SEO questions before launching new web site

I'm launching this big database (1.5+开发者_StackOverflow社区 million records) driven website and I want to know some SEO tips before..

1. Which links I need to tag as rel="nofollow", rel="me", etc?
2. How to prevent search engines to follow links that are meant to users only? Like 'login', 'post message', 'search', etc.
3. Do I need to prevent search engines from entering the 'search' section of the site? How to prevent it?
4. The site is basically a database of movies and actors. How to create a good sitemap?
5. I need to prevent search engines form reading user comments and reviews???
6. Another robots.txt or .htacces configuration is needed?
7. How to use noindex the right way?
8. Additional tips?

Thanks!

1. If you just have internal links, no reason to make them nofollow
2. Make them buttons on forms with method="post" (that's the correct way to do it anyway)
3. Don't think you need to do that.
4. Perhaps see how IMDb does it? I'd consider just listing all actors and all movies in some sort of a sensible manner or something like that.
5. Why would you need to do that?
6. Depending on whether you want to block something (via robots.txt) or need .htaccess for something else
7. No idea
8. Remember to use semantic HTML - use h1's for page titles and so on.

1. Use nofollow when you don't want your linking to a page to give it additional weight in Google's pageRank. So, for example, you'd use it on links to user homepages for comments or signatures. Use me when you are linking to your other "identities", e.g. your facebook page, your myspace account, etc.
2. robots.txt allows you to give a set of rules to webcrawlers on what they can or can't crawl and how to crawl. nofollow also tells Google not to crawl a link supposedly. Additionally, if you have application queries that are non-idempotent (cannot be safely called multiple times), then they should be POST requests—these include things like news/message/page deletions.
3. Unless your searches are incredibly database-intensive (perhaps they should be cached) then you probably don't need to worry about this.

4. Google is intelligent enough to figure out a sitemap that you've created for your user. And that's the way you ought to be thinking instead of SEO; E.g. how can I make my site more usable/accessible/user-friendly—all of which will indirectly optimize your site for search engines. But if you want to go the distance, there are semantic sitemap technologies you can use, like RDF sitemaps or XML sitemaps. Also, Google Webmasters Tools offers site map creation.

5. No, why would you want to hide content from the search engine? Probably 90% of StackOverflow's search engine referrals are from user-generated content.

6. What? Configure your web server for people, not search engines.

7. This is easy to find the answer to.

8. Don't make your site spammy, such as overloading it with banners or using popup ads; use semantic markup (H1, H2, P, etc.); use good spelling/grammar; use REST-style URLs (even if it's not a RESTful application); use slugs to hide ugly URI-encoding; observe accessibility standards and guidelines; and, most importantly, make your site useful to encourage return visits and backlinks—that is the most sure fire way of attaining good search ranking.