Archive for the 'Internet' Category

Recently, a bunch of former Google Answers Researchers launched Uclue, an affordable paid Q&A/Research service. It’s somewhat similar to what Google Answers was - you ask your question, set a price, and wait to see whether a researcher can supply your answer (you get a refund if they can’t).

We couldn’t have launched successfully without the enormous help received from the Uclue researchers, who worked hard on design, testing and marketing.

Even more striking, though, was the help we received from those not participating as Uclue researchers. Dozens took the time, unprompted, to submit detailed and helpful feedback about the operation of the site. We received many great suggestions, and are still only part-way through implementing them.

And it didn’t stop there. A software product manager provided some unsolicited but useful insight into aspects of our business model.

We got help with the visual aspects too. The base software package that we are customising, AskPert, uses a rather basic table-based layout:

A less-than-charitable Digg commenter took one look at this and asked whether we had employed “the CraigsList design team”! Another Digger with username soiq was more constructive, and designed us a new layout:

All he wanted, he said jokingly, was a year of Uclue membership in exchange for the PSD files.

We received another unprompted redesign too, this time from Philipp Lenssen (publisher of Google Blogoscoped):

And that’s the design we decided to use.

So, as the site and the service come together, it really is due not only to our work, but also to the work of others. Thanks everyone!

Here’s what happens when a low-traffic blog like this one is linked to from slashdot.org.

And that’s just from a link in a comment that I posted to Slashdot. A link from an article would surely melt my server…

Stock photos are images that have been taken in anticipation that they will be useful in the future. When you need a photo of a juicy hamburger or a pink cat, you will find one on a stock photo site.

image by Gerla Brakkee

Stock photos used to be produced by professional photographers and licensed for tens or hundreds of dollars per use, but now there is a huge range of fantastic images on the web available to be used freely. Sometimes there are “strings attached” such as a restriction to noncommercial use, or the requirement that the photographer is credited.

When I need a stock photo for a web page I’m putting together, my first port of call is always Stock Xchng, where there are over 200,000 photos. Registration is free and has been without hassles or spam, but you do need to register in order to download full-size images.

Just type in your search terms and browse the resulting photos. The photographer chooses the terms and conditions. All images are free of cost, but some require attribution or written permission for use. I always use images that are completely unrestricted, but I always credit the photographer anyway, and send them the URL. That way, if some future maintainer of the web page removes the attribution they are not in breach of the photographer’s conditions.

Many stock photos are produced for commercial purposes, and can be rather bland (see “Ten Top Stock Photography Cliches“), but there are lots of fresh ideas at Stock Xchng.

And I even contributed a few of my own photos to the site.

Google Blogoscoped posted a challenge to find a query that returns the same top result when the query is reversed. Philipp Lenssen, the blog’s author, calls that a Googledrome.

Dave Pettit found that searches for Oprah and Harpo both led to oprah.com, and I managed to find a couple more Googledromes.

I had a hunch that it might be possible to construct (rather than “find”) a Googledrome for the home page of Google Blogoscoped. Naturally, a search for “google blogoscoped” returns Philipp’s blog as the top result. I searched for “depocsogolb elgoog” and found that Philipp’s blog was the last of four results.

That’s not surprising, because the blog’s home page doesn’t include the reversed words - they are found only in the text of a link leading to the home page, and the page with the link is going to rank higher than the linked-to-page. I had to find a way to exclude the linking page from the search results.

I inspected the Google cache for the home page (because that will match Google’s indexing). For each word on the page, I searched for its reverse. About half way through the page I struck gold: the words “live” and “evil” both appeared on the page. Searching for depocsogolb elgoog live brought the Google Blogoscoped home page to the top of the list.

But now I had a problem with the other version - searching for evil google blogoscoped returned individual blog posts which used the word “evil” more times and more prominently than the home page. I had to find a way to eliminate those pages from the results.

I could have continued to seek words whose reverse also appeared on the home page, but an easy option was available. Single-letter words are their own reverse, and by adding ‘e’, ‘g’ and ‘3′ to the search I was able to eliminate the individual posts from the results and bring the home page back to the top of the listing.

I ended up with a Googledrome consisting of the following two searches:

• g e evil google blogoscoped 3
• 3 depocsogolb elgoog live e g

This Googledrome won’t last long, of course. The home page of Google Blogoscoped is in a state of constant change, and when it is re-indexed the query results may change too.

Also, Philipp has hinted that he might change the rules to disallow single-letter search terms.