The iPhone’s launch has brought attention and sophistication to the mobile web. Whether or not you’re sold on Apple’s execution, you probably look forward to the enhancements on other devices that it will inspire.

My second cell phone was from SprintPCS in 2000. It had a web interface that allowed me to read predefined news selections. I can’t recall much else about it. I next switched to Voicestream and got a Nokia 3390 Gold, with AOL IM on it. I’m not sure if I had web access, but I did start text messaging with that phone. From there it was to Cingular to get the new (at the time - in 2004?) Nokia phone that unfolded to expose a full QWERTY keyboard that was split with the screen in the middle. I loved many things about that phone, but the email client was a clunky Java app and the web browser still left much to be desired.

Most recently, two years ago, I got a T-Mobile Sidekick II. Easy to use, the Sidekick has a cheezy interface, but offers POP email, AOL IM, SMS, and web browsing. The web browser on the Sidekick does not read mobile stylesheets. Instead it pulls content through Danger‘s server, formatting and positioning things down the narrow page. I use my Sidekick a lot, most often while on commuter trains. All of my email is copied to it, and I use IM when I’m away from the desk. I use the browser regularly to access sites like Google News, the Huffington Post, Google Reader (mobile), and Digg for the iPhone. I’ve yet to encounter a link I can’t click through and read in full. The only handicap is that I’ve disabled JavaScript to improve rendering times.

Now we have the iPhone with Mobile Safari that renders pages in full and allows you to zoom in and out to read regions of content. Opera Mini, found on Nokia phones and available as a download where Java apps are supported behaves similarly in its next version. Google leveraged the iPhone’s launch to be a real mobile player by embedding its Maps and YouTube apps onto the iPhone.

Now Google’s about to launch a mobile device.

What’s my point in all this? Critical mass is imminent. You can’t ignore this stuff anymore. I wrote about this a few months back and recently took a pass at formatting my content for the iPhone. While the potential offered by these devices is their improvements at making all web sites accessible on mobile, the real promise is in making content usable on them. Optimization. And to create an optimal experience, you need to understand the experience. I’ll likely gravitate toward the iPhone, perhaps replacing my Sidekick with one as my T-Mobile contract is up in November, and I’ve been a bit disappointed with T-Mobile of late. As I incorporate mobile into my practices, I’ll be sharing what I learn here, beginning with some recent announcements and resources I’ve been following:

• Put Your Content in My Pocket by Craig Hockenberry at A List Apart
• Mobile Web Design was released last week by Cameron Moll
• The iPhoney emulator which I, sans iPhone, used to develop C77Studios.com/iPhone
• Opera Mini Simulator - Opera Mini is a popular mobile browser. The simulator enables you to preview your site as it would appear.
• Apple Developer Connect - Web Development for iPhone
• Multi-Safari - If you’ve upgraded to Safari Beta 3 for testing iPhone dev, you can run older versions of Safari for your day-to-day testing
• Added 9/4/2007: The good, the bad, and the ugly - iPhone edition on WebStandards.org

If you’ve got any pointers or just want to connect on mobile development, email mjr@c77studios.com. You can also access my latest posts to del.icio.us and contact info at C77Studios.com/iPhone.

For more posts about mobile web development, visit http://www.c77studios.com/archives/category/mobile.

This post came about when I was linked to on the ExpressionEngine Blog.

Happy as I was when I saw traffic start coming through from a link on the ExpressionEngine Blog, I was dismayed to see that my RSS feed, where the link came from, still had my older entry URL, without the /post. My previous design accommodated multiple entries and single entries on the /index template. ExpressionEngine will serve multiple entries as specified on the {exp:weblog:entries} tag, or filter down to a single entry if it catches its url_title on the URL requested.

But single entries weren’t long enough for the homepage in the new design, and I coded all my single entry links to a /post template. I checked and rechecked my homepage links and archive links, but the RSS feed slipped through the cracks. A single entry link off the homepage made it out into the wild, and I was starting to get some decent traffic on a less-than-perfect page.

I shot off an email requesting a change while working on a solution using my old template as a testpad. What did I use? ExpressionEngine’s segment variable.

The segment variable lets you grab a slice of the URL delimited by ‘/’.

On the URL

http://www.c77studios.com/post/ee_aggregating_content_using_magpie_and_expressionengine/

I could output the value ‘post’ by saying

<? echo {segment_1}; ?>

To do so, you need to turn on PHP parsing for the template (Allow PHP?) in the Template Preferences Manager and set the PHP Parsing Stage to Output, so that the PHP code is run on the value output by ExpressionEngine.

My first step was to test things by doing a quick ‘hello world!’.

<?
if ('{segment_1}' == 'ee_aggregating_content_using_magpie_and_expressionengine')
{
echo 'hello world!';
}
?>

That worked, so now I had a functional conditional to catch the value. Since this was an urgent fix, I wanted to catch this specific circumstance and get them to the right spot.

<?
if ('{segment_1}' == 'ee_aggregating_content_using_magpie_and_expressionengine')
{
header('Location: /post/ee_aggregating_content_using_magpie_and_expressionengine/'); exit;
}
?>

That accommodated this circumstance and eliminated the issue. NOTE:

<?
exit;
?>

is necessary in an ExpressionEngine template to suppress the rest of the page when redirecting using header(’Location:url’);

Crisis averted. But what about other URLs out in the wild? How to keep this from happening again, and how to not have single case code sitting around in my template?

<?
if (strlen({segment_1}) > 0) {
header('Location: /post/{segment_1}'); exit;
}
?>

There. Now it’s dynamic. If a request is made 1) on the /index (homepage) template and 2) has a segment variable on it, grab the segment variable and redirect to /post/{segment_1}. Any old URLs out there with the entry-specific url_title directly on the root will bounce to the /post template, right where I want them to be.

UPDATE: Unfortunately this method isn’t foolproof. I had an inbound link with a querystring parameter (c77studios.com?ref=mjr). Even though there’s no ‘/’ on that URL, the link was translated to c77studios.com/?ref=mjr by the browser or server, which created a segment_1 that doesn’t correspond to a post, which redirected to a dead end. I was able to fix the link, but haven’t yet updated the code to handle it.

For more posts about ExpressionEngine, visit http://www.c77studios.com/archives/category/expressionengine.

About a year ago I put together a spreadsheet in Excel that does a lot of that work for you. I’d definitely call it a 0.5 version of what it could be. It’s pretty basic, and relies on some hard-coded values, but it also puts to use Excel’s date functions to dynamically define durations that skip weekends and holidays, color-code durations by label ("Agency" or “Client” in this version), and do all this just by entering a start date and the number of business days to spend on a task.

A few things to note:

1. This may require one of the Analysis Toolpaks or just a recent version of Excel. I seem to remember adding one of these (Add-Ins under the Tools menu) when I put this together, but my current installation of Excel doesn’t have any checked. Perhaps I found a workaround.
2. This is, of course, unsupported. Download it, play with it, improve it, put your logo on it and call it your own. I’m just sharing some knowledge and a starting point here.
3. Column A is hidden. In Column A I stored the holidays we were given off where I worked so that, in addition to weekend dates defined by Excel, the chart would know additional days to exclude from the schedule.
4. The Start Date in Task 1 is a hard-coded value. The End Date is determined by the number of days allocated, and subsequent Start Dates are determined sequentially following that task. This probably isn’t always the case, but the values can be overridden as you need to.
5. To use this functionality, do a Fill Down (Ctrl+D) on fields B:G for the number of Tasks you need, editing the Task, Days, and Owner fields for each new row.
6. Dates across the timeline are linear and the timeline can be expanded by doing a Fill Right (Ctrl+R). The first date should be entered manually.
7. The duration stripes use conditional formatting. (Conditional Formatting under the Format menu) They’re in an if...else sequence that first blocks off weekends and holidays in gray, then applies a background color for Agency tasks, and then a background color for Client tasks. It does this based on a formula that checks if the date above is within the range of the task (Start Date/End Date, already calculated or manually entered).
8. The values that correspond to “Agency” and “Client” are in the conditional formatting formulas. You can add additional values there.
9. The conditional formatting grid can be expanded out and down by highlighting cells and doing a Fill Down or Fill Right (Ctrl+D or Ctrl+R).

That should be enough to get a somewhat savvy Excel user started. This won’t work magic on its own. It does have enough of the thinking done, though, that it can be expanded and styled to cut down on the number of steps to put together a project schedule.

Download dynagantt.xls

I recently rearranged and redesigned things here on C77 Studios and one of my goals was to pull in external content server-side. I’d been using JSON code to pull in my del.icio.us links, which was frustrating me as the JavaScript call to the remote URL would sometimes stall the loading of the full page.

Magpie is a PHP RSS parser and caching mechanism that pulls RSS feeds from a remote URL and provides an API to grab content and include it in your page output. What’s even better is the Magpie ExpressionEngine plugin that was available by default in my install.

The plugin does the same thing that the PHP Magpie parser does: it provides an API to grab and parse RSS feeds as well as cache them on the server using a timeout parameter. The caching mechanism is useful. Not only does it queue content for fast page rendering if the remote URL happens to be offline, it also prevents bombarding the URL and getting denied. I saw this happen when I reduced the timeout value and found my del.icio.us feed unavailable for a while.

The tag-based syntax for the Magpie ExpressionEngine plugin is really simple:

{exp:magpie url="http://feeds.feedburner.com/ModernHomesNewEngland" limit="4" refresh="60"}
{items}<p><a href="{magpie:link}">{title}</a></p>{/items}{/exp:magpie}

Really simple: Open and close the exp:magpie tag with a feed URL, limit (number of items to show), and a refresh value for the cache (in minutes). For the items pulled, grab bits from each, in this case the title and link for each post.

The example here, that appears in the footer of the page, is from ModernHomesNewEngland, another blog I maintain, so this lets me pull all my blogs together in one portal. I’ve also, as I mentioned, pulled in the latest 12 items from my del.icio.us bookmarks (on the right side of the page), allowing my del.icio.us account to server double-duty: the portable bookmarking tool it’s intended to be, and a tool for me to post quick links to my blog. Finally, I’m using the Magpie plugin to pull my latest Twitter post in as a caption in the photo, capturing a quick thought that I can post anytime, anywhere using my cell phone.

If you maintain a personal blog or homepage, and contribute to other blogs, or wish to pull your del.icio.us posts in as content for your blog, Magpie on PHP is a powerful tool to grab them server-side rather than having a user’s browser do all that work in JavaScript. From a content perspective, this is the ultimate goal: Turn your homepage into a publication with lots of good content and resources. Web 2.0 and the API Web have us posting bits and bites (and bytes) all over the place, and it’s all good content. Magpie helps us easily aggregate all that in one central spot and ExpressionEngine’s tag-based plugin makes it even easier.

For more posts about ExpressionEngine, visit http://www.c77studios.com/archives/category/expressionengine.