Welcome to the Q&A Forum

It's really going to depend on a lot of things. A good thing to do would be to draw up a list of what you want the platform to accomplish and then start triaging from there. In my mind these are the important questions:

• Open Source or For Pay? I'm not especially impressed with a lot of the OSS solutions out there (e.g. osCommerce) but the advantage they do have is that the developer base is enormous. This in part depends on the IT resources within your organization. If you don't have dedicated developers it'd probably be better to go with an open source app (e.g. osCommerce) and run with hired guns. You will take a hit in terms of flexibility, configurability, and so on, and likely in the long run end up paying more for add ons, extra development, maintenance, etc. than you may have by in-housing.

• Scope of features: What do you want it to do? If it's just connect to payment gateways and receive money then there's several free/low cost solutions out there. But what about order tracking? Auto-emailing on confirmation? Harvesting and sending tracking numbers? Inventory integration? Logistics? Accounting tie-in so you can determine margin and ROI? You need to figure out what you want this application to do. The tighter the scope, the easier it is. Related to this is, how many SKUs? How many categories? Do you need Search?

• SEOability: Several of these platforms offer some kind of SEO integration, with the most basic being control over meta tags/on page elements to reduce duplicate content penalties.Others go whole hog and offer what are essentially complete CMS packages that allow granular control over every page. You need to determine A)what your SEO strategy is and B)your resourcing to support it. There's not much point splashing out cash on an advanced total CMS integrated system if you only have one overworked intern who will be creating content for it.

This wikipedia page is pretty good for breaking down the major features and what's offered:

http://en.wikipedia.org/wiki/Comparison_of_shopping_cart_software

Brief impressions on what I've worked on in the past:

Magento: Very good feature wise, but not newbie-friendly (e.g. you'll need a dedicated developer/contractor). Pricey. Scales up for large very well, large developer pool.

AbleCommerce: Terrible. .NET based. We call it 'baby's first shopping cart' and it shows. Every client I've had that started on AbleCommerce outgrew it in one quarter.

osCommerce: Mediocre. Tons of developers and plug-ins but the can end up being completely unmanageable if not grown properly. PHP based and all the security problems that come with that, and open source code. Last year there was an exploit that was used to attack literally millions of sites because people weren't updating their installations.

Speaking personally I prefer hand-rolled solutions, but we've made so many over the years it's almost second nature.

If you follow their advice it's actually pretty easy; what they suggest in the article is just disabling the 'custom' error functionality for those php functions so it doesn't present 404 any more. This means copying and pasting those 5-6 lines into one file inside your root web directory. They end the article with another code snipped that will actually log those errors into a log file on your system instead of letting them go 'invisibly' (because once you disable the custom error how do you know if it's happening any more?).

SEO wise it's hard to say. Obviously in an absolute sense less errors = more better, but the 404s should be considered in context of your overall site's performance. What you might want to examine is why did your error rate apparently triple around February?

From a pure 'does Google notice and rank this' perspective auto cross-posting doesn't really have a measurable effect that I've seen.

However, it entirely depends on the audience you are publishing to and their level of engagement. If you have people actually reading those social platforms and sharing your content in turn, then it's obviously a slamdunk that you should be doing it. For some clients I've had a tweet live for weeks, getting forwarded around, showing up on blogs, etc.

There's two questions to ask really:

1. Do you have an engaged audience that is going to use this content?

2. How do you tell if it's even being shared/read?

You'll note that I haven't addressed how measurable Google's response to this kind of thing because in my opinion... it isn't really measurable as of yet.

Isn't this more of a DNS question? Shouldn't it just go:

site.tempurl.com is an alias for newsite.realurl.com?

And then you change the base domain/config in WordPress so it never references tempurl.com ever again?

The error basically means that your PHP page called a function (opendir) and it errored out. This is built into php in case you wanted to specifically trap for that event  (e.g. to write your own custom error page in this instance). This blog post does a pretty good job of summing it up:

http://perishablepress.com/eliminate-404-errors-for-php-functions/

There are many, many, great frontpage SEOMoz articles that explain a lot of this stuff. In the abstract I would say that 'strategic' SEO (e.g. overall site structure and planning) does require a grounding in technical issues but that 'tactical' SEO (e.g. per keyword or category content creation) doesn't necessarily, as long as there is a clear understanding of what is to be accomplished and how.

Here's a good article from the Learn SEO series on this site:

http://www.seomoz.org/learn-seo/canonicalization

(The whole Learn SEO section is pretty good)

The short answer is, canonical tags tell search engines what you want the 'real' URL for any set page to be when presenting it as a search result. Consider these URLs, which all theoretically go to the same place:

mysite.com

www.mysite.com

www.mysite.com/

mysite.com/default.aspx

www.mysite.com/index.html

If search engines treated all of these as 'different' pages, it means that your 'link weight'' or relevance for them would be split 4 times, depending on how many other people linked to them, how you link to them internally, how your sitemap references them, etc. But you don't want that; you want 'one' default page for your site. The canonical tag lets you explicitly specify which of the above should be considered the 'canonical', or original, page.

This has a variety of applications, but these are the most important:

• Controlling capitalization. Is there a difference between Default.aspx and default.aspx? It's unlikely, but canonicalization can take care of this

• Reducing search engine confusion for identical pages, as above.

• Reducing querystring over-indexing of duplicate content. If you can have over 100 iterations of SearchResults.aspx?Start=0&End=100&Sort=Asc, SearchResults.aspx?Start=10&End=110&Sort=Desc, etc. etc. you can canonical them to the original, plain, SearchResults.aspx to help avoid duplicate penalties.

I'm sure there's more, but I'm just writing this off the top of my head.

I think you have to consider the scale of the enterprise; it's easy to think of Google as being some kind of video game that responds to your input as soon as you make it (note to self: make Video Game about SEO. Possible titles: Link Awakening, The Adventure of Link, A Link to the Past). Also consider the number of links you're talking about; I'm surprised you saw any movement for just five links.

To me, Google's response (including time it takes them to notice) to a link is kind of ineffable. Charles' advice is pretty good. If you're solely concentrating on number of links and what affect they'll have you might be doing the wrong kind of SEO.

I've noticed these guys in link profiles for several sites I manage. They'll usually show up around 1,000-10,000 times in the backlink profile. From what I can tell they index websites, build up keyword relationships, and then when you search for something on their site (e.g. poker) they'll present a list of related sites with stats about them. The stats seem to be yanked straight from Alexa. Where the backlink comes from is that every time 'your' site shows up for a search result they'll put a little iframe that contains your site. This means if your site's name/keywords are pretty broad, you could be showing up thousands and tens of thousands of times as being linked from these guys on their pages that Google indexes.

And Google indexes, boy do they ever. At the height, they had over 53 million pages indexed. That has apparently shrunk now to around 25 million. I believe their strategy is to generate a crap-load of automated content in the hopes they can cash in on obscure long tails.

So my questions for you guys are:

1. Are you seeing them in your backlinks too?
2. Should I block their spider/referrers?
3. What is their deal man?

I'm not saying forum sigs are in and of themselves bad, I'm referring to blitzing every forum under the sun related or not with them by third party link builders. If it's relevant and you're building it up in-house, sure. I've even done that in the past.

While I realize that it's probably possible for third party back linkers to be useful I think both the pricing model and market are against it. When payment is 'per link' there's no real incentive for the provider to do anything other than quantity over quality. I know I've seen link profiles for both competition and projects I've taken over from other SEOs that are plainly silly, and obviously reek of badly done link building. This includes forum profile pages, signatures, Yahoo Answer spam, blog spams, markov-chain generated splogs, sponsored blogs barely in english, thousands of pages based on scraping, and so on, and so on. It's possible that there are good ones out there (after all, if they were good I wouldn't even notice their activity) but I've yet to see any.

You're thinking about it the right way at least. Many clients come to me with "get rid of that bad listing right now!!!" and I have to respond with "I'm sorry but if I controlled the internet I wouldn't be working for you" and then they storm out and it's generally a bad scene.

All those things you mentioned are good tactics to push something down the SERP. Some others off the top of my head:

• Press Releases
• Guest Blog/Friend Blogs about the company
• Bizarrely, YouTube works sometimes if the account is co-branded
• Registering for services like Manta, TechVibes, citdirectory, macrae's, other industry directories
• Legitimate forum postings made in the company's name

There's others, but I think you get the idea.

There's quite a few sectors that have a bad rep SEO wise. Real estate is one of the ones where it's because the competition is very tight. However I think this is primarily the reason, the competition. Other sectors that receive attention (adult dating, poker, pornography, forex, mlm affiliation) are usually because they're competitive and packed to the gills with black hats. So, that said I wouldn't consider Real Estate a 'bad neighborhood' like the aforementioned. At the same time though, I wouldn't bother grooming links with them unless I was specializing in Real Estate SEO.

As these people have said it depends on the content, though I use a rule of thumb from the journalism days:

If you're reporting on something, that is something happened (e.g. new iPhone launch, founder of InvisibleChildren caught masturbating in public) it's usually only 250-500 words.

If you're explaining something, that is teaching a new concept from scratch, I'm usually comfortable with around 1,000 words.

If you're commenting on something or editorializing it can be anything in between depending.

This obviously has exceptions, I find that technology matters have shallow reports for example whereas social science/phenomena are generally longer.

 <code>```
<code>Your changed one is missing the wildcards. Unless I'm wrong, it would only</code>

<code>match 'shewula.nl' exactly, and not anything else.</code> 

<code>This might work, but I'm just eyeballing it and am not good with regex:</code>
```</code> 

<code>RewriteCond %{HTTP_HOST} !^(.).shewula.nl(.)$ [NC]
RewriteRule ^(.*)$ http://www.%{HTTP_HOST}/$1 [R=301,L]</code>

<code>edit-ugh stupid editor auto-inserting code blocks</code>
```

It could be a matter of your home page 'leaking' authority/juice to those internal pages, depending on your internal link structure. We had a client once whose blog would show up stronger than their home page, landing pages, etc. We determined it was because every single page in their domain was linking to the blog, and that was the only page that was being linked that much. We did a bunch of nofollow (and canonical to reduce redundant pages) and the situation normalized.

My only guess is that it's not in the OpenSite Explorer index. Keep in mind that the internet is a pretty big place, and even Google, with far more resources than SEOMoz, doesn't see 'every page'. I am totally guessing here (because this probably gets proprietary into how Roger Mozbot actually works) but I think OpenSite Explorer's index is a combination of the 'top' web sites by traffic, with the spider opportunistically picking up URLs from us Pro Mozzers after we submit for crawls. I have a client in a very niche industry, and probably less than 20% of their backlinks show up in OSE. Part of that is probably bad links, but the other is that their index can't contain the entire internet.

Note, this is all theorizing, I haven't seen or heard anything confirming this.

One last thing, to get an idea of the engineering problems involved, this site is always neat to check out:

http://www.worldwidewebsize.com/

I think it goes that after the initial expiry date there's a two week 'grace' period where the original owner can renew, and then another period. This article explains it all, though it's from 2005:

http://www.mikeindustries.com/blog/archive/2005/03/how-to-snatch-an-expiring-domain

I've never actually sniped an expired domain but to be frank I haven't tried that often either. The impression I get is if a domain is actually valuable there's an army of domain squatters/speculators out there who can snag them way better than I.