#!/bin/sh

curl -o /tmp/webkit.dmg http://nightly.webkit.org/builds/Latest-WebKit-CVS.dmg

hdiutil mount /tmp/webkit.dmg

if [ -d /Volumes/WebKit/WebKit.app ]; then 
    rm -rf /Applications/WebKit.app
    cp -pR /Volumes/WebKit/WebKit.app /Applications/
fi

hdiutil detach /Volumes/WebKit

Sure, it could be a little more robust (or you might prefer one of the other two builds), but it works. Put this in a new file, chmod +x it, and drop it into /etc/daily to have it run each morning.

(updated 2005-12-21 to reflect new webkit packaging and fix minor bugs)

(Apparently most all of the new iPods do this, but I never noticed because I have a 3rd generation).

When I started thinking about this topic about a week ago, it was originally going to be a rant about NetNewsWire and how its click-through behavior was so frustrating. Upon doing a little research however, I discovered this problem is not specific to NNW, but all OS X applications, and I’m certainly not the first person to talk about it (John Gruber has a number of interesting reads). I have stumbled upon something however that I’ve been unable to find mentioned elsewhere online, and that’s the inconsistency between single and double click-throughs in an application.

First, a definition from the Apple Human Interface Guidelines to make sure we’re all on the same page, and not talking about banner ads —

An item that provides click-through is one that a user can activate on an inactive window with one click, instead of clicking first to make the window active and then clicking the item.

Often, click-through is discussed in terms of whether or not an inactive application “receives” the click. Developers have the choice of making click-through active on most any object within an application, and there are various schools of thought on what should and should not allow click-throughs. I’m more interested however in “double click-throughs”, or whether an application receives a “double click” message when it is inactive.

An illustration

As I’ve already mentioned, this all begin with an annoyance I discovered in NetNewsWire; here’s my basic workflow

• browse through headlines in NNW
• open full article text in Safari
• after reading article, close Safari window and click back to NNW

This works okay, unless in clicking back to NNW I want to select something other than what is currently selected. For example, if I currently have CNN.com as my selected feed, and I click on my Slashdot feed with NNW being inactive, then all that happens is that NNW becomes the active window. I must then click once again to actually select the Slashdot feed — this is completely expected behavior when click-throughs are turned off. What is not expected however, is that if I perform this second click too quickly, then NNW receives a double-click action and opens the Slashdot homepage in my browser. The problem is that NNW does not allow single click-throughs but it does allow double click-throughs. The only way to make a new selection in NNW while it is inactive is to click once to make it active, pause for a moment, and then make your actual selection. Talk about annoying!

I realize I’m kind of picking on NetNewsWire, but that’s only because this is where I first noticed the issue. Apple applications are no exception - this inconsistency in click-throughs can also be found in iChat’s Buddy List window and Mail message listings (though I’m sure there are many others). iTunes’ track window seems to behave consistently however — neither single or double click-throughs are registered — so apparently it is possible.

The Remedy

I’m certainly not a skilled interface designer, so I’m not about to jump into the debate (religious war?) of when click-throughs should and should not be allowed. I am however quite skilled at the role of “end-user” and can recognize inconsistency when I see it. I’ve only toyed in Interface Builder quite briefly, so I don’t know if it has different properties for single and double click throughs. If not, I would suggest some mechanism for consistency be added, and if it’s already there then I would suggest to developers to use the same value for both.

As a teacher, I’d love to have lectures podcasted and have the ability to embed Word or Keynote docs or hyperlinks somehow as well! I know album art and links to the iTMS are what will pay for some of this, but why not add capacity and watch a new market emerge?! — message board post

During his WWDC Keynote, Steve Jobs demoed the podcasting capabilities of the upcoming version of iTunes and displayed a “New Music Tuesdays” podcast setup by Apple. What made this one particularly interesting is that album artwork changed during the podcast to display the currently playing track. They’ve apparently extended their album art metadata to support multiple images that each span a very specific portion of the file. It’s certainly a very interesting case of what can potentially be done.

However, I’m not so sure that embedding a Keynote presentation into an audio file is the best idea — it definitely seems to be taking the wrong approach. Podcasting aside, an excellent language called SMIL exists to create multimedia presentations consisting of multiple types of embedded media files, and since Quicktime already plays nice it would seem that iTunes could play the files without much work. One could certainly enclose a SMIL file into an RSS feed, but I don’t know if it would integrate as easily as existing podcasts; it would certainly be interesting to find out.

Don’t you just love when you run across a company that just “get’s it”? They just really grok their industry and what it’s all about… they see the big picture, and therefore tend to be the trend setters. There aren’t all that many companies I know of that I’d put in this category — Apple would likely be at the top of the list (though they do have their faults in some areas), but I’d also list Lenny’s Sub Shop, Gotee Records, and Visible School to name a few. Today I ran across another — the English Standard Version Bible (or rather Good News Publishing). This is a Bible publishing company that really cares about standards… how strange is that? Somehow I never knew this, but for the past two years or so they have offered a Web Service to perform queries against their Bible database using parts of Google’s search APIs. Additionally, you can simply use RSS to perform a query if you wish. Online Bible searches are nothing new, but never before has a publisher made it so easy to access their Bibles without having to do it on their site. And just in case there was ever any doubt as to the amount of forethought and planning that went into their online services, they’ve even published a best practices guide. When it comes to publishers, they get it.

Another Longhorn feature talked about was the enhancements to WinFS and Explorer (was called “Windows Explorer”, but I think it may be called “Document Explorer” now). WinFS was originally supposed to be a database powered file system, providing rich metadata and search capabilities. That idea has long since been scratched off the Longhorn road map, but they are still working on an extensive document metadata index. As it turns out, this should be very similar to Spotlight. Longhorn will also have “virtual directories”, akin to “smart folders” in Tiger. Blow for blow, it seems Microsoft and Apple have very similar plans for changing [for the better] how people will interact with their files. I say this not accuse one company of copying the other; on the contrary, I see this paridigm as a rather natural progression based on current trends in the search industry. What really gets me though, is the timeline — Tiger debuts in two days, while the most recently announced debut goal for Longhorn is “Holiday 2006” (given Microsoft’s track record, this can be read as “mid to late 2007 at the earliest”). We’re talking about a 2 year difference here… that is several lifetimes in this industry; I can’t possibly imagine what Apple will be rolling out two years from now.