Every now and then, we're asked by a client if they should have a separate version of their website for mobile devices. These are generally called 'mobi' sites, because they often have a '.mobi' top-level domain name instead of the common '.com' or '.org'.

The World Wide Web Consortium (W3C), the de-facto standards body of the web, have published a set of best-practices for the mobile web. The idea is to make websites accessable to mobile devices which don't have the capabilities of desktop or laptop browsers. The standard 'feature phone' screen was 240 pixels wide, and 320 pixels high, and the browsers in them weren't very good. In fact, the standard says to plan for only 120 pixels. The standard says you shouldn't do pop-up windows or pdfs, navigation should be simplified and at the top of the screen, pages should be less than 20k in size, only count on being able to display 256 colors, don't use image maps or javascript. The main problem is that this standard was set in 2008, a year after the iPhone was released. The current iPhone's screen is 960x640, has a fully-compliant browser, runs Javascript perfectly well, and can display full 24-bit color. The Android and Blackberry smartphones aren't too far behind. In other words, the current generation of hardware has left the standard in the dust. The 'feature phones' around which the mobi standards were created are rapidly dying out.

The upcoming standards of HTML5 and CSS3 will provide better support for mobile devices with their Media Query properties. Unfortunately, no browser fully supports HTML5 or CSS3 right now. It'll take at least another year to eighteen months for the major browsers to support these new standards. And by then, the need for them will be greatly diminished by having mobile devices with better screen resolutions.

So where does that leave us from a design standpoint? None of the current generation of smartphones can handle flash (not just the iPhone as seems to be common wisdom). So as long as your site isn't flash-heavy, it should display fine on the current generation of smart phones. There may be a small group of users that don't have an adequate phone to view your website, but they will rapidly diminish over the next 18 months.

Choosing a good password is always a problem. The essential tension is that a good password should be easy to remember, yet hard to guess. Most people err on one side or the other. Most people choose passwords that are too easy to guess, such as the name of their spouse or pet. Any word in the dictionary makes for a quick security breach. We have other vendors who send us passwords like 8#{aUv7. Yes, it's hard to guess, but also impossible to remember, guaranteeing that it'll have to be written down, copied, pasted, stored, and generally left laying about. Here's some hints of forming usable passwords:

• Combine two or more words, like 'noneshallpass'
• Some people convert their passwords into 'leet speak' by substituting zeros for ohs (o->0) and ones for ells (l->1). These are fairly easy to guess as well. The next step up would be to substitute threes for ees (e->3)
• Even better, if you speak a few words of a foreign language, use them. Such as: 'verb0ten'

NIST standards call for users to change their passwords every 90 days. This is commonly accepted wisdom in the security community, but I remain dubious. I think one well-chosen, well-guarded password is better than a string of mediocre passwords that change so often the user has to write them down to remember which one they're using at the moment.

In closing, have fun with the idea.

We've seen the first two waves of Content Management Systems (CMSs) as being the custom CMS that gave rise to the proprietary CMS. The custom CMS market is dead. The proprietary CMS is under a great deal of pressure from the third wave of CMSs....