Essence of a Domino hosted website

If you are a Notes developer, you are used to Views, providing a flexible interface to underlying data. But when you are developing websites, this might very well turn against you. Find out the importance of permalinks and why Google does not like Notes/Domino websites.

The importance of the permalink

Permanent links or 'permalinks' are a core concept in online communication. Search engines and users remember the location of the information you provided: search engines store it, users bookmark it. So finding the same information with different URLs is a very confusing situation for both. And what's worse: suppose you update your Domino application or your Notes administrator decides to re-organise the directory structure of the web server? Suddenly, all information shifts location. Search engines get broken links. Users bookmarks get 404 errors. I found out that the best thing you could do, is separate the content from the navigation, and provide just one single URL for the content. A URL that stays the same forever: the Permalink.

Site substitution rules

To establish permalinks, Notes administrators and developers have to work together. The administrator provides the substitution rule, the developer makes sure all hyperlinks created by the application (website) use this substitution. Here are the substitution rules for this site:

• /domino/ points to the database itself: /hosting/emd/blogs/mme/mme-blog.nsf. This ensures that all content in this database can always be located from one single URL, even if the database itselfs changes name or ends up in another directory. And also: it reduces the number of subdirectories in the URL, which is essential for search engines: they prefer shallow structures over deep ones.
• /robots.txt points to a page: /hosting/emd/blogs/mme/mme-blog.nsf/robots.txt. Search engines expect to find this robots.txt file in the root of the website. With this file, you can instruct search bots which content they may spider and which content is 'off limit'.
• /sitemap.xml points to a page: /hosting/emd/blogs/mme/mme-blog.nsf/sitemap.xml. In Google, you can submit a sitemap in XML format indicating your most important pages. However: Google expect it in the root directory of the website. The format can be a RSS feed, so you can re-use your RSS feed for the sitemap. I render just as the RSS itself, but with content-type 'text/xml'.

Indexes and Leaf pages

The trick for having one unique URL for each content block (read: document) and still keep the flexibility of the Notes views for indexing is simple: make sure that all the links in the views point to the document in one dedicated view instead of using the normal view link. So in this blog, all viewlinks point to the document in the /archive/ view. In fact, you could make even nicer URLs when you use two separate substitution rules, one for the indexes and one for the data:

• /index/ points to a view which contain all indexes. These are documents with an embedded view, showing a single category, pages, etc.
• /archive/would then be the single access point for the 'Leaf' pages, the documents with the content itself.

Where is the sample database?

I will provide sample databases which illustrate the principles and which you can use for your own development as soon as I have my registration functionality ready. I am willing to give, but not to Anonymous users :-)