I was discussing this with my barber the other day (just one guy in a single rented shop). In the end we agreed that he should just use Facebook and Google Places but he should also buy a domain to use for email. He can reach out to existing customers via Facebook and get his shop found via local searches.

He didn't have the time or money to spend building and ranking a website but owning a domain would allow for consistent branding and mean he could salvage something if/when FB jumps the shark.

I'm going to speak up for XML sitemaps - they're incredibly useful. I agree that they don't do much for the big search engines but having an exhaustive list of all the pages on your site and their modification dates in a machine readable format is a huge boon. If I need to process every page (like running a spellcheck, checking for valid HTML or counting internal links) I just write a quick script that works through the sitemap.

Google don't need a lawsuit, they've created a mechanism to poison Bing's SERPs.

Maybe they've been doing just that since Dec 31st.

I'll happily chase nofollow links - they pass a little PR and plenty of traffic.

I'm pretty sure that pass quite a lot of domain trust too (see Todd's comment about Wikipedia). What I haven't figured out yet is whether Google differentiates between nofollow links in the post text vs nofollow links in comments.

As I understand it, Google is defining "page speed" as the time it takes to download all the components of the page (HTML, javascript, CSS, images, Flash etc) not the rendering time and not the time to execute javascript that runs on page load.

There is a big difference between Google's interpretation of page speed and that of a user sat in front of a browser.

After watching P1R & Alan slug it out on Twitter, I had to weigh in here with my 2c...

It's my understanding that crawlers run a process like HTML Tidy on the pages they download in order to better parse it. Yahoo states that it uses HTML Tidy for YQL so it's likely to be used elsewhere too.

HTML Tidy is good but it can't deal with complete garbage. If your code is that messy, crawlers aren't likely to be able to parse your pages and you just won't make it into the index.

Most of the time, invalid markup can (and will) be cleaned but the issue here is what happens during the cleaning process:

• What happens when your affiliate link anchors have two "rel" attributes?
• No alt tags? Good luck with image search!
• Your logo wasn't closed properly? Better hope that G doesn't mistake your content for boilerplate.
• Two id attributes on your headings? I doubt you're going to see "Jump To" in your snippets.

I don't think that there's a check in Google's algorithm that penalises invalid code, but that doesn't mean that validation is not useful for SEO.

As I understand the legislation, it does not allow search engines to republish your content. There's a big difference in UK law between making a copy and making copies available.

Publishers still retain copyright, but they (implicitly) allow search engines to make a copy.

I don't think that this legislation covers the Google cache, as we know that it's possible to opt out of the cache but remain in the index. So Google cache is not necessary "for the purpose only of providing said search engine services". The cache essentially republishes copyrighted works and is probably illegal in the UK.

Pagination doesn't seem to be working today - both the upcoming and hot topics pages are affected.

• Some of my comments are reported as 1 day old - they're not, I was sick yesterday
• I'm sure that I've had some of my submissions go hot but they're not reported, nor are my desphinns. I wonder if the submissions are being confused with "went hot", my profile says I have 8 submissions and I'm sure I have more.
• The "profile recently viewed by" data is currently blank.
• Searching throws a PHP error if at least one result is found:

• The graphical buttons on the feeds page are 404
• Sphinn Live is very cool - can we have a feed please?  :-)