Searchwiki can only be an experiment

On November 22, 2008 (a week earlier in Europe according to reports), something strange started to happen with Google's biggest source of revenue - that ubiquitous service in which the monolith has virtually no competition - Google Search.

Around November 21, we saw a heavily publicized launch for what Google termed as SearchWiki. (Even the WSJ prominently carried this story.) Some claimed that it was not a "Wiki" at all since it was not essentially a socially-driven search system, while some others said that Google was just trying to copy Wikia. Read the Google Operating System blog post about the launch.

Few actually saw the fully-functional version of the modified search system until November 22. But taking us by surprise, by afternoon, Google made a hushed withdrawal, replacing the original search system. Read the TechCrunch post about the withdrawal.

Soon, Searchwiki made a reappearance in Google Experimental (Google's description mentions a cheesy "sound-effect" recorded by Sergey Brin himself - aargh!)
Worst of it all, the Experimental version didn't work - (It doesn't work the moment I am writing this blog!)

What really is happening at Google? The way it's happening, it doesn't seem intentional. The reception to SearchWiki has been very divided. While one side maintains that it will kill the manipulation of searches by SEO's and foster better search results providing for a combination of search and social sharing capabilities, the other side says that unmoderated commenting will prove to be a haven for spammers, and cannot do well because the mass customer is always "dumber" than the search engine developer.

My take? This technology is so new and immature that it should only remain an experiment. Search results are more vulnerable for manipulation in Searchwiki, say for instance, if a website starts hiring people to just promote their website for specific keywords. Of course, Google says that your preferences will not affect another user, but that doesn't make sense since we already have an array of bookmarking options. I believe that Google's plan was definitely to slowly transition to a user-influenced search system. Also imagine searching for "ExxonMobil" and getting the first page full of links to exxposeexxon.com. The "dumber than thou" argument also underscores the fact that Google should not impose something that's so noisy and complicated on every user especially given its reputation for simplicity. Obviously, the ramifications of such a search model can be huge since the internet is getting more and more search-driven.

It was not too wise of Google to launch such a huge change directly in its most used service. Actually it's much worse of a trip-up than what facebook did by not involving its users when it launched its new interface. In case this was just an effort to collect a larger sample of usage data, Google should've informed the users that it will be temporary. This unacknowledged blooper from Google is certain to fuel the argument that Google is foraying into too many things than it can effectively focus. Since every product Google launches leverages at least some share of Google's impeccable search capability, a faultering step here can be disastrous. Will Google be able to maintain the small-firm approach that it had 10 years ago even as it has become the biggest "portal" to the web, given the mess in which Yahoo has submerged itself? The task is tough for sure.

As an aside, an SEO professional came out with a script to turn wiki on and off, which was kind of neat, had Google stuck to its earlier plan. Also read: Networkworld critical about Searchwiki

Update: As of 11 p.m. on November 22, SearchWiki has made a full-fledged comeback. Now that it's going to stay in one form or the other, you should read this awesome illustrated guide to SearchWiki entirely so that you get an idea of the do's and don'ts in a changed search environment. I maintain my stance that SearchWiki should be allowed to gradually graduate from a experimental release - there are too many jitters as of now. And I'm not happy that they removed the "Note this" feature from search. I used to love the ability to privately bookmark my searches. Give it back to me, Google!

Update (Nov. 24): To de-personalize your search, add the string "&hl=all" to the address bar once your search result comes up. It not only removes SearchWiki but also prevents your web history from influencing your search. Get your de-personalized Google search Firefox plug-ins here.

Google Chatbots!

Translation chatbot:

Google has released 23 translation chat-bots. (If you didn't know, a bot is simply a (ro)bot). You just add any of them to GTalk (or any other jabber client) and send a message. They (some ultrageek translators who type at lightning velocities onto a gargauntan monitor filled with at least 1 million open chatboxes 24/7/365 and yada yada) will reply back with the translation within a second.

List of bots:

ar2en, de2en, de2fr, el2en, en2ar, en2de, en2el, en2es, en2fr, en2it, en2ja, en2ko, en2nl, en2ru, en2zh, es2en, fr2de, fr2en, it2en, ja2en, ko2en, nl2en, ru2en, zh2en*

To add the bot to your buddylist, use this id format:
[firstlanguage]2[secondlanguage]@bot.talk.google.com

Example: en2es@bot.talk.google.com

Sample chat:

me: hello
en2es: Hola
me: what's up
en2es: ¿Para qué son
me: welcome to ootty
en2es: Bienvenida a ootty
me: nee poda manda
en2es: Nee Poda Manda
Sent at 10:57 PM on Monday

*Explanation on the language codes:

EN = English, NL = Dutch, DE = German, ES = Spanish, FR = French, IT = Italian, KO = Korean, RU = Russian, JA = Japanese, ZH = Chinese.

Group Chatroom Bot:

Add any or all of

partychat@partych.at
partychat0@partych.at
partychat1@partych.at
partychat2@partych.at
partychat3@partych.at
partychat4@partych.at
partychat5@partych.at
partychat6@partych.at
partychat7@partych.at
partychat8@partych.at
partychat9@partych.at

to GTalk and follow the instructions at partych.at or http://techwalla.googlepages.com/ to create a closed (can be password protected as well) chatroom of friends of your own, and carry on a persistent (the group chat option in GTalk is not persistent) chat with them. We can be simulataneously active in 11 different groups at the same time using the ids above. Using /help or /commands on the partchat bot will provide online help. This was created by Akshay Patil of Google.

Sample conversation:

me: what the hell?
partychat: you are not in a party chat, for help using PartyChat, type '/help'
me: /help
partychat: To enter a party chat, type '/join chat_name [password]' (password may not be required).
To exit a party chat, type '/exit'.
For a list of PartyChat commands, type '/commands'

Find a more detailed treatment of Google's chatbots or IM bots here.
Pretty cool, ha? Enjoy!

Have you piped the internet yet?

Pipes. They are the next generation of mashups - yeah, if mashups present a programmable web, pipes take it one step further - they bring it to the layman, who doesn't necessarily have programming skills. Pipes bring to us that version of the web, which we create by searching, filtering, slicing and dicing it to make it look the way we like it - all using easy drag and drop tools. If Yahoo! could get one step ahead of Google anywhere, it's through its Yahoo! Pipes service, which offers a pipe editor interface.

Tim O'Reilly's article claims "Yahoo!'s new Pipes service is a milestone in the history of the internet." There are dissenters too, like Tim Anderson,who argues that Yahoo! Pipes lacks a business model and allows users to strip the websites off the advertisements which help them run, leaving the internet robbed off its chief driver. Well, I'll say, it's a great idea - let's not nip it in the bud! It's not even a year since it launched and it's going to be in beta for longer - let's give the innovators and business moguls of Yahoo! and Google and Microsoft time to think out how they could cultivate a winner out of Pipes. At the least, let's hope that Google and Microsoft come up with clones to take on Yahoo, instead of killing this innovation.

Pipes are just like giving the crayons and the paper to the little boy. The little boy starts drawing - random lines and different colors initially, but not so much a picture yet. Things don't work easily at first, but he's not one who gives up so easily. He pulls up a magazine and starts browsing through the pictures until he finds one which he can copy. Well, for us, NYT through Flickr is such a good picture to start with. When you click on the link, you see a map peppered with red dots. They all represent pictures from Flickr. How are these pictures relevant? Well, a New York Times news feed has been "piped" with Flickr using keywords, to give the pictures related to the New York Times content. Cool, ha?!

The little boy is impressed. Just above the map in the NYT-Flickr pipe, he finds a 'Clone' button and clicks on it. He's led to a Pipes editor. It's not so intimidating because it looks just like a Visio flowchart editor, or even a diagram object in Powerpoint. Well, he feels NYT is pretty high standard for him - he can't digest that as much as he does 'The Hindu'. He's been growing with the paper, and the paper, with him. He goes to the The Hindu webpage, and clicks on the 'XML' RSS feeds button. There comes a list of all The Hindu's feeds. He chooses Kerala and replaces NYT's feed in the pipe. There you go - He's just created his own pipe! Save it as a different pipe, and run it - That's all he needs to do to get a feel of how many articles in The Hindu have photos that correspond to them on flickr. Zoom into Kerala on the map, and have fun!

A Mashed-up World...

This is a 'worldly' greeting courtesy geoGreeting, a site powered by Google. Visit the howdy link to know more!

(Added 2-Dec) geoGreeting is just another 'mashup', which, for the uninitiated, is a technology to separate the presentation and the application layers of a webpage. This is quite similar to the J2EE ideology of Model-View-Controller. You can have different presentations for the same information, provided the information sources allow you to access them. And, when you mashup, you can combine information from multiple sources, and present them in a cohesive and sensible form. For instance, the 7 wonders mashup combines information from Google Maps and Wikipedia. This is possible because Google and Wiki follow an open-source policy. Similarly, an enthusiast could combine coffee outlet information from Starbucks.com, process the addresses through Google and come up with a Starbucks map. Now, he could get the weather information of each location and put that up too. This is one great advancement in web design and is a pointer to what's going to come up in terms of interactivity and personalization in the new internet. Programmable Web keeps track of all these mashups, and has posts on the newer ones that are coming up. It also has a experimental Matrix which is interesting.

Talking about mashups, and what thet means to the future of Internet is this Wiki mashup page. Now that wiki will tell you everything, I don't have to brag on about it!

You also get to know about the recent term Web 2.0, and why the likes of Sir Tim Berners-Lee reject this idea, though positive about what the Internet is becoming. And, you also understand the flip-side of mashing-up, when reliability gets compromised in an effort to pull data from various sources.

Here's my bonus - some interesting mashups I came across!
Flickrscapes
Xrez Gigapixel Images
Web Traffic Comparisons
GooCam
GahooYoogle Search

Hmm! Didn't I just tell you another device to fritter your time away? Just mash it up!