YuviSense: Codin Kid

Wow, that was quick! Thanks Scoble, for the link. My traffic just doubled (I had a grand total of 150 hits the day before. Today, it is 300!). So, here we go, with the links.

The Top Ten

Without much ado, here’s the top ten sites he linked to:

The Links is the number of links to that host, while Posts is the number of posts that contain atleast one link to that Host. The Link percentage is the total percentage of links going to the host, while Post percentage is the total percent of posts containing atleast one link to the host.

And, here’s your chart:

Note that this is just the top ten sites, sorted by number of links.

Flickr?

I was actually surprised by the first spot being taken by Flickr. But then, some dirt digging bought up this piece from Flickr’s TOS:

The Flickr service makes it possible to post images hosted on Flickr to outside websites. This use is accepted (and even encouraged!). However, pages on other websites which display images hosted on flickr.com must provide a link back to Flickr from each photo to its photo page on Flickr.

That means, you give some juice to Flickr whenever you use any of your own images. Neat trick, because a search for Share Photos on Google gives Flickr the first spot. Google’s own Picasa is relegated to a mere sponsored ad: I had to go to the third page to find Picasa, and heck even this feature page at Microsoft.com about digital photography in Windows is placed higher than Picasa:D

Incase you are wondering, Scoble’s Flickr page has 14 links, but that still places it as the 5^th most repeated unique link. And of those 245 links, 165 are to his own pictures, 12 are to Laughing Squid, 11 are to Thomas Hawk and 3 search queries. And, here’s the chart:

Blue pacman, isn’t he?:D

Finishing up on Flickr, here’s how links to Flickr have changed over time:

Lots of cliffs there, ain’t it? He linked aggressively to it in September, but after that seems to have forgotten…

Day Job

At #2 and #3 are the expected ones: His former and current paycheck yielding day jobs. As you can see, PodTech is about to come second, but it had already come second if you consider the 18 links to ScobleShow.com which redirect to PodTech. So, let’s draw some charts, shall we?

First, let’s see how much he’s comparatively linked to both:

But, how was it before he left Microsoft?

Geez, PodTech is very low. The fun fact here is, that before Joining PodTech, he actually linked only 7 times to PodTech and 93 times to Channel9, to make a total of 100. To a 15 year old, that’s kinda funny.

And, after joining PodTech?

Channel9 is not as bad as PodTech was, isn’t it? He still links to them!

Let’s do this more accurately:

Pretty nice shape. Looks like a, uhm, well? My imagination is pretty retarded… Comments on the shape are welcome..

As we can see, there were only 7 links to PodTech till the day he joined them. Extending the projection, we see that at that rate, there would have been only about 15 links to PodTech if he had not joined them!

Also, the links to PodTech have really taken off only from September: Till September, they had only 19 links! This means they went from 19 to 121 in a span of 113 days, at almost a link a day! In contrast, Flickr gained only 23 links in that period, while Channel9 got just 12!

Goooglee?

First time I saw the result, I couldn’t believe that Google was at #4. I knew he linked to some searches, but 107? I thought I had found a bug, or maybe discovered a Microsoft conspiracy. But, on detailed analysis, it came out right.

First, here’s the graph of how it went up:

See the sudden increase after September? It’s his link blog. Infact, his link blog is the most linked to single URL, with 43 links in 41 posts. Here’s a chart of just his link blog:

See? It’s growing pretty fast. But, there’s more:

That is a staircase if I ever saw one! So, he occasionally goes into search mood where he links to a lot of searches, but eventually forgets about it. Rinse. Lather. Repeat;)

This shows that he’s linked to 54 Google Searches! That beats his link blog, but unlike the linkblog, he’s been doing that on and off, rather than continuously.

MemeTrackers

TechMeme is at #5 and tech.memeorandum.com is at #7. Expected. But, it should be a lesson for anybody trying to change domain names after the game has started.

Here’s the comparative Chart:

As you can see, it took TechMeme some 6 months before it had the same link popularity as Memeorandum. It all evened out in the end, but if they had stayed the same (or chosen to be TechMeme from the beginning), they would’ve had much higher link popularity…

On the other hand, there’s a rumor going on that if you link to TechMeme, you have a higher chance of being featured there. Now, is there a connection between that and this?

MSDN Blogs

Another surprise, since the only link to MSDN blogs that I remember seeing was a link to Raymond Chen. Here’s the graph:

Hmm… Not exactly consistent. Some mini-stairs near the end. This means he links to them once in a blue moon, but those blue moons tend to occur frequently.

But, he seems to link far more to individual blog posts rather than to the blogs themselves. Here’s the chart:

That shows that he’s interested more in what the MSDN Tech bloggers say than the bloggers themselves.

There is, however, one exception to this: Raymond Chen. All 5 links to him are to his home page. And of that 5, he typed the URL 3 times from his memory, the remaining 2 times he copied from browser’s address bar or his RSS reader. [You could get to know this pretty easily: If the URL is typed from Memory, then it’ll be just http://blogs.msdn.com/oldnewthing. But, if it were copied from somewhere, it’ll be http://blogs.msdn.com/oldnewthing/default.aspx. ;)]

And, as a tribute to Raymond Chen, here’s the graph of links to Raymond Chen:

Perfect Staircase:D

ZDNet Blogs

Another surprise, because my feeds don’t have anyone from ZDNet Blogs Gillmor and BTL are the major players here, getting 19 and 15 links respectively. But, Gillmor gets 7 links directly to his blog, which might mean that Scoble’s talking more about Gillmor than what Gillmor’s talking about. Or, maybe it’s because Links are dead

Here’s a breakup:

And, here’s a chart of the linking frequency to ZDNet Blogs:

And as another tribute to Steve “LinksRDead” Gillmor,

Seems like after April, Links to Steve Gillmor are dead:D The fact that the blog’s not been updated since July might have a small relationship to this:D

Dave Winer

He’s the only one to get a standalone blog up in the top ten. Of the 50 links to his blog, 8 are to his homepage while the rest are to specific entries. Here’s his chart:

Pretty consistent for links to a friend. But then, Scoble’s linked to more Google Searches than Dave Winner:D

I’ll deal with bloggers alone in another post.

MSN Spaces?

Caught me by surprise. But then, ah, his wife’s blog was over there!

See the flatness after August? That’s because MSN Spaces is now Live Spaces, and all the PageRank built by the MSN Space bloggers are smashed(unless they’ve gotten some type of redirection that works with Search Engines in place).

Another tribute, with this Chart with the only two line title:

What happened in the gap between January 06 to August 06? No link? Huh? Or, bug in my code? I searched, but I couldn’t find any bug. Infact, I actually looked through the link list, and found only 20 links to Mariyam on MSN Spaces and 18 on Live Spaces. Maybe, she was using a newer blog engine? Different URL?

The Long Tail

These high profile sites might seem to be dominating, but then there’s the long tail:

As you see, 73% of the links are spread out across 1838 hosts! You can look at it in the positive, that 73% of the links are spread out, or that 27% of the links are concentrated in just 20 sites. Take your pick:)

Now, splitting that up a bit more,

70% of the sites linked to have gotten only one link! That means, if you get a link from Scoble, then there’s a 70% chance that it’ll be your only link

The Next Ten

I’m not goanna elaborate on these, but I’ll just give you a look:

Note that the Scobleizer Wordpress URL doesn’t count images or smilies. The Wikipedia links are evenly spread out, but he has 4 links to his own article

Notables

A few interesting stuff that didn’t get in there:

I gave up after I came to 200. Anything more you’d like to know specifically? Just let me know!

Word Analysis coming soon!

Notes

1. Deannie explained the mystery of the closing parenthesis in the comments. Thanks!
2. Posting the application online: I’ve got some requests, and though I have no intention of keeping it with myself, there is a problem with sharing that app that might not be apparent at first short. I’ll write a post shortly about that. And no, it’s not technical.
3. I’ll do a short post on comment analysis soon.
4. Now, let’s see how much time Scoble takes to find this one:D
5. Please note that I had attempted some humor in this post, so if it offends anyone, be offended:D

Everyone and their aunt knows that Google’spulled the plug on the SOAP Search API and pulled the SDK from their site. I expected that, but I figured that it’ll be replaced by POX over HTTP. Or JSON. I could’ve accepted even, uhm, CSV. XLSX even. But, I never actually expected it to be an opaque Javascript block, and still be called an API.

But, fear not! Davanum Srinivas has waded through the obfuscated Javascript and produced this piece of Java code that’ll give you a JSON feed of the search terms. It didn’t take me much time to convert it to C#:

using System.Net;

using System.IO;

using System;

public class GoogleAJAXSearchAPI

{

    public static void Main(String[] args)

    {

        String endpointURL =

                    “http://www.google.com/uds/GwebSearch?” +

                    “callback=GwebSearch.RawCompletion” +

                    “&context=0&lstkp=0&rsz=small&hl=en&” +

                    “sig=8656f49c146c5220e273d16b4b6978b2&” +

                    “q=Yuvi&” +

                    “key=xxxxxxxxxxxxxxxxxxxxx&v=1.0″;

        WebRequest wrq = WebRequest.Create(endpointURL);

        WebResponse wrs = wrq.GetResponse();

        StreamReader sr = new StreamReader(wrs.GetResponseStream());

        while (!sr.EndOfStream)

        {

            Console.WriteLine(sr.ReadLine());

        }

    }

}

The q parameter is your search term, while the key is yourGoogle AJAX search API Key. Note that Google may change this any time, without notice, causing code that was working to suddenly stop working and people will blame it on C#. Now, all we need is a library for parsing JSON….

After reading Sriram’s old post on building a search engine, I wanted to teach myself screen scrapping. So, I just knocked up Visual Basic 2005 Express Edition, and for no apparent reason, named the project PaleRash. Then after learning the hard way that most sites aren’t XHTML compliant (heck, even the old digg!), I looked around for a .NETty forgiving HTML parser, and thankfully found the awesome HTMLAgilityPack by Simon Mourier.

Now, with the technical side settled, I needed a target. A simple blog with a semantically meaningful structure. And the first name that popped up in my head was Scoble.

I’ve been reading Scoble since I arrived at the blogsphere, mostly as a news source: A human filter who filters the news so that it might not alter my faith in Microsoft Religion, add some opinions of his own, and boldly debunk some of the facts that could make the heretics lure out the faithful.

Anyway, so after a weekend of coding, a week of debugging, and another weekend of rewriting, I finally had a base which could be modified to run link and word statistics on any blog. The results were interesting, to say the least.

Now, since writing a single post about all that would be a crime on humanity, I’m going to split it up. So, here’s the part zero, just the lame General Statistics. Stay tuned, for I have more in the works.

General Posting Statistics

I used all the posts from scobleizer.com, from the time he switched to Wordpress.com (1/10/05) including the “Hello World” post, till 22/12/06, or the time during which he didn’t post for two days. That gives us 447 days of blogging, which equals to 2405 posts, at an average of 5.3 posts a day.

What’s a statistics post without graphs? I’ll forewarn you that today’s charts aren’t pretty interesting. Here’s our first:

That worm’s pretty straight. So straight, that after September, I could barely see the straight trend line I drew! So, though a bit rough in the early days, it has pretty much stabilized now.

General Links Statistics

In those 2405 posts, he’s linked to 4992 pages, 4321
of them unique, spread out across 1858 domains(including subdomains). That gives you an average of 11.16 links a day and 2 links a post, with 86% of links being unique.     

I love graphs!

Dang, another straight line! A slight drop in linking frequency at May-06 to Sep-06, but then it’s resumed to normal propotions.

General Word Usage Statistics

He used 464,078 words in all, at an average of 192 words a post and 1038 words a day. He however used only 14,898 unique words, including derivatives, meaning that each word repeated, on average, 31
times. In contrast, Shakespeare used 884,647 words in all, of which 31,534 were unique, and repeated each word 28 times on average^[source]. Translation: Shakespeare knew twice as many words, but Scoble wrote as much as half of all Shakespeare in a year and a half! I guess that maybe if I had considered his full blogging career, he might’ve surpassed Shakespeare in quantity, but…

One more Graph!

Another Straight line! So, the number of words he’s using is pretty constant. There’s a small cliff in September, where he started writing bigger posts, before getting back to his normal routine of smaller posts…

General Character Usage Statistics

2,103,816 Characters (or more precisely, Unicode Code Points) were used by him, including spaces, full stops, and all sorts of punctuation. That makes an average of 874 characters a post, 4706 characters a day. Poor keyboard. Poor Fingers.:) He used an average of 4 characters a word, which seems okay to me. This excludes HTML markup.
Unicode Code Points 

Yippee!! Another Straight Line!

That ends it for today. As I said, today’s post is pretty tame. As my Math teacher once said, Data is not interesting: Information is. Today’s post is data. Coming posts are information:D

Lot more charts to come!

You know those teasers they air at the end of every episode of most Animes? Here’s the equivalent:

Teasers

• Scoble linked more to Ze Frank than to Scoble Show (The domain, that is. It redirects to podtech)
• Before Joining PodTech, he’s linked only 7 times to Pod Tech. Today, PodTech is at #3 in the total number of links, and is actually #2 if you count in links to scobleshow.com, which redirect to PodTech.
• In a move that confounds me, he had more closing parenthesis than opening ones!
• The top ten sites he links to make up about one fifth of his total links.
• He has linked more to typepad.com hosted blogs than any other. Livejournal blogs are so low on the list you’d think they barely exist.
• Wordpress’s WYSIWYG editor saved him and his keyboard quite a lot of keystrokes. A lot.
• He used the word Microsoft more than the word Google:D
• He used the word I about twice as much as the word You. That egoistical bastard:D

Part I coming soon!

Any guesses on the most linked to site?

P.S. Incase you are wondering, I did respect the robots.txt file at scobleizer.com.

P.P.S: I have a whole mountain of data, and am pretty short on Ideas on what to do with it. Any suggestions?

P.P.P.S: Scoble’s now busy with a political campaign. Let’s see how long it takes till he finds this post by an almost unknown Z-lister (if he ever does!)

BSNL kicks up usage cap and bandwidth

That means, my usage limit is now 2.5 GB instead of a crappy 1 GB, and the speed can go upto 2 mbps! Was the guy who made that drunk?:P

The perfect new year’s gift:D

Found via Krishnaa.

Brad Adams is the guy who’ll forget more about the CLR than you’ll ever know: He built parts of it! Today, he has a post on building consensus within a team. Though it’s directed at a PM, many parts of it apply to normal life as well.

I’m not quoting anything, because that means I’ll have to copy paste it:D Head over there and read the full thing.

Phil Plait aka Bad Astronomer is an awesome astronomer and a damn good writer as well. Infact, he was the first science blogger I read (ever!) and is still the one I check first.

Today, he posted his top ten astronomy images of 2006. Awesome read/pics. If I had the money, I’ll print them up and post em over in our school, not that I’m not going to do it anyway.

Now, that reminds me that I have to head off to Wikipedia…

Maybe, BadVista.org is a joke?

Let us see. They link to gNewSense and utotu as the Linux Alternatives to Vista.

But, there’s something wrong with utotu.org. First up, you get a security warning about the SSL certificate being from an unknown authority. Why the heck do you need SSL for a homepage?

But, it gets funnier. The site is actually in Spanish! I mean, huh? I guess maybe you need to know Spanish to be pure and free? And, maybe utotu means something funny in some other language?

And gNewSense. Say that fast five times, and you end up say Gee-Nu-Sense. That is Gee, Nuisance!

If this set was a joke(I sincerely wish it is. I respect Linu(x|s)), then it’s a particularly elaborate and well executed one. Kudos!

Asa Dotzler too thinks this is a joke, but the commentators on that post don’t think so. What’s this world coming to?

Update:

Seems like they published the scripts they used to make Ubuntu into the gNewSense. Funny how it requires 35GB of free space. Now, the only thing that we need to show that this was a joke is an official announcement from somewhere…

When the Coke advertisers created Santa, they had to make it red, right? Though according to the site, it was red before, I guess Pepsi would’ve made him blue…

Merry Christmas people!

Via Sunil Jagadish, the FSF launches BadVista.org. The purpose is clear from the url, ain’t it?

I just gave it a visit, and the first sight I had got me ROFL…

I had to clip the image. Here’s the full text:

This website wants to run the following add-on: ‘MSXML 5.0′ from ‘Microsoft Corporation’. If you trust the website and addon and want to allow it to run, click here…

From the website:

Can you set yourself or your company free? Can you ever be free from Microsoft?

Oh the Irony! Oh the Irony! I can’t stop laughing! Oh my!

Yeah, I do know that it wasn’t the fault of the FSF: I think they just used some JS code that accessed the XML DOM. And, if I had used Firefox, I wouldn’t have seen it. But, still….:D

Sudar, if you don’t know already, is one of the persons who’s responsible for what I am today. He was the one who got me interested in computers.

Yesterday, he underwent LASIK Eye Correction Surgery. He’s been wearing glasses and contact lenses from ever-since-I-knew-him, so I guess it would’ve been very liberating to go out without something forgein on your eye…

In his blog post day before yesterday, he asked me to post updates to his “condition” on his blog comments. Instead, I’ll just post them here and link to it from his comments.

Updates:

20/12/06: The surgery goes fine, and he returns home on the same day! Information right now is patchy, but I’ll update as soon as I find anything.

23/12/06: Looks like everything’s fine. He’s not yet seeing TV or even reading books however.

27/12/06: I found his comment on Aswin’s blog! Is that three weeks for him?:D

Other than the horrible colors and the misuse of the <font> and <table> tag with no hint of CSS, there is one almost reliable way of detecting wether a page was done in frontpage.

The <meta> tag. In all the pages that frontpage generates, you’ll find this:

<meta name"GENERATOR" content="Microsoft Frontpage x.x">

While Frontpage puts it there by default, an enterprising user could hand edit the html and remve  that line. But, everytime the page is saved, this comes back, unless you have “preserve html” enabled.

What this means is that, if the page is designed by someone who doesn’t know to get rid of that line, then it will show up in the final markup…

While I have nothing against Frontpage being used by Students and the occassional computer user, I do STRONGLY object to that being used in a government website with transactions by the millions going on.

I just changed my theme from Freshy by jide to Tulip Time by Highland Design. Though the old theme was pretty colorful and, well, “fresh”, It just wasted too much space for me. So, I was on the lookout for a three column theme, and after hitting upon a lot of themes, I finally found this theme. It was love at first sight, and kudos to John, especially considering that it was his first theme!

The cool thing about this theme being, It gives more prominence to the blogroll on my right, and the calendar. The calendar reminds me that I’ve to post more often, and the blogroll reminds me that I’ve to read more:D Hope I’m able to do both…

The only thing I added? Categories in the Blogroll now show up! And, the Meebo Me! Widget as well…

Show up, and comment!

Microsoft just got themselves a new home page.

Took ‘em long enough!

But, I think I like the new one better. The old one was pretty crowded and portalish, to say the least. The new one looks more like a blog’s home page. Heck, it even has a glass background with a collapsible sidebar! But, anyone think of any reason WHY the sidebar would be collapsed?

What do you think?

UPDATE: It seems that this one’s been on beta for quite some time.

XNA Studio express has RTMed!

For those of you who don’t know what XNA is about(shame on you!), it’s like VB 1.0 for Games, but strangely you’ve to program in C#. So, instead of mucking around in DirectX/OpenGL and C++, you muck around with C#.

Now, the only thing that I don’t like about XNA is that the IDE is C# only:( Though I am able to use it in VB.NET as well, you won’t have support and some IDE specific features I guess…

And, if you haven’t, download it here!