Jan 27, 2009 Posted by john on Jan 27, 2009 in Software & Media | Comments Off

Why Cuil Should Buy Iterasi… No, Seriously.

Ahh, the power of blogging. It’s one of the things that makes the internet great. Anyone can sit in the comfort of their own 40-hour/week job and detail how they are the only person who truly understands the large-scale economic trends that drive our global financial markets. Anyone can become an analyst of complex political interactions… or even corporate strategies.

I think this tendency is very important to the blogosphere, because without the ability to become an arm chair CEO, people might begin to take us too seriously. There are always blogs telling us why Facebook should buy Twitter, or why Facebook should by Xobni, or why Twitter needs to support Laconica. Most of these suggestions are at best an interesting idea, at worst they are complete non-sense.

The suggestion that I’m making, however, is spot on the mark. Why? Because, my friend I know what I’m talking about.¹

What I am talking about is this: In the new landscape of the web, Cuil and Iterasi go together like bacon and, well, anything.

What The Hell is a Cuil? (and who are these Iterasi?)

Let’s take a minute to define terms for those who are not ridiculously geeky. Like always, techno-geeks are welcome to skip over this incredibly well written and super-interesting portion of the post.

Cuil

Cuil (A Gaelic word for “knowledge,” pronounced: “cool”) is a web search engine and so-called rival to Google. It was started by Anna Patterson– former Google employee and creator of the search engine that Google still uses– with the intention on building upon what the founders see as search engine stagnation.

Searches for Cuil on both Google and Cuil yield basically a lot of initial “They’ve arrived” posts, with little since then but the sound of virtual crickets.

Like many bloggers, I’ve written about Cuil previously because I was super excited when it came out of stealth mode early last summer. Also like many bloggers, I’ve pretty much forgotten about it since then.  Despite the something like $55 million in funding and the initial excitement about a better search option, the interwebs have been nearly silent since July 08. In fact, searches for Cuil on both Google and Cuil yield basically a lot of initial “They’ve arrived” posts, with little since then but the sound of virtual crickets.

Now, Google could be selectively de-valuing results about its rival just a Cuil itself could be de-valuing something that they don’t want to give away. From my CEO’s seat in the arm chair, however, it seems like Ockham’s Razor might cut through the arguments to reveal, somewhat sadly, that Cuil is no real threat to Google.

I could go on for days about how cool this is, wouldn’t it be more fun if it were?

Iterasi

In a nutshell, Iterasi is a social bookmarking site. But that’s a nutshell about the size of a filbert, and Iterasi is more like a coconut.

When you bookmark a site in Firefox, Explorer, or some other browser, you’re saving a link to that site. The idea, of course, is that you can get back to it eventually. Social bookmarking sites allow you to store your bookmarks in “The Cloud” and to share them with others. Ma.gnolia is one such social bookmarking site. The utility of a site like Ma.gnolia is that you can access your bookmarks from anywhere on the web, rather than wishing you’d brought your computer with you when you want to get back to a site. It also provides nice organizational features. Using Ma.gnolia, you can store those links, title them, index them, and share them with people.

The problem with this is that, well, websites change. Someone may come to Positively Glorious! to read an article about something, but want to save more than the address- they may want to remember some other contextual feature like specific text, the layout, or a particular image. If they merely saved the address http://positivelyglorious.com, then the next time they arrive at the site, it will likely be different. Not only will it have a different article on the front page, the very layout of the front page might be changed. If they are viewing a particular article, perhaps they wanted to capture a sidebar element that is now gone.

By contrast, Iterasi let’s you keep the whole page. Everything. Images, layout, advertising. Everything. In fact, it takes a snapshot of that page and stores it for you, like this:

No more do you have to go back to a site and see that you have the address- but not the information you wanted.
The image you see is an embedded version of the page that is stored in the Iterasi cloud. Clicking this image will bring you to the archived version in my Iterasi profile, where you can see all pages that I have “noterized” (e.g. bookmarked) and marked as public.

No more do you have to go back to a site and see that you have the address- but not the information you wanted. It’s especially good for capturing web moments that are short-lived, for instance if you write an article that ends up on the front page of Digg, an event which may never happen again. You can take a snapshot of the page, as you see it. Right now.

I could go on for days about how revolutionary Iterasi is, as a company as well as a concept. Suffice it to say that it’s just about the coolest thing since sliced… arrays.

Cuil and Imagery

So, now that we know what Cuil is, let’s take a minute to describe some nice improvements that Cuil makes in search results. I’ll briefly touch on some things that I like, but mostly I want to talk about one of the most potentially interesting (and currently virtually pointless) features of a Cuil search: Images.

When you search Cuil, you are presented with a lot of information right on the first page. Cuil search results are returned in multiple columns, often with an AJAX-y drop-down list off to the right and a menu bar at the top. Both the menu bar and the drop-down list give you instant access to more focused results that are categorized based on the information Cuil found in the search.

The following is an example of a Cuil query for the term “bookstore:”

Currently, the choice of images presented with any given search is random at best, and completely confusing at it’s worst.

Note that the drop-down list to the right says “Explore by Category,” and the top menu bar gives more refined category choices as well. Ignore for the moment that you may not care about searching within “Universities and Colleges in Pennsylvania,” “Christian Bookstore,” or “Bookstores in Canada.” It’s not so much the results, but the concept that’s beautiful. Right here, we are immediately presented with  2-3 times the information that Google has provided, and all we’ve hit is one button.

Cuil also places images next to most search queries. The “bookstore” search is a good example of this. I think that the use of images is one of the most potentially valuable aspects of a Cuil search. Unfortunately, it is also completely pointless. Currently, the choice of images presented with any given search is random at best, and completely confusing at it’s worst.

As an example, let’s do a Cuil search for your favorite blog on Mettacontemplation:

You’ll notice immediately that the first search result is this very blog (meaning that my SEO work is not completely pointless). That search result shows an image that, according to Cuil, is somehow representative of my blog. When I saw this, my first reaction was “… the hell is that?”

My second reaction was “No, seriously! What the hell is that?”

The Good, The Bad and… well, you know the rest

This image that Cuil chooses to represent my blog is one that I’ve never seen, never used, and that is completely unrelated to the content of my blog. Now, I’m not one to immediately discount good work based on one mistake. I mean, my blog is hardly the example of focused writing on a single definable topic. Cuil’s images are served from a secondary set of servers: cuilimg.com. I’m assuming the algorithm used by this system figured that my blog equates to “circus.” Who can blame it, really.

So I gave it another try. Here’s a search for “WordPress:”

This result illustrates better than most why Cuil’s current use of imagery is almost random, and why it borders on downright confusing.

There are some good images. The graphic used for wordpress.org is a random image of a boxed software product. Ignoring, for the moment, that you can’t get WordPress as a boxed software product, they at least correctly identified it as a software product. This is more that can be said for the WordPress Planet,  which has somehow become a way to set up your own Yahoo site. Perhaps the connection there is that WordPress will allow you to set up a website?

It gets really bad when we look at the graphic for WordPressMU, the multi-user installation that allows many people the ability to create a blog under one installation. WordPressMU (MU for “multi-user”) is the software that runs wordpress.com, edublogs.com and even redivivablogs.com. Oh, one more thing: WordPress is software developed and owned by Automattic which, last I checked, has not been bought by Microsoft. Thus the question: “Why is the Windows logo used to represent this software?”

The questions continue: Why also is an image for RailsChina used to represent WordPress (written in PHP)for the iPhone? Why is there a random graphic of an eagle used for WordPress Themes?

And then it gets ugly. Just to make sure you see this, in case you can’t scroll down enough, I’ll give you a blow-up:

Yes. That is, for all intents and purposes, the back of a naked man flexing. Now, don’t get me wrong, I like looking at beefy naked men as much as the next guy, but is there even a tangential connection here? Lorelle on WordPress is a site offering “tips, advice, and techniques for blogging and using WordPress and WordPress.com.” Does Cuil think that ”WordPress” similar enough to “Bench Press” or something?

I could likely go on for pages about how wrong this is, but I think everyone will agree that the phrase “What the hell, Cuil?” sums it up nicely.

Iterasi’s Awesomeness: Useful Imagery and Website History

The entire concept behind Iterasi is, in their own recent words, so that digital history will not simply disappear. In the latest post on the Iterasi blog, they discuss this in relation to archive.org:

At iterasi we have taken a different approach. We do not know where all the great nuggets of information are on the Internet. But we think you do. So we allow you to build your own archive. We hope you archive pages you find valuable and, if appropriate, share them publically so that others may benefit from what you found. They may contain historic events or they may contain your favorite recipes. Whatever you archive, I bet others may will find value in what you archive.

A great thing about Iterasi is that they store all noterized pages on their server, and provide  items like screenshot images, shrunken urls, and embeddable frames for use in other pages. All of these link to the resulting page at iterasi.com (clicking on any of the above images will illustrate that).

Think about Iterasi’s words for a moment. They do not know where the great nuggets are, but we do. We search them, we Digg them, we Twitter them, and we frequently bookmark them.  Putting this all together  and combining it with archiving only seems natural.

“Your Search Results, Now and in The Past”

If you look at the code for a Cuil search results page, and look closely, you’ll notice that the images are pulled from that cuilimg.com server I mentioned. Since the concept of pulling images in parallel with the search results is already in use, and the infrastructure to put it in place is already built, one could just pull screenshot images from Iterasi to populate that page. The result would be something awesome

But this isn’t it. This is a very rough mock-up of what could be returned as a search result.² WordPress.org shows a screenshot of, well, wordpress.org. WordPressMU is no-longer associated with Microsoft, and Lorelle on WordPress could no longer be, sorry ladies, a beefy naked guy’s back. Although it doesn’t work on this mock-up, each actual image could link back to the Iterasi archive. Thus, you have the choice of seeing the site, or the archive of the site.

Not only can I immediately see the site I’m looking for, I can view the site’s immediate past. And all of this happened merely by me pushing one button: ”Search.”

There’s another value to doing this. Iterasi maintains the ability to schedule a page noterization. Thus, the functionality could exist that estimates the average “change time” of a page and then schedule page snapshots regularly (SEO gurus could even come up with a standard <meta> tag attribute to hold “change time” on the client’s side). Then, the image on the result (or another field such as “last snapshot”) could link back to the Iterasi archive, while the text of the result would link to the actual site.

Ahh… now we have some incredible search functionality (and something a whole hell of a lot better than Google’s crappy “Cached” option which is very nearly useless). Now we have a search result that shows a screenshot of the resulting sites- possibliy allowing instant recognition- and a web history of the site in an archive that’s linked to the page. Not only can I immediately see the site I’m looking for, I can view the site’s immediate past. And all of this happened merely by me pushing one button: ”Search.”

Let me say that again, loud and clear. All I did as the user was push one button!

Coda (and disclaimer)

The flies are buzzing, so I’ll step away from the dead horse carcass and leave you to your thoughts. But first, I’ll give a bit of a disclaimer.

I have no interest in Iterasi.

That is to say that I am very obviously interested in Iterasi (i.e. they are friends who would probably buy me a cup of coffee) but that I have no actual interest in Iterasi (i.e. they are not friends who would buy me an oceanfront villa on Kauai). I think their concept rocks, and their community-based organizational structure is something to learn from. I also think of them as friends (or, a friend. I’m still a bit confused on that, actually).

That said, there’s a really good chance that this article results in a reaction similar to “What an idiot! Well, he’s out. Who’re we going to give that oceanfront villa on Kauai to now?” In otherwords, I don’t know if Iterasi wants to be bought, and suggestions like this might just piss them off and make them write cease and desist letters telling me to mind my own damn business.

If that’s the case, it’s a good thing that I’m one of those “armchair CEO” bloggers that no-one takes seriously. I mean, about 17 people are going to read this post anyway… and 15 of them will be eating cold pizza at 11 am while playing Bagpipe Hero in their underwear.

Ahh, the power of blogging.

1. Okay, before we blow anyone’s top, feel free to scan ahead to the  disclaimer. I’m not starting with that because I want to get right to the bacon. [↩]
2. This post has been on hold for about 2 weeks while I try to make time to create an actual search result page, I finally gave up and roughed it in. [↩]

Comments are closed.