Polyforms

Playing around with GitHub pages. I just might move away from tumblr. I hate posting code here; it's too much of work and still the code does not come out well (check previous post). Plus, I'm loving the idea of doing everything through the linux terminal :D  

I have been studying from Foundation of Computer Science by Jeffrey Ullman and came upon an exercise (on page 32; "Sorting Keys").

Suppose we have a constructor STUDENT as following:

struct STUDENT{ int StudentID; char *name; char grade; } A[MAX];

and we wish to swap two STUDENT constructors on the basis of the condition:

A[j].StudentID < A[small].StudentID;

where 'small' is the smallest known StudentID value till a particular iteration of a loop.

Now, one way to this would be to swap every variable of the strucute with each other. Like this:

TempA.StudentID = A[j].StudentID; A[j].StudentID = A[small].StudentID; A[small].StudentID = TempA.StudentID;

and so on for every variable in the strusture.

This will take loads of time which only increase as the number of variable in the strucutre increase.

So, the book highlights a solution to this: "Since it is time-consuming to swap whole structures, a more efficient approach is to use a second array of pointers to STUDENT structures and sort only the pointers in the second array. The structures themselves remain stationary in the first array. We leave this version of selection sort as an exercise."      

Here is an idea about how to do it..

I finished reading Jony Ive's unofficial biography by Leander Kahney few days back. Nice book. This along with Steve Job's biography give an interesting perspective towards the products of Apple; you get to read about the story behind every product. Which is pretty romantic and unique in the consumer hardware industry. It gives every device a sense of heritage.

After the beginning, where the author talked about Ive's childhood, schooling and british education, the book focused a lot on design in Apple during the era without Steve and for quite some time it gave a feeling that the book was more about Design at Apple rather than about Jony Ive. Though, later it did focus back on him and I guess it is important to talk about design at Apple, even the ugly part, because it gives a sense of scale to Ive's achievement. 

Jony is a likable man. He shares the same amount of Intensity, focus and desire for perfection as Steve him but does not degrade other people in the process.

Nice book. Would only recommend to people who like or are interested in Apple, design or both.

Image credits: funkyspacemonkey.com

I will shamefully admit that I was unaware Aaron Swartz before he took his own life about a year back. I was and still am using lots of things he had worked on during his amazing life; things like, RSS, Creative Commons, Markdown and of course Reddit - the most influential thing ever in my life.  After becoming aware of Aaron and reading loads of stuff about him on the internet, I started reading his blog Raw Thought. It has some of the most thought provoking essays you could read from a man so young. 

I have been going through the blog's archive everyday, even sneaking time to read between work. My favourite post remains the last essay he would ever write: What Happens in The Dark Knight. 

Aaron has been inspiration for me, an inspiration even after his death.

Isaac Asimov is the master of science fiction. Decades since they were written, Asimov's stories still inspire awe and imagination from it's readers. Almost every story has a philosophical viewpoint which makes the reader sit back and think.

The Complete Stories consists of several short stories, set in the future, which focus on problems of a certain individual or humanity as whole.

SPOILER ALERT.

I have observed few recurring characteristics in Asimov's stories. One is how dumb-downed humanity becomes in his future universe. Despite (or due to?) all the technological achievements, the man loses his grasp of what all that made that technology possible. There is too much abstraction in his universe; technology works but no one knows how. The story "Feeling of Power" is a perfect example for this case. The people of that world are so dependent on computers that they are not only incapable of doing basic mathematical calculations but are also unaware about the methods to do so. According to them calculations can only be done by a computer; the human brain is just not capable of doing it. But when a person does discover some methods for performing mathematical calculations, the people in higher authority are in complete awe and can only think about how such calculations done by a human brain can potentially replace the computers completely and give them an edge over other nations. It might be hard to believe that such a situation can ever occur but come to think of it, is this not one of the big questions being asked in today's always connected world- Is technology making us dumb? I'm not the right person to answer that question, so I'll refrain from doing so.

Suppose we have a system in which a user gets points and he/she can "level-up" after collecting certain amount of points. The amount of points required to level up increase with each increasing level. Pretty much a real life scenario which can be seen in video games like Pokemon (in which case the level up is for the Pokemon). One way to design such a system could be to manually assign threshold values for every level. This could work if the number of levels is not high; 10 or maybe even 20 levels can be assigned a threshold value individually. But if the case is such that the number of levels are not defined or are very high then this approach fails simply based on the fact that individual assignment of threshold values would be too cumbersome.

We can easily use Fibonacci Series to help us out in such a case.

sum = prev_no_one + prev_no_two threshold_value = sum * k

Here, 'k' is a constant value which will determine how conservative or dynamic the system will be (Larger value will mean larger difference in the threshold value of each level and vice-verse).

Example: 

Threshold value for a system of 10 levels:

int main() {

 int prev_no_one = 1;  int prev_no_two = 2;  int sum, threshold = 1;  int k=5;

 for(int i=0; i<10; i++) {

      sum = prev_no_one + prev_no_two;       threshold = sum*k;

      printf("%\n", threshold);       prev_no_one = prev_no_two;       prev_no_two = sum;

  } 

}

Output:

15 25 40 65 105 170 275 445 720 1165

The above number are the threshold values for each levels respectively.

I'm working on an app which requires communication between a JSP and an Android app through HttpServlet.

I need to send and receive data, in XML format, from the app to the server and vice-versa. There are loads of APIs to parse and/or serialize data. Here are the options I'm considering: 

1. DOM: Too old. Too heavy; though, the XML files generated by my application would not be too heavy, using DOM just did not seem right.

2. SAX: Again, too old. But I was ready to use it. I even a taught myself how to parse XML the SAX way. But turns out that It is not possible to serialize data through SAX but only parse. I'm sure there must be a hack for it; I'll look into it later.

3. JAXB: Looks neat for the server but has Android does not support it. So I'll have to include the API library into my Android APK. Which will take 9MB(!) of my APK. Not good. 

4. StaX: This one is neat and simple too. But again, no official support in Android; I'll have to import a (small) library.

5. XMLPullParser: Android suggests the developers to use this. I have to take a look into it but apparently it is similar to StaX. Good. 

For the past two months I have had the opportunity to unbox and play around with the N9, thanks to a colleague at my internship.

I have finally rooted my i9300 and installed CM 10.1 on it. It's been a little more than a day and I'm liking the new experience. 

Some hurdles did come up during the rooting and, as well as, the flashing process. First, I was unable to root through Odin on my laptop. It always failed. Tried it on another computer and my phone got rooted instantly. Then, I was unable to boot the ClockWorkMod recovery. I looked around the forums a bit and downloaded ROM Manager from the play store. After that, CWM worked fine only to point out that my copy of CM 10.1 was corrupted. So I had download it again. Finally, everything was successfully done and I had a brand new ROM on my i9300!   

I have been destro hopping like a mad man! From Mint 14 Mate to Mint 14 Cinnamon to Mint 14 with fluxbox to back to Mint 14 to Mint 15 and now to Crunchbang! Add to all this a little experiment with Arch Linux in between. It's really fun and quite addictive, to be honest. 

Now, I think Crunchbang (#!) is a destro which I think suits me very well. Mint is a really nice destro too but there was something about it occupying about 15GB of my HDD space that made me uncomfortable. It had loads of useless software on it, which I never planned to use.  Crunchbang, on the other hand, currently uses about 5.0 GB of my HDD space. 5 GB after me some useless installation. Not that I'm low on HDD space or enything but minimalism!

The default RatingBar on Android is ugly and too big for most occasions. So, I decided to design my own. All you'll need is a decent graphic editor to make the rating stars (actually, anything; you can even make thumbs up and use those instead of stars).

Alternatively, like me, you can download Android Stencils and Sources and take the star stencil from there; its perfect.

Here is how you go:

1. Create two graphic file in .png format. One image should be of a filled star and the other non-filled. Remember to make the background transparent. 

2. Place these two file into the drawable folder, found in /res.

3. Create an xml file in the drawable folder and put the following code into it:

4. Add the following code into styles.xml file, found in "values" folder. Place the code inside "resources" tag 

5. In the layout file, where you are defining your RatingBar, add the following argument into the <RatingBar> tag:

style="@style/style_name"