After that, my favorite startup and news source will be no more, and to say that’s a shame would be an understatement. Greg took the work he pioneered on personalization at Amazon and brought it to news, and the results were effective and useful, becoming a large part of my news discovery for the last several years.

Findory will be greatly missed by its fans, especially since there is nothing out there that can deliver as well as Findory did. I’ve read 3,362 articles through Findory, each one going a little towards making the service better for me, and without Findory, the web is a little less useful. Thank you, Greg, for delivering for the users the last four years. We’ll miss Findory.

Hopefully, whatever comes next for Greg will be even more exciting, and this time it’ll hit that magical combo that brings success. Good luck.

Comments

Tag:

New downloads of Opera Mini contain the Yahoo search, and older versions are also being switched over (and not all of them are pleased with the change).

(Found on Findory)

The new browser for the mobile browsers is new itself, Yahoo’s OneSearch (picture below). OneSearch works interestingly for mobiles by examining the search query and showing a set of results based on the type of search query. In the screenshot, since you are searching for 49ers, you are getting sports scores, stats, photos and news, while a search for iPod would default to shopping search and a search for a city would include weather, traffic and local news.

It’s a good idea, similar to what Ask X does in its sidebar, but I’m not sure I’d want to rely on it as my main search engine. Most of the time, I know what I want, and I know how to use search verticals, so I like to get the regular search engine to make my own decision. However, other searchers, just typing stuff on the phone’s screens, may find this to be the perfect solution for their searching.

Comments

Tag:

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

Greg , the man who built Findory after an incredible run at Amazon, announced that development will slow, and Findory will last through most of the year, but eventually it will fade away.

Here’s the comment I left on his blog:

Wow, I am so sorry to see Findory go. It has easily been one of my favorite online tools since the day I discovered it, and I have no idea how I am going to live without it.

Thank you so much, Greg, for all the hard work you have put into Findory. You may consider yourself lucky to have worked on Findory, but I consider myself lucky to have used it. Findoy has taught me what we need to demand of personalized systems, and any company that enters this space will have no choice but to follow the rules you established, or fail by not learning.

I hope you consider ways for Findory to live on, maybe by giving back part of the codebase to the community, or letting the users fund the website’s basic upkeep. If you can’t, let me be one of many to say that I consider it a priviledge to have gained from Findory in many ways.

I’m not sure how I’ll get on without Findory. Findory was how I discovered new blogs and found great stories from blogs no one had ever heard of. Findory was a news junkie’s greatest gift, a site that let you read your news, and then learned how to bring you better news as you did so. I’ve read 2,617 articles through Findory. I have three Findory RSS feeds that give me hundreds of new things to read every day. It was a part of how I conducted my day, every day, for two years.

I’m hoping that the hard work Greg did on Findory does not dissapear, just because Greg is moving on. I’m sure there are many possibilities for the website. If I were at Microsoft, and had some purchasing authority, I would buy it in an instant, build it into Live.com, and watch as Microsoft wins awards for the great work started at Findory. Otherwise, I think the personalization code, put into the hands of the open source community, could revolutionize hundreds of sites across the internet, built into smarter web applications. Hell, I’d pay a fee to keep Findory up and running, and I’m sure other users would too, so even that’s a possibility.

Whatever happens in the future, I guarantee Greg has left his mark on the web, first at Amazon, then entirely through his own great ideas at Findory. He should be proud of the work he has done.

Comments

Tag:

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

Not only do many Google services not work or have minor problems in Opera, but some of them are hard-coded to not even run in the world’s greatest browser. This Javascript will get Google Docs & Spreadsheets, as well as Picasa Web Albums and its slideshows, up and running, and is updated frequently, with fixes for Google Pages and Google Groups on the way.

If you don’t know how to use a user JavaScript in Opera, Ionut Alex explains how.

(Found on Findory)

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

1. Yahoo – 130 million unique visitors
2. Time Warner – 120 million
3. Microsoft – 118 million
4. Google – 110 million

The reason Google will soon top the list? 22% growth from the previous year. Of course, this does not appear to take into account the fact that Google now has YouTube, itself a top ten website, and that Orkut.com is showing surprisingly strong growth lately. According to Alexa, Google now has three top ten sites, besting any other competitor.

These graphs show google.com vs. yahoo, msn.com, live.com and youtube.com, and the second replaces youtube.com with orkut.com:

UPDATE: A VC has some great Comscore numbers and charts. It’s very interesting how different the US and worldwide numbers are. It looks like properties get popular in the US, then, if they last, they eventually get popular worldwide and have huge staying power after that point. As a result, while the US top ten has some recent additions, the international is topped by the big three, followed by a major gap, and then way down, some old guard (Time Warner, Amazon, eBay) and some sites that have just been around a while (Apple, Adobe, Lycos), plus Wikipedia.

Comments

Tag:

Add to Del.icio.us | Digg | Yahoo! My Web | Furl

Bookmark WebProNews:

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

Just enable the layer, which should now appear in your Layers panel, and zoom in to see all the pretty red, white and blue stars. Clicking on any of them reveals the candidates in that area, with links to web, image and news searches. I’d prefer if Google actually listed the more likely candidates first (as opposed to what appears to be a random ordering), but this is a really cool way to inform voters of the elections in their area.

Found on Findory)

Comments

Tag:

Add to Del.icio.us | Digg | Yahoo! My Web | Furl

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

Findory had famously been running a form of personalized AdSense, by feeding to AdSense keywords based on the logged-in user’s account, but Greg Linden says that even with very well-chosen keywords, AdSense just didn’t deliver the results he hoped for.

In his blog, Greg says that AdSense’s targeting is too stubborn, insisting it knows best (like delivering ads about car parts when you use the keywords “search engine“), while Findory had hoped to use its detailed personalization history to make the ads more targeted.

Since AdSense would not bend to the Findory database’s will, Greg decided to find an ad system that would, and it turned out to be one that maybe even has some legacy code Greg wrote back in his Amazon days.

According to Greg, Amazon’s system delivers slightly lower earnings than AdSense, since while books have higher payouts, but not only do you have to get the user to click the ad, they also have to buy the book.

However, Amazon listens to Findory’s data more often, which means there is more of a chance they can tweak it to get better earnings, something that proved almost impossible after a year of tweaking AdSense.

One suggestion: While getting Amazon’s personalization data to play well with Findory’s might prove impossible (or not even a consideration), Findory could keep track of which users do and don’t click ads and buy books, and target the ad system by (a) if they don’t click ads, introduce some randomness to find some ads that work and (b) if they do click ads and buy books, target the ads more towards the books the user is likely to buy than the news articles the user tends to read.

After all, I might read a lot of articles about Google, but what if the only books I buy are about Superman?

There’s a lot of room here, and I wonder if Findory can do a better job targeting ads than Google can.

All they need to do is raise the eCPM a nickel to declare success, and it would make Google realize that it needs a broader set of data than just the words on the page to create an even more effective ad system.

Comments

Tag:

Add to Del.icio.us | Digg | Yahoo! My Web | Furl

Bookmark WebProNews:

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

You can read the story at the Google Code blog, but the gist is that HP built this incredibly powerful engine and then sat on it for the last 11 years. Now, it is finally being released to the community at large, and hopefully destined to be greatly improved upon by OSS coders. Google says that it is “far more accurate than any open source OCR package out there”, so this is quite a gift to the coding community.

Google’s even looking to hire people to work on it.

(via The Scoop > Findory)

Tag:

Add to Del.icio.us | Digg | Yahoo! My Web | Furl

Bookmark WebProNews:

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

The 2.6 gigabyte file is only available for the first 100,000 downloaders, and is getting enough coverage that time is running out. Get the file now.

If anyone at Microsoft can answer: Can those who download and install Pre-RC1 upgrade to RC1 and then Vista, or will they have to clean install? Obviously, there is a big difference.
(via Winbeta > Channel 9 > Findory)

Download progress: 18.4%, 1 hour 30 minutes remaining.

Tag:

Add to Del.icio.us | Digg | Yahoo! My Web | Furl

Bookmark WebProNews:

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.

Apparently, since Google hasn’t figured out yet that indexing the message boards of Vibe Magazine is an awful idea (and completely against the purpose of Google News), spammers are hitting that board hard with “free ringtone” spam. Take a look (and, if Google does fix it, check out Jeremy’s screenshot).

There are lots of problem sources in Google News. If Google would just have one person whose job, even 20% of the time, was to do searches on highly spammable words like phentermine, they’d notice that News includes sources like livearticles.org, which has none of the high-quality news articles Google is supposedly looking for.

The next time a perfectly legitimate blogger complains that Google News didn’t accept him because of “tone”, I don’t think I’m going to take Google’s side. Either do some spam control, or block the keywords altogether.

(Found on Findory)

Add to Del.icio.us | DiggThis | Yahoo! My Web | Furl

Nathan Weinberg writes the popular InsideGoogle blog, offering the latest news and insights about Google and search engines.

Visit the InsideGoogle blog.