Microsoft just launched it’s own social networking service, Wallop. As usual, Microsoft is joining the game late, adding a few innovations, and hoping to take the industry by storm.

New Features, But at What Cost?

Wallop features a few new technologies designed to improve the social network experience. Adding friends to your list is more strictly managed to reduce the incidence of stalking, etc. Most of the system is based on Flash, making it cross-browser compatible, dynamic, and slick looking. Microsoft is also establishing and supporting a Wallop Developer/Modder Community. Modders can create and then sell different Wallop Widgets to other Wallopers.

By making their system Flash and regulating what gets uploaded, Microsoft has increased quality but really clamped down on add-on volume and creativity. For example, I build web sites for a living, but I don’t know any Flash. I’ve just been pushed out of customizing my site beyond the presets and provided functions, and that may be too limiting for my tastes.

We’ll see if Microsoft got this venture into social networking and the balance between control and creativity right.

Invitation Only

Wallop is by invitation only. Since Wallop was released today, I don’t know of anyone who is a member yet. I have submitted a request to become a Walloper, so hopefully I can get in and go check it out soon. I will let you know if it’s cool.

Web accessibility, in a nut shell, is making your web site as usable as possible to as large a group of people as possible. One of the major parts of web accessibility is text size—each person has a text size that they are more comfortable reading. Unfortunately, one of the biggest complaints about many modern web designs is that the text size is too small.

Resizable Web Site Text

Most web browsers allow you to set your default text at different sizes, and most allow you to resize the text on a web site. In order to be usable to the largest possible group, web sites should have resizable web site text. If you can’t read the text, simply make it bigger, right?

IE’s Text Size Bug

Internet Explorer 6 has a well-known limitation—it can’t resize text that is set in points or pixels. What this means is that a site needs to have font sizes coded in percentages or ems in order for Internet Explorer 6 users to be able to resize the text.

Text Resizing on the Web

I performed a little research to see how well the web is doing with resizable web site text. The results were very interesting.

I went to Alexa.com to find my sample of web sites. I visited each of the top 100 sites in the U.S. using Internet Explorer 6. I tried to resize the text on each site, and if a significant portion of the text (75% or more) changed sizes with my settings, the site was labeled as a site with resizable text. If less than 25% of the text changed, the site was labeled as a site with a fixed text size.

Here are my results:

The blue bars are for sites with resizable text, and green is for fixed text sizes. The four columns are for ranks 1-25, 26-50, 51-75, and 76-100.

Click to Enlarge

Top Web Sites

The top web sites mostly utilize resizable text—7 out of the top 10 and 16 out of the top 25 sites have resizable text.

Bottom Web Sites

30 out of the last 50 sites have fixed text sizes.

Overall Text Results

Overall, 53 out of the top 100 sites have fixed text sizes. This means two things for the web design industry—first, it’s great that so many sites have recognized (or at least not messed up) this issue. Second, we have a ways to go until the majority of the web is easily accessible to vision impaired users.

ChaCha uses humans for search, but they’re no longer the only ones. Some search problems are a long way off from being solved, and rather than wait for technology to catch up, Google is using humans to label images for its image search.

Consensus Labeling

Google uses humans by making it an image-labeling game—you and a partner label the same image at the same time. You only earn points and move on to the next image when you and your partner give the image the same label.

I must admit, it is fun to try to label images quickly and see what kind of score you can get. At times, it is frustrating because you and your partner see the same image so differently. You can pass to the next image, and hopefully that will help you to keep earning more points. Google certainly benefits from having many, many images labeled quickly.

Labeling Issues

There are a few issues with the quick consensus labeling solution. One is quality—if you are hurrying, images usually end up labeled with one word. A cover of a book about “the darkness within” got labeled “red” instead of “depression,” “psychology,” or “emotional exploration.” Forcing a consensus also tends to limit the length and complexity of labels.

Another problem is difficult images. There are images that are abstract, hard to see, or complex and that would require more time to properly analyze. These images get passed over by humans in a hurry to agree with their labeling partner.

Human Solution

Even though there are problems, you have to admit that it is certainly more effective to use humans to label pictures than machines. Humans can evaluate the entire picture at once, make associations, and use emotions. “Holiday,” “barbecue,” “graduation,” and “play” are all labels that a machine would have an exceedingly difficult time in coming up with.

Duplicate Labels

So why doesn’t Google follow the tagging phenomenon? There are tons and tons of pictures already categorized, labeled, and even described on Flickr and other photo sharing sites. And a simple game doesn’t have much stickiness—there needs to be a stronger incentive to come back and label hundreds and thousands of pictures. It will be interesting to see how Google and others will solve the image problem, along with the other semantic, meta-data problems of the web.

DMOZ is another name for the Open Directory Project, the largest, free, human-edited directory on the web. It is constructed and maintained by volunteers from all over the globe who make submissions and edits to the listings. As the largest directory, it provides listings and results as a help or starting point for many search engines, including Google, HotBot, Lycos, Netscape, and others. Go to BruceClay.com to see a graph of how central DMOZ is to search engine results.

Google Draws Results From DMOZ

Google, like other search engines, often uses the listings and descriptions from the ODP in its results. These results and descriptions can have a powerful effect on how effective your site’s search engine results are. Wouldn’t it be nice to tell DMOZ and Google what to say about your site? To write a description that accurately describes your site and entices visitors to come?

How to Submit a Site to DMOZ:

Luckily, you can tell Google and DMOZ how to portray your site. DMOZ is edited by humans, and so to get a listing in the ODP, you need to make the humans there aware of your site (which, if your site is new or small, may not be very likely). There is a link on most DMOZ category pages for suggesting sites and descriptions for those sites. Official guidelines for the DMOZ URL suggestion tool are here.

1. Go to the category that you feel fits your site best
2. If available, click the “suggest url” link
3. Enter the web address of your site
4. Give your “elevator pitch,” the concise, accurate summary of your site that you want DMOZ to display (Make sure your description is objective and doesn’t sound promotional)

Chances are, if you make it easy for them, the editors will simply use the information you gave them and give your site a good DMOZ listing.

What to Do About a Bad DMOZ Listing:

Sometimes the editors don’t understand your site and give you a poor or incorrect listing. Sometimes your site changes and your listing needs updating. You can do one of two things about these problems:

1. Update your listing. You can update your listing by going to the category where your site is listed and clicking the “update listing” link. Make sure to clearly show the error, short coming, or change that needs to be made in your DMOZ listing.
2. Use the NOODP meta tag on your site. The NOODP meta tag allows you to ask search engines to override or ignore the ODP listing that your site has in favor of content found on the pages of your site. Matt Cutts’ blog entry shows you how to use the NOODP meta tag.

DMOZ provides what is, for the most part, a valuable service. There are allegations that the directory is flawed, spammed, or worthless, but the editors are only human. Besides, flawed, spammy, or worthless, the ODP still has a strong influence on the web and you have the tools to influence what they say about you or to avoid ill effects from potential DMOZ problems. Use DMOZ submissions as one more edge in your effort to improve your rankings and marketing message on the Internet.

I’ve spoken about the virtues of clean, semantic, valid markup before, and lots of designers recommend that you validate your code. What doesn’t seem to get addressed, however, is how one goes about validating the HTML of a web page.

A Major Hassle

It used to be that you would have to validate pages one by one using the W3C’s free validator tool. This method works, but it sure can be time consuming, especially if you have a large site or a site that is mostly valid already. Getting through a lot of pages or finding the few that have problems is very difficult this way.

Validation on the Fly

Luckily, Marc Gueury decided to do something about this important but difficult process. He wrote an HTML Validator Extension for Firefox. It’s based on HTML Tidy—basically, you install a validator in your browser.

The validator extension for Firefox is great because all you have to do is browse from page to page to see which are valid and which are not—an icon in the corner of your browser tells you whether your page is valid, contains warnings, or contains one of two kinds of errors (errors that the extension knows how to fix and errors that it doesn’t). To see the warnings and errors, you simply need to view your page source. This all takes place on your machine, so there is no external server to fiddle with or wait for.

Accessibility Evaluation, Too

The HTML Validator extension also has accessibility evaluation built in—if you want to see how accessible your site is, you can turn on evaluation for one of 4 different levels. This part of the tool is helpful for a quick look at how accessible your site is.

Not Quite Perfect, But Close Enough

The validator extension isn’t quite perfect—occasionally it will make mistakes (especially with heavily nested elements) and it seems a little bit “touchy” at times, but for the most part it is a powerful and convenient way to validate your website. I used it to make my site valid, and I wasn’t even sure I was going to validate it at all before I found the HTML Validator extension.

Provo Labs is a Utah firm (run by Paul Allen) that incubates web-based companies, providing funding and guidance to young firms. It has experienced some recent difficulties, and some people have asked, has Provo Labs failed?

No.

Paul Allen wrote a post about how Provo Labs has made some changes to return to its original vision and trim back on some over-extensions it has made in the last few months. He was very open about admitting some mistakes he’s made and about how what Provo Labs can achieve in the immediate future is more limited than the original scope of his vision. However, he didn’t use the words fail or failure in his post. Provo Labs still exists, and can still do amazing things.

Lindsey Snow posted a comment on Paul’s blog post that stated “I use the word failed because Provo Labs has failed many people.” In one sense, she’s right, Provo Labs didn’t live up to many people’s expectations, least of all founder Paul Allen’s. However, I would disagree in other senses. Provo Labs is a startup incubator. To work for a startup is to take a risk—many startup businesses fail or don’t live up to expectations. Estimates range from 40 to 80% of new businesses failing in the first 1 to 5 years! Your chances are 50/50.

Is a Startup That Folds a Failure?

Sure, people depend on the company they work for to support themselves. But is the startup that folds a failure? How did the employees of the startup support their families for several months or years? Did they learn anything? The skills that got employees hired at a startup and the skills they acquired while there will help them to do better in a new firm. Besides, jobs in stable, established companies often last just a few years anyway. This seems to me like a case of unrealistic or overoptimistic expectations, not a case of abject failure.

Is it the Incubator Firm’s Fault?

If the firms that receive funding from an incubator fail, is it the incubator firm’s fault? In this particular case, is it Paul’s fault? Paul apologized for being spread too thin, but who can be the CEO of 12 firms? The small teams in each company receiving funding from the incubator firm are ultimately responsible for the success or failure of their venture. It is dangerous to rely on the incubator firm too much for guidance and focus—especially as the sole source of inspiration or ideas. If you don’t believe in the idea, or if you don’t feel as passionately about it as the originator, or if you can’t see how to guide a company based on that idea, then you are in trouble even if the originator of the idea can hold your hand every step of the way.

In Paul’s case, he is reducing the number of firms he is personally working on because the firms in the incubator were unable to function without his attention, and no one can be expected to spread his or her attention so far. Paul is just realizing that, as the head of an incubator firm, he can’t personally attend to every firm. As the head of an incubator company, he is more of a portfolio manager than a business manager. Paul’s meeting to set things straight was meant to realign employee expectations to the way they should have been from the beginning—you will sink or swim based on you, whether we assist you or not.

All that being said, don’t take me or Paul as unsympathetic towards the employees and families who worked hard because they were promised much and were then let go. It’s a hard place to be in. I just don’t think that they can justifiably feel offended or unfairly treated. What startup begins without high expectations?

The primary challenge for search engines is to provide search results that are relevant and accurate—to find and deliver what it is that people are looking for. ChaCha Real Live Search is taking a new approach to finding and delivering the results that people want from their searches.

Machines Aren’t Human

Instead of fighting the fact that machines aren’t human, ChaCha is embracing it—the whole idea behind ChaCha is to acknowledge that machines will never be able to return the results that humans are looking for as well as a human could. Instead, ChaCha is using humans to deliver results. ChaCha “guides” are paid to deliver one or more relevant results for a query, using whatever abilities and skills humans have that machines don’t (intuition, inspiration, and the ability to see subtle, implicit relations between items).

My take on this was, “oh, they’re combining Google with DMOZ”—building a search tool with the human recommendations built in. It will be very interesting to see how well ChaCha can store and track the results and information that are generated by their guides. The success of ChaCha is dependent on how well they “leverage their human resources”—on how many results and how much relevance they can get from a limited group of people.

Humans Aren’t Machines

Another severe potential problem with ChaCha is that humans aren’t machines—they have foibles, imperfections, and biases. For example, how can ChaCha guarantee that some of its guides aren’t going to try to deliver results that benefit them or their friends? It needs to establish a guide’s level of expertise to make sure he or she is delivering good results. It may be difficult to have enough quality guides to handle the potential volume of searches.

Paying for the Search Results

Another critical factor for ChaCha’s success is how it is going to pay for its guides—ChaCha serves video ads while you wait for your human guide to gather results, and it is dependent on a critical mass of these ads in order to survive. If they can’t garner sufficient advertiser interest, they will choke as the volume of searches increases.

It would be really, really neat if the video ads that were served were related to the query—and the ad spots were auctioned off just like Google or Overture text ads.

Google Support

One of ChaCha’s main pillars is its live chat feature—you communicate with your guide through chat. This means that essentially ChaCha is Google Search’s support team. Google doesn’t provide much customer service, and so ChaCha is filling the gap for them. There is a pretty close resemblance between Google’s and ChaCha’s home pages.

Watching ChaCha will certainly be interesting—who will deliver more relevant results, the scaleable-but-uninspired machines or the intuitive, emotional but non-replicable humans? How will ChaCha manage its mix of humans and machines? Can ChaCha out-do Google’s results for relevancy?

Note: please don’t violate ChaCha’s terms of service by posting invitations for people to join ChaCha on this blog. I don’t want any guides losing their accounts because ChaCha found out they weren’t following the rules.