

Ruby

0.93%

shell script

0.75%

Tcl

0.75%

Objective C

0.51%

Emacs Lisp

0.17%

Erlang

0.07%

Scheme

0.02%


Here we can see that C/C++ has much more impact than Java. This can be contributed to large projects in C/C++ like Linux where there is a lot of effort involved. On the other hand even though Java has larger number of projects they are smaller in size and effort.


Open Source Documentation

After my Bachelor’s in Mechanical Engineering I started to work at an Information technology firm. The first and the most important thing which was told to us (new software developers from non computer background) and then pounded every now and than is to write proper comments/documents for the code one writes. Thus, for whatever lines of code we develop there was about 25-30% of comments so that the next developer who works on it can better understand the code. Now when I am in Open Source world the situation is opposite. There is hardly any documentation for the code developed which may make the life of the developer studying the code very difficult. Open Source development is always criticized for not having proper and formal documentation. Often the developers are geographically dispersed and communicate with emails or IRC. The following table gives us the relative amount of comments with respect to the total lines of code for the Open Source projects. Thus, the data represents the percentage of comments for the Open source code for a particular programming language across five thousands Open Source projects considered.

Language

Average comments (%)

Matlab

0.46

Java

0.35

Ada

0.29

Pascal

0.27

PHP

0.24

rexx

0.24

C#

0.24

AWK

0.23

DOS batch script

0.21

C/C++

0.20

Objective C

0.19

Assembler

0.19

JavaScript

0.18

Ruby

0.16

Tcl

0.16

shell script

0.15

Emacs Lisp

0.15

Perl

0.14

Scheme

0.14

Lisp

0.13

Python

0.12

Lua

0.11

Visual Basic

0.10

CSS

0.08

Boo

0.04

HTML

0.03

XML

0.03

It can be seen that the percentage of comments/documentation is really very poor. Matlab has the highest comments of 0.46% of the total lines of code which is very poor with respect to the classical software development.



Mystery Spot

A couple of weeks back I visited the mystery spot located in Santa Cruz, CA where it is said that the gravitational force does not act in the vertical direction. I knew that there is no proven scientific theory behind this phenomenon and being an analytical science based person was looking to solve the puzzle or at least criticize the hypothesis. The tour guide demonstrated some things, which were certainly mysterious. He showed us

1 > Rolling a Golf ball on a “seemingly” slanted surface which stops in between and rolls back in the opposite direction.

2> A pendulum which needs much more force to make swing in one direction than the other.

3> The height experiment where an individual seems taller than the other at one position and the shorter in the other. The guide then clarified that the surface is horizontal with the use of bubble slide

There is one thing one should note that all the surfaces are tilted which causes disorientation with respect to the horizontal and vertical.

There have been many theories to explain the phenomenon

1> A alien space ship landed here which causes this disorientation (the most weird theory!)

2> There is a large carbon dioxide gas flow nearby which changes the gravitational force

3> A heavy metal deposit which causes the anomaly

None of the theory is sufficient to explain and answer the weirdness of the place weather it may be ‘feeling light headed’, ‘top heavy’ or hanging tilted of an individual at about thirty degrees on a horizontal bar. The honest answer is that I (with all my friends) am not sure weather whatever we saw and experienced there is an optical illusion or in fact a real anomaly. The fact that nobody still has been able to explain; makes me believe that there is something about the mystery spot. If you are in Bay area, you should certainly visit this place to experience all yourself. The Mystery spot is located at 464 Mystery Spot Road, Santa Cruz, CA 95065 (http://www.mysteryspot.com).

Are you Orkuting?

Orkut recently launched the service where you can insert the various RSS feeds in Orkut itself. So now when you visit a profile of a person you can get the complete information with his favorite videos, testimonials, RSS feeds, blog, personal website and the other profile information provided by the user. It has become an ideal place to portray your complete social profile. It also provides functionality to rate a user on dimensions like cool, trusty and sexy, status on Gtalk, recommended communities and latest friends who have login. It has much richer functionality now than it had when I first started using it sometime in 2004 (that is obvious). Still I feel it lacks some features, which can make it richer.

1> The messages in Orkut have little functionality and more looked upon as spam. I personally never read the messages and delete it blindly.

2> The scrapbook does not have any method where I can set rights or preferences. I mean there is no functionality where I can set permission for a certain user to view it and some not.

3> Deletion or selection of scraps from a particular user or in certain time-frame. We need a scrap search engine.

4> I guess there is some limit on the number of photos in the Orkut album, which should be increased.

5> Orkut album should support slide shows from the Picasa web album.

6> Some maps functionality for the user addresses

7> Better analytics features and not just number of users visited in last week, last month and since one year

8> Better server and uptime (some time I am fed-up with the number of donut I get). I agree that the service is much better than the past but still there is a large room for improvement.

9> Better monitoring and filtering of sensitive and spam material on communities (considering the issues in India where some people were offended by some anti material against a group on the website)

10> Support for reminders other than birthdays like anniversaries, some special festivals etc...

11> Some Google gadgets for Orkut

Smell or taste?

The biggest challenge in the cosmetics industry is the need for innovative packaging techniques that can be differentiated from the competition. Has anyone thought of the effect if the vendors stop spending money on packaging and concentrate on the quality? Personally, when I buy goods I look into its quality, reliability, and then cost-benefit. I hardly ever care if that product has an attractive packaging. However, it seems the opposite in the ‘pink’ cosmetic industry. The notion of quality is more tied with the innovative, attractive eye-catching packaging of the product than the actual quality of the product. That means the quality evaluation and perception towards the product is directly influenced by sleek packaging, shape and color. The brand name also plays a big role as in every other arena of products. But can you imagine the cost of the cosmetic products if the manufacturers stop spending and researching on attractive packaging and aesthetics of the products. I mean what will be the effect if one of the big companies like Estee Lauder Companies, Mary Kay, Procter & Gamble Cosmetics or Revlon introduce cosmetic products in two modes one which has all sleekness, color and shape attributes (expensive) and the other in a simple container and a ordinary package (half the price of the first). This will serve both kinds of the customers. Those influenced by the ‘smell’ and the others influenced by the ‘taste’. I was just wondering that could this be a workable business model. Will such a marketing strategy work where the cost of the product is half of the “attractive” product with same quality and “brand”?