You might have noticed already that First Person has two new widgets. The first one shows the articles I have shared on my Google Reader shared page. I first tried Reader some time back, and I have been a major fan of it ever since. I make it a point to check my Reader at least once a day every morning (usually more than that). The one feature that sets Google Reader apart from other feed readers is its sharing feature. Whenever I like an article I am reading, I just click on the ‘Share’ button and my friends in GMail who use Google Reader, get to read that article, and it also appears on my shared items page. This concept of sharing is great, and provides an excellent way to read good stuff and discover new feeds. But not everyone uses Google Reader, and while I have been persuading many of my friends to start using it, they don’t seem too enthusiastic. It seemed such a sad thing that only a few Reader friends could see my shared articles, until I saw that good old WordPress had a feed widget and my shared items page offered a feed! It didn’t take me long to put two and two together, and now, you’ll be able to see what I am sharing in Reader right here on my blog! That way, you’ll never be short on good stuff to read.

The second one is also an RSS feed widget. This one shows you my latest tweets. I rambled on some time back about how I wasn’t writing as regularly as I wished to, even though I had lots to say, and that’s when the microblogging concept seemed useful. I had known about Twitter a long time back, but it simply didn’t appeal to me at that time. Now however, I am giving it a shot. You can see my tweets in the widget here, or follow me. You can find me on http://twitter.com/prasannapandit.

Also, my blog, First Person completed one full year this August. So happy birthday First Person, and all that.

I came across this interesting news item today, that made me sit up and take notice. A 22 year old Orkut user was arrested in Pune, here in India for posting “derogatory” messages against Sonia Gandhi in an Orkut forum. You can also read the IBN news article here. It seems that people from the Congress party lodged a police complaint against this guy, and the police contacted Google, which in turn was only happy to give out personal details of its offending subscriber “in compliance with valid Indian legal process”.

This is an excellent example of what I was talking about in one of my articles about online privacy. I haven’t read what the accused had posted on Orkut, so I can’t really talk about that. But one thing is for certain. If you write something which can be deemed offensive to someone, all it takes is a simple police complaint to land you in prison – don’t expect Google or any service provider to protect you. Because they won’t.

I repeat, constant vigilance!

The Internet is well known for the anonymity it provides. All of us have enjoyed this anonymity at some point of time, haven’t we? There’s this feeling of privacy, of being secure as long as we are not foolish enough to divulge our personal details. Everything, from searching for illegal stuff to finding maps of frowned-upon places in your town – it’s all okay as long as it’s done on the Internet, right? Or are we falling prey to our own complacency as we spend those long hours on the Internet? How many of us end up leaving a trail of all our online adventures?

I usually wonder about this when I see companies like Google and Yahoo providing a host of services all under the umbrella of a single account. Let’s take Google for example. Let’s say you have signed in to your GMail account in your Firefox browser (as we all are, at some point during the day). And you want to do a web search, so you open a new tab by clicking on the web link at the top. The funny thing is, your Google session is still active (you can see your GMail ID at the top) and all your web searches are being recorded. In fact, you can see your past searches and trends at the web history page. That includes all kinds of searches – web, image, blogs, books, products – you name it. You frequently chat with your friends and family on GTalk. You open Google maps to find the route from your home/office to different places. You have all your photos tucked away in your Picasa account. Your Google Reader feeds give you your daily reading material. Your credit card info is in your Google Checkout account. Your orkut account lists hundreds of your friends. Your plans this week are in your Google calendar and your documents on Google docs, some potentially confidential and possibly incriminating, I might add. And have you ever wondered how the “relevant” text ads in GMail get their “relevance” from? Isn’t it a script that scans your personal mails looking for keywords for their text ads? So what’s wrong with all that, you ask me. But there is. All your personally identifiable information is on Google’s servers and all your search escapades can be mapped to a single account – yours. Anonymity on the Internet, anyone?

Don’t get me wrong – I am not saying that Google is out to cash in on your personal information. Not at all. I am one of the many people who remain grateful to Google for all these wonderful products. Heck, I use Google services as much as anyone else does and I can’t imagine my day without it. All I am saying is, a truck-load of your personal data is there on someone else’s machine, whether or not someone takes advantage of it. And that data is only as secure as the company wants it to be. I am sure we all agree that there will always be people in any company who will have the clearance to look at all the data residing in their company’s servers (those are the people with keys to the fabled executive washroom). Is it so darn difficult to imagine someone (the government, the terrorist who has kidnapped the CEO’s niece, the dictator who has nursed hatred towards you ever since you beat him in ping-pong back in high school) who arm-twists that company to trace and divulge details about you? I am not saying switch off that modem and cut yourself off from the world. All I am saying is you need to be careful, and see where you are putting all your data. Just do a bit of a background check on the company before you trust it. And decide what data that company can be trusted with. Never give away your passwords to any of your existing accounts (for those of you who just said “duh!”, here’s an example : a particular social networking site asks for your GMail password when you sign up, presumably to invite people on your contact list. And you can’t skip the step. I get invites from my friends every other day!). And of course, when in doubt, just say no thanks, or log off.

Constant vigilance!

So here’s the thing. How do you search the web for all images of a specific resolution, irrespective of the title? I decided today that I wanted to change the image header on my blog (the wide image that appears just below my blog’s title). I needed just a flashy, colourful and eye-catching image; I didn’t have anything specific in mind, as long as the size was precise. Of course, I could have just used one of the bigger images I have and cropped it to 780 X 95 (that’s the resolution I want), but somehow I didn’t like the idea. And so I turned towards my trusted image search partner, Google Image Search. That’s when I got a big surprise. There’s no way I can specify the pixel details (a X b) of my image. Just putting “780 X 95″ in the search bar doesn’t help much. In the advanced search, all I get are the options – Small, Medium and Large. So what category does mine fit into?

I then turned towards Yahoo. This one seems exactly like Google except that I get a Wallpaper option for size too. Big help. I looked at many other image search engines but none of them (including Microsoft’s Live Search) have a solution for my problem. Interestingly, Altavista offers preset resolutions (640 X 480, etc.) to filter the search, but a queer one like mine isn’t listed there.

I finally found what I was searching for when I typed “780 X 95″ in Google (web search, not image). It took me to a website called Free Web Page Headers and that’s when my search ended. And yet, it’s funny how you realize that advanced options are not really that “advanced” only when you want something that’s not listed there.

Update: Reader Teebot offers a good solution. Just use Altavista’s image search to search for images of one of the available resolution options, and then modify the URL to your required resolution. I tried it, and it works. So how about a front-end for custom resolution, Altavista?

Update 2: Just insert imagesize:axb followed by your search keyword in Google Image Search, a and b being your resolution. For example the query imagesize:640×480 mountains returns images of size 640×480 for the keyword mountains. Thanks to reader Rizwan for this.

Edit: Reader now has a search bar. And you can hide the side navigation too.

I finally tried out Google Reader today (don’t know what took me so long). Reader is Google’s RSS/Atom feed reading tool, which allows you to subscribe to websites. This means that whenever the website is updated, you’ll be able to read it (full articles or summaries as the case may be) inside Reader. Obviously, it’s a great way of keeping track of new material posted on websites without having to open each website everyday. Of course, the website has to offer feeds for this to work.

So how good is Reader? Like all things Google, Reader’s biggest plus is its simple and easy-to-use UI. It took me less than two minutes to subscribe to a bunch of websites. You can subscribe by entering the feed URL of the website and if you don’t know it, you can just type the keywords and the search option will give you one. For certain social sites (Blogger, LJ, Flickr, etc) you can simply enter the user name of your friend, which makes the whole thing very simple. I see WordPress missing from the list, so you’ll have to add me manually. Categorizing them is also simple enough and in an instant, all my friends’ blogs went into the Blogs folder. The amazing thing is, you can actually share your Reader with others using the URL Google provides you.

On the downside, Reader cannot recognize if you have already subscribed to a site if you have changed its name. Of course, it is still under development, so I’m sure it will be fixed (if it wasn’t intended to behave like that).

For those who don’t want to use browser add-ons to read stuff from their favorite sites, Reader provides an excellent option, since everyone has a Google account anyway. To subscribe to this blog, use the RSS URL http://prasannapandit.wordpress.com/feed/ .

A couple of years back, a gentleman from a US university gave a talk on genre based searching for web pages in our college. He highlighted the shortcomings of normal text based searches that search engines like Google do. At that time, I really didn’t find any problems with the results Google presented. I have been slightly wiser ever since my search for Eclipse (the software) ended up in Google trying to enlighten me about a construction company called Eclipse Architectures and about solar and lunar eclipses .

Let’s take another example. Let’s say you want to find out who Matthew Perry is friends with in real life. And so you type ‘Matthew Perry’s friends’ in the search field in Google. But there’s a problem. Matthew Perry is one of the stars of the hit TV sitcom FRIENDS. So obviously there are hundreds of websites containing the keywords ‘Matthew Perry’ and ‘Friends’. The results that are displayed are not what you wanted at all. Putting them in double quotes slightly improves the situation but you’re still stuck with irrelevant links.

Now assume a situation where you can actually specify which category your search must be directed towards. If you wanted to know Matthew’s role in FRIENDS, you would choose the category ‘TV shows’ and for the above example your category would probably be ‘People/Community’. The websites would be indexed according to their genre and the results would be displayed with the ones in the specified genre being displayed first. Sounds interesting, doesn’t it?

A moment of reflection shows how it might possibly be done (I thought about all this last night instead of counting sheep, so don’t blame me if my “solution” leaks like a sieve). Firstly, we would need a list of all the popular genre and a list of key words that help in putting a site under that genre. For example, the probability of finding the words TV, cast, episode, friends, prison break, etc. would naturally be higher in a website dedicated to TV shows. Of course, doing this step is the hard part. Once categorization is done, searching of the required text is done only on the web pages which have been put under that category. Many papers have been presented on genre based web searches. Here’s one.

It would be really cool if we could have this kind of searching. But until that happens we’ll have to keep rewording our search phrases in Google until we get a satisfactory result.