Another example of “search becoming a modern version of the command line interface”:http://www.geekfun.com/archives/000484.html, this time on mobile phones.

ChristianLindholm.com: Qix from Zi Corporation could revolutionize search on mobile phones

bq. The SW will index the content on the phone and enable search from the idle so that the number keypad enters text mode on the second keypress and starts suggesting suitable words from the menu, phonebook or bookmarks which then can be accessed with the Select-key

No sooner do “I rediscover Textile”:http://www.geekfun.com/archives/000488.html for long enough to install it, than I come across “Markdown”:http://daringfireball.net/projects/markdown/ which seems to have similar aims:

bq. Markdown is a text-to-HTML conversion tool for web writers. Markdown allows you to write using an easy-to-read, easy-to-write plain text format, then convert it to structurally valid XHTML (or HTML).

I’ve not really had a chance to dig into it. From a brief-glance, it seems that the text pre-HTML transformation is a bit more natural-looking, particularly for headings, which isn’t suprising, since the creator, John Gruber, says that the formatting conventions used in plain-text e-mail is a big inspiration for the syntax. On the other hand, the link syntax for Textile reads more naturally to me.

I’m not sure whether I’m going to bother with Markdown given that 1) Textile’s been more widely adopted 2)I’m planning on setting up a wiki, and both formats are similar to but different from many of the native Wiki markups — learning 3 and keeping them straight is going to be too much for my little smooth brain to handle.

Some “discussion of Markdown”:http://www.43folders.com/2005/02/in_praise_of_ma.html over on “43Folders”:http://www.43folders.com.

After posting about my frustration in “trying to use the GreaseMonkey Textile Script”:http://www.geekfun.com/archives/000487.html to ease the formatting of entires into my “MovableType Blog”:http://www.geekfun.com (aka what you are reading right now), I kept thinking about it, and the more I thought, the more I fumed. “Textile”:http://www.textism.com/tools/textile/ bills itself as a “A Humane Web Text Generator,” but my experience was that it was totally unforgiving of human failing — if you made a mistake and didn’t realize it until after you’d had textile convert your text, the experience of fixing the text and re-using Textile was painful.

I was stalwart, perhaps even giddy, in my determination to channel my righteous indignation into a scathing blog entry. Then I realized that the problem was a simple implemenation issue. The client-side implementation may present problems, but as long as its implemented on the server-side, and does the HTML generation as part of the page rendering proces, everything is cool, as long as the user gets to work on the unconverted Textile tagged text whenever they make edits.

So, I went looking for something more suitable and found Brad Choate’s “Textile plugin for MovableType”:http://bradchoate.com/mt-plugins/textile. The install was easy. I just had to download the archive file, upload two files to my installation, tweak the permissions, and regenerate my posting bookmarklet to show the text formatting drop down menu.

I’m assuming that everything else is going to just work. I guess I’ll see once I hit the Post button.

_Update:_ It works, once I got rid of spurious spaces and other cruft. Bravo to “the creator of Textile”:http://www.textism.com/about/ and “the creator of the MTTextile”:http://bradchoate.com/brad/.

That will be all.

GreaseMonkey is a Firefox extension that lets people write useful JavaScripts that execute against selected web pages. I’m tried using it with a script that will process any text box using Textile to turn a simplified markup into proper HTML.

I tried using it just now to enhance the very basic edit control presented in Movabletype 2.66 when using FireFox. It made it very easy to create the links I used above with a minimum of typing. I also tried to use it to make numbered and bulleted lists, but I had trouble getting the syntax right, and ended up having to repeatedly rerun it. As a result, I accumulated all sorts of difficult to read crap in the HTML, including multiple duplicate paragraph tags, and break tags where I didn’t really want break tags.

I thought it might be a buggy implementation of Textile, but I just visited the cannonical implemenation and found that it is also rather unforgiving about rework.

Adaptive Path is a User Experience consultancy. They maintain what is essentially a group blog featuring occasional long-form essays from their principals, as well as links to other items of interest. They even use Movable Type to maintain it, but near as I can tell, there is no way to receive updates via RSS or Atom.

Dumb.

Jeremy Zawodny notes that most of Google’s applications present two UI modes.

“1. Type into the box and hit the button. Look at results.
2. Use other navigation to browse. Repeat.”

With search and others use mode 1, and others use Mode 2

He closes by asking;

I’m not convinced that the “one box” view of the world is going to be the primary mode of interaction over the next few years. Are you?

I’m not convinced either, but I do think that the ‘one box” interface is pretty powerful and 3 years ago I observed that Google seemed well on their way to becoming the command line for the internet age, which sounds a bit like I’m damming them with faint praise, but the truth is quite the opposite. Command lines offer you an entrypoint into an almost infinite world of information in a tiny space.

Google is evolving into the command line for the information age, and I think it is great.

I just stumbled across its ability to give a phone directory listing for people when you include a city along with their name in the search box. Once it works properly for businesses, this will soothe a long-time frustration with web phone directories.

It always seemed to take too much clicking and typing to get a useful answer, but I never looked up numbers often enough to justify having any dedicated UI exposed to facilitate finding them. Now, I don’t need to, because the functionality is hidden on the google toolbar.

The great power of the command line is that it can provide access to a nearly infinite amount of functionality and information in a small amount of space. The classical limitation of a command line is that the means to accessing that information is opaque. The user must know before and the proper commands.

This limitation is addressed by the Google example in two ways. First, Google uses what can loosely be termed “natural language processing” to guess at what answers the user is looking for. Then, it uses the space afforded in a web page, and the navigational ease of a point and click interface, to offer those answers to the user.

Google has added a fair amount of functionality to that interface int the past 3 years, and I’d say they still have more they can do.