http://www.example.com/innerpage.html (link hardcoded), vs

/innerpage.html or ../innerpage.html (link drag ‘n dropped in DW).

DW just does it when a page gets moved or renamed, & spiders are as proficient as browsers when they recreate the absolute address at showtime. As they should, it’s a fundamental requirement.

I admit I’ve never tested the home page issue as for client sites I usually use .htaccess to explicitly set the default home page. You can also set the order they are looked for as well, but I haven’t needed to do this yet.

Consistency: When you drag ‘n drop a home page link in DW (template or page) it does create a /index.html style of link, not a simple /. And later if you add code & rename it to index.php or index.shtml, all the links get updated. If a fool like me can get it working, it’s a no-brainer for everyone else.

Off topic: Wry smirk when you talked about moving content to a CMS. Had one recent project where we were asked to move content from one (proprietary) CMS to another. Bad idea. The whole CMS thing is great in may ways, but it’s a one-way street unless import/ export is really, really good.

Michael: Not sure what you disagree with, as the search engines will treat example.com/ as a different URL from example.com/index.php (for example).

Searching and replacing is often necessary with relative URLs, but it’s much more difficult to do particularly if inconsistent naming conventions are implemented.

You can use any of the following to get to the same page from within a site that uses multiple folders:

/
/index.html
../
../index.html
../../index.html

If Dreamweaver prevents that kind of ambiguity while implementing relative URLs, then that’s better than one could hope for, I suppose.

It is also possible to crash servers (I have done this) when trying to compensate for changing file extensions on a directory index file. That is…

index.html
index.htm
index.php
index.shtm
index.shtml
index.cgi
index.asp

can all perform the same function but some servers are set up in such a way that trying to redirect mistyped URLs crashes the servers. Servers are usually set up to run down the list in some order to look for the preferred index file but only if no index file is specified in URLs being fetched.

Hence, the cleaner, more absolute, more compact your internal linking structure is, the less likely other people are to make a mistake when linking to your root URL or a directory folder.

“While you can also use absolute-path links for local links (to documents in the same site), that approach is discouraged–if you move the site to another domain, all of your local absolute-path links will break. Using relative paths for local links also provides greater flexibility if you need to move files within your site.”

Most of us are too lazy to change this because we didn’t have a reason to. And it worked when we put the site.

Now we do have a reason, although I disagree with you on the canonicalisation point. Use DW, FP or whatever to move a page, and the relative links are all updated. Same if you change a page’s name. When you use absolute links, you have to use search/ replace to do the link update. Again, not a big deal but most of us are lazy & it’s not as automatic as having the software do it for us.