Did YOU contract HIV during your last ***? Odds are greater than YOU reading this Post!

Looking at a news article on the odds of something good, bad or surprising happening to someone, I wondered what are the odds that a particular but random blog post would be read today by you or me i.e. someone in particular.

Delving deeper into the matter, I thought I should make certain assumptions about this problem before it starts to get beyond my mathematical/analytical abilities. So, the problem ended-up being a simple one to analyze, owing to the assumptions! Although, I would love to see comments deriding my analysis. And yeah, the HIV part comes in the end ;)

First, I would need a rough estimate of 'number of blog-posts made in a day' . It turned out to be easy. I found it in Technorati's quarterly State of the Blogosphere (April'07) report. In the article's summary it says, "...1.5 million posts per day."

Now, how many posts does 'that somebody, you or me' read in a day? The report says nothing about that. 5? 10? 20? Let me assume 21 posts, with no statistical backing.

So, is it 21/1.5 million? I'd not like to make matters complex but let's add another detail. If I start reading those 21 posts of mine - what 21 posts would I read? Most probably that link to each other. I assume probability of reaching a post randomly is 0. Thus, the 'Post in Question' that is to be read may or may not lie in my linked path or may lie too far ahead in my linked path and I may not reach that in those 21 posts I read.

I have 3 situations:
1. Post can not be reached as it is not in my path
2. Post can be reached as it is in my path but would not fall in 21 posts I read today
3. Post is reached

Matters can get more complicated, by taking into account that popular blogs are more linked and hence a post in them would be easier to stumble into. But, in all fairness to my intellectual abilities (excuse 1: c'mon man there is no statistical data?, excuse 2: the odds of this post being read are too low to take the pain :D), I assume that all 3 situations have equal chance of happening.

So, the proability is 1/3 * (21/1.5 million) or, 1/(1.5 million/7) .

Thus, the odds of 'You' reading 'This' post are 1 in 214,286. This link tells me that Odds of a person contracting HIV by having unprotected sex with someone who 'may or may not' have HIV are 1 in 200,000.

You have already read this post. Is it time for you to run to your doctor? Btw, you may like to take a look at coincidences when talking of large numbers here.

P.S. Please link back to the article if you want to reproduce it on your own site.

Disclaimer. The author is not claiming anything regarding process or probability of contraction of HIV/AIDS. Thus, reader is advised to take this article as a piece of humour and not medical advice.

Read more...

How to show Images uploaded on Blogger in a Modal Window using Lightwindow

If one has been following this post, he/she would be happy that 'phydeaux3' suggested the possible cause and a solution to the problem. phydeaux3's original comment in full body & soul can be viewed here.

Quoting phydeaux3's comment:

"Actually, the problem most likely resides in the peculiar way blogger image hosting works. By default, the large linked image is in a format (I'm using yours as the example and only showing the last relevant bits of the url)

/s1600-h/mpro.jpg

If you view that image directly, you'll find that Picasa Web Albums (which is the actual image host for blogger) makes this NOT an actual image file, but an html page, with an image embedded in it (view source). What that means is, the large image cannot be used as in in an image tag, but only viewed directly. But if you cut out the -h in the folder name
s1600/mpro.jpg
That gives an image that CAN be used in an image tag, and presumeable in your scripts as well (but cannot be viewed directly, as it will prompt for a file down load..yeah it's odd). So remove the -h from the last folder name and try it.

(there also may be referrer restrictions on images larger than 800px in size, but I haven't messed with it lately)."

Here, I put the above to the original test (without '-h' in the target url):
Without a doubt, it works! Hail phydeaux3 :)

Read more...

Lightwindow fails to run with Images uploaded on Blogger

Modal Window (Box) utility, Lightwindow - based on Prototype & Script.aculo.us libraries - fails to pop-up on clicking the (image or text) link if used with an image uploaded on Blogger. At least, it does not in my case. It works in case of other links (e.g. click on Lightwindow, Prototype etc. links).

To illustrate the point, here is an image uploaded on Imageshack (Click on it : works clean as a whistle!)...

Read the full story...

Now, here is an image uploaded on Blogger (Again, click on it - Caution: the page would behave abnormally & you might have to close or reload the page!)...

I hope, I am not the only one with the problem and someone would have the brains to sort this out. I guess, the problem is with how 'lightwindow' deals with the URLs or may be not! Nevertheless, it's a great library, and this little problem should not stop anyone from donating towards 'buy macbook pro' cause :P

P.S. It's worth noting that ModalBox, another lightwindow like library, also fails in the above case!! (May be the problem is with Prototype or Scriptaculous as ModalBox also uses both :|)

Update: We've a solution - suggested by 'phydeaux3' in the 'annotations' below - here is a link to a working demonstration.

Read more...

Adsense incompatible with Blogger?

When I tried to insert 'Google SiteSearch' code generated through Google Adsense into the template of my Blog on the recently revamped Google's Blogger - it gave the following error on saving:

Your template could not be parsed as it is not well-formed. Please make sure all XML elements are closed properly. XML error message: Attribute name "checked" associated with an element type "input" must be followed by the ' = ' character.

Since, the new Blogger had been in beta earlier - I suppose these quirks have already been discussed and thus, I got the solution i.e. to write: checked = "true" to get rid of the problem in the SiteSearch's auto-generated code. But, the bigger question is "Is the adsense generating non-standard code?"

Read more...