I just stumbled upon a possible new Google search layout:

I’ve seen this layout once before, but slightly different. Note that the horizontal menu has been replaced by a vertical one on the left hand side, with options to search Images, Maps, News, Groups, Videos, Books, and Froogle on the top and bottom of the page. Also, the ads seemed to have disappeared. I’m not sure I like how the left menu is eating into the search results real estate–it would be better to stick with the traditional horizontal menu near the search box.

Is anyone else seeing this? Or has my IP yet again been singled out for beta testing?

It looks like Yahoo has caught wind of the Rollyo effect, and have launched the Yahoo Search Builder, which allows users to tailor searches based on their interests.

The idea is to foster the creation of focused search interfaces. For example, if you’re into computer games, you can make a game-focused search for your website. The creation of the search is simple, and is very similar to services like Rollyo. On a basic level, search is focused by specifying a list of sites that are to be searched. Yahoo also allows you to add keywords that will be included in every search, exclude keywords, and exclude certain sites from being shown in the results.

After customizing the look and feel, users paste some code on their sites to publicize their searches.

Yahoo’s effort is lax compared to that of Rollyo, which gets much deeper into user involvement with the use star-powered searches from celebrities and featured searches. I’m sure Yahoo is working on the community aspect, but, it may be too little too late.

On a more nit-picky point, is anyone else bothered by images implanted in text forms? At first, I mistook the Yahoo search box for an image with a border around it (so I was wondering where the search box was). They need to either make the image lighter, or get rid of it entirely to avoid confusion.

TagFetch is a search app that will do a quick tag search of some major sites that support tags. The current set of sites include flickr, reddit, del.icio.us, Technorati, Feedster, YouTube, NewsVine, and Tailrank. The feature set is small right now, allowing you to do tag searches, and save searches (if you log in) for use later on.

But, the key is TagFetch’s speed. When I’m trying to track a subject, it can get cumbersome to search through a lot of tagging sites. Most of the time, I find that tag searches are pretty slow, especially for the larger sites that have huge tag clouds. It’s nice to have TagFetch do all the heavy lifting.

Some features that I would like to see:

• Searches for digg, other photo and video sites, wikipedia
• A more robust “View More” feature that expands the search results (instead of clicking through them four at a time)
• RSS feeds for searches that aggregates results when they change
• Suggestions for related tags

Overall, TagFetch is a nice, simple service that works as advertised.

There’s certainly been a lot of blog searching buzz in the past few days, including announcements of new features from the big dogs like Bloglines and Technorati. Among the buzz is the newly launched, Australian based gnoos, which is trying to break into the blog searching world with a slightly different angle.

Their interface is as clean and sleek as their response time is quick. These are key factors in persuading users to return to your site time and time again. Michael mentioned that he would be able to spend quite some time on gnoos without being fatigued, and I whole-heartedly agree with him. There’s just something pleasant about the design decisions that were made for gnoos’s front end.

Some of the differentiating features include the ability to tag, rank, comment, and show the text of each blog post result. The first three features are similar to swicki’s social search aspects, the difference being that gnoos isn’t necessarily tied to a specific site, and is more of a general purpose blog search. Technorati could learn a thing or two about the type of community search that gnoos is bringing to the table via user interaction with search results.

The fourth feature is gnoos’s ability to pull up a blog post when clicking on the “more” link. The entire blog post is shown inline with the search results. Depending on who you’re speaking to, this could be a blessing, or a wolf in disguise. gnoos is allowing visitors to view blog content without visiting the blog or RSS feed. For sensitive content producers, this could be controversial, either due to financial or copyright concerns. gnoos does have a copyright policy in their Terms of Use page, where they say that they

will never knowingly infringe (or be involved in infringing) the intellectual property rights of others.

In addition, gnoos might run up against the same wall as Google did with their page caching, where private posts that were deleted might be easily recovered. Hopefully gnoos won’t ever have to deal with privacy and copyright issues. Personally, I don’t mind my posts being on 3rd party search engines. The way I see it, the more exposure my content gets, the better.

Perhaps the most interesting tactic that gnoos is bringing to the table is localization. gnoos is heavily localized to Australia, and shows it by allowing visitors to filter by “Aussie Blogs” and “Aussie Media.” I find that localization is something that has been overlooked by many startups, and indeed, it is the secret weapon that gnoos has to gain significant market share in Australia. What would be cool is for gnoos to take the lead in localization and start localizing blogs for other parts of the world. However, I doubt that this is their strategy.

Currently, the content on gnoos is much smaller than Technorati. Their fast querying speed may be a result of their small database. If this is the case, we’ll have to see how well they scale up as they index more of the blogosphere.

Good luck to the underdogs!

I had the opportunity to try out the beta version of Qube v2.0. Made by the people at Qelix, Qube is trying to enter the search arena, but, with the concept of “browserless search.” I know that this has been done before in the past by other companies, but, I can’t recall any recent forays into this space (those who do know some, please let me know).

Qube is a downloadable application that is run in the background and checks to see if you’ve hit any hotkey search combinations, and, it will automatically search the selected items. For example, I can hit Ctrl + Right Mouse Buttom to search the current text that is below my mouse cursor (in this example, the word “gain”). Then, I can bring up Qube’s search window, which will have the web search results for “gain”, and also a dictionary definition (if available).

There are also various options for searching selected text, phrases, or manually entering in searches. The search capture algorithm tries to be as flexible as possible, allowing you to search not only normal text, but window titles, text in GUI buttons, and other elements that you can’t normally highlight in the OS.

Update: Apparently, I missed the QFlyer feature altogether. Whenever you do a search hotkey when Qube is minimized to the system tray, the QFlyer will pop up and show you the search results, as seen below. This is the whole “browserless search” aspect of Qube, allowing you to quickly scan search results within any application. Thanks to Ebrahim for pointing out my oversight.

During my usage, I found their search capturing to be a bit buggy, capturing text that wasn’t right below my cursor, or garbling the selected text before searching. They’re still in beta, so hopefully those issues will be resolved upon release.

Qube currently searches with MSN, and, they have plans to roll out searching with FAST (Yahoo). I’m a little puzzled as to why they are excluding Google from the mix, since Google has by far the majority of the search market (and for good reason). Adding other searches like Wikipedia or Answers.com would be useful as well.

The interface for browsing your results is snappy, allowing you either view each result page in a preview window within the Qube application, or launching your default browser to the search result. Manually typing in the search box will instigate a Google Suggest-like dropdown box that gives you suggestions for your search terms.

There are few more features that haven’t been rolled out yet, which are Siel and AdRoll. Siel is a community driven search, which appears to be in the same space as Swicki. The details of AdRoll are under wraps until it comes out of development. Ebrahim Ezzy, Co-Founder of Qube, has disclosed only this:

All I can say, for the moment, is that we’re not making ads sounds like a feature. AdRoll is unique concept still unattempted. Not link exchange, pop-up network nor a nasty prank. AdRoll aims to enable free, point-based advertising in a proactive manner. It may sound overdramatic at the moment, but we surely know what we mean when we speak of it.

I’ll be looking forward to see exactly what AdRoll is. But, a guess would be that AdRoll involves using the context of searches to help deliver more relevant ads, perhaps integrated with Qube’s interface in a non-interfering way.

Overall, I think that Qube has some good ideas about providing flexible ways to search outside the browser. The only problem for me, is that when I’m doing research, I’m usually already inside a web browser. Opera already has a built-in right click context menu that lets you search with the major search engines. Perhaps other people who aren’t as net heavy as I am would save same time with Qube’s search grabbing features. Also, the QFlyer would save time for anyone in a web browser as well, since it’s a quick pop up window that appears near your search item.

As past history has suggested, the “browserless search” space has yet to produce a killer app. I’ll be checking in on Qube as more features are rolled out, and we’ll see where this goes.

Digg this post