Just an announcement that we are going to be discussing Why you MUST have a Twitter Strategy at Atlanta Web Entrepreneurs on August 21, 2008.

I’m going to present a short intro/overview to Twitter and then, god willing and the creek don’t rise, we plan to have two (2) video conferences, one from Triangle Tweetup and the other from a soon-to-be-announced Industry luminary with over 25,000 Twitter followers!

After the 8pm break we’ll have a roundtable-less discussion and Q&A led by our featured participants:

• Sanjay Parekh (sanjay on Twitter),
• Tessa Horehled (driveafastercar on Twitter) and
• Paul Stamatiou (stammy on Twitter).

Anyone that wants to attend should first be sure to have a Twitter account and to follow atlantaweb. We’ll use that list as a roll call for the meeting and we’ll announce our special guest on the atlantaweb Twitter account by 6pm Wednsday August 20th.

For more details and to RSVP see go here.

It’s been almost fifteen years now since the web first hit it’s tipping point and transitioned from an academic’s playground and a mere curiosity for the average person to the decidedly mainstream global change agent that now drives trillions of dollars in global value creation annually. During that time we’ve gone from asking "What’s this ‘World Wide Web’ thingy the geeks keep talking about?" to rapidly seeing Web-based services dominant the activities of practically every business person alive. In the days of the first Internet "gold rush" a.k.a. the "dotcom bubble" it seems that everyone and their brother grabbed a .COM domain, or ten, and set out to strike it rich.  Back then you really didn’t even consider getting anything besides a .COM for your website business but that was okay because many good brand names could be created from still available .COM domains.

Fast forward a decade and great .COM domain names have became scare and even good .COM domain names are hard to come by especially with all the domain squatters.  In addition a lot more top-level domains have opened up, and many small countries such as Tuvalu (.tv) have decided to cash in on their (un)natural resources. And just as fashions change, some creative types who needed a good domain name chose to forgo the .COM status quo and the "www." sub-domain convention and instead compose domain names from words by ignoring the domain level separators (i.e the periods ".") From this trend popular websites with domain names like http://del.icio.us (aka "delicious") were born.

If you are not familiar with delicious it is essentially a website to store your web bookmarks, but storing them "in the cloud" as opposed to in your browser. And one of it’s best innovations, since mimicked by thousands of other sites, it the ability to allow users to categorize with freeform "tags" and later recall their bookmarks by the tags they assigned. These tags are just words, any words your chose, such as "marketing", "video", "php", "bestpractices" or even "shoes."  What words you use to tag with is totally up to you. 

I’ve been using delicious for several years now and at this point not a day goes while surfing the web that I don’t tag at least one website for future reference.  You can even use it to create lists of sites groups by a tag and then send those links to others so they too can see your list of links. But I digress; there is a lot more to delicious but the subject of this post is the .COM domain so we’ll my detailed description of del.icio.us for another day.

As delicious got more popular with the many influencers on the web, Yahoo stepped in and bought them.  Since then delicious has languished for years, still there but never updated. Probably the best thing that has happened to delicious during that period was Firefox built delicious tagging into their browser as their favorites list — if you choose to let Firefox use delicious for you — and the fact that since it hasn’t changed it’s been a pretty stable target for people who wanted to use to delicious API to create add-on functionality and integrations.

So, after years of languishing it turns out Yahoo has been paying attention to delicious behind the scenes, and behold; there is a new delicious! [1]  What’s more, Yahoo has redirected all attempts to access delicious at http://del.icio.us to instead find delicious at http://delicious.com/ and thus, in one fell swoop, have extinguished the quirky domain name that was in part why the web’s tastemakers first took note of delicious. The powers that be at Yahoo probably choose to do this because of usability data I expect they’ve collected that probably told them that the "in crowd" got the funny spelling but that the vast majority of users were simply confused.

Which brings me to the crux of my post where I posit the following:

Are .COM domains still required for commercial success in a mainstream website? Or or those rushing to get domain names with all the new top level domains simply exercising futility?  Were all these idiosyncratic domain names merely a fad and now we’re back to business with .COM, or did Yahoo jumped the shark on this one?[2] 

So what do you think? 

Damon Clinkscales blogged about Twitter Spam last month where he advocated proactively cleansing one follower’s list of "follow spammers" to help reduce the load on Twitter, improve Twitter’s reliability, and increase the value of the Twitter community in general.

I agree!

Still, I think Twitter could take a proactive step reasonably easy that would make it so we don’t have to. I think Twitter could reduce most of the type of Twitter follower spam I got today by applying two simple criteria (And I think Damon also got that same spam today. BTW, nice blog theme Damon! ;-)

I think a strong indication of Twitter follower spam is simply:

1. Their following/follower ratio (or their ing/er ratio for short), and
2. Their follow rate (i.e. how quickly they follow someone after that last time they followed someone.) 

This spammer I got today followed me with 4 different Twitter accounts within a few minutes and each account had around 2000 followings and just over 10 followers making their ing/er ratio about 20-to-1 and I’ll bet their followers were all auto-followed. It’s also clear from the fast & furious tweets that I was not their only mark.

I think it would be reasonable for Twitter to auto-block anyone with a ratio of greater than 15-to-1 ing/er ratio. Twitter could even remove the auto-followers from the calculation; those that follow within around 90 seconds of being followed wouldn’t count as a follower. Doing this Twitter would still give someone the ability to follow 15 people for every one that follows them, and heck they could give them their first 150 people[1] for "free" (i.e. not counting against the limit.) If someone really wants to follow 15,000 people they need to be interesting enough to have at least 1000 people follow them. Shouldn’t be that hard…

Also, Twitter could limit followings per day to, say, 75.  That should be enough for anyone, even the most hard-core twitter newbie (150 "free" + 75 more), and it’s not unreasonable to require a newbie to wait a few days to follow lots and lots of people. 

If I were in charge of setting these limits, I’d set the ing/er ratio to 5-to-1, give them only 25 "free" and then limit to 25 followings per 24 hour period, but I shot high because I was trying to be "reasonable." Of course, Twitter could allow for special cases by allowing people to request to have those limits manually raised if they provide a good justification for it.

What do you think?  Would this work to reduce most Twitter follow spam?  I think so.

I figured that I’d just quickly learn how to build a WordPress plug-in and create one for exposing RESTful web services; after all with a year of programming Drupal modules WordPress’ plug-in API can’t be that hard, right?  Well turns out it wasn’t that easy and I think I have run into a design limitation with WordPress and I’m beginning to wish I’d just taken the brute-force approach and said to hell with writing a plug-in.

Although I am not 100% certain, and I hope someone can point out that I’m just doing something wrong, it seems sadly like I’m pushing the edges of the WordPress API and exposing where it’s design falls short. By the way, I’m working with WordPress v2.5 because why upgrade mid-project when god knows if WordPress will release another in the remaining days before this project is done and I’ll just have to do again before deployment?

Here’s the details.  I started writing a plug-in called “RESTful Services” with a goal of implementing URLs that behave in the following fashion; {format} could potentially be html, xhtml, json, xml, rss, atom, etc.:

http://example.com/services.{format}

Provide a list of RESTful services in specified {format}, defaults to html

http://example.com/services/{service}.{format}/{data}?{params} 

Provide a RESTful service in specified {format}, defaults to html, with optional provided data and parameters.

But before I got all those options working I just wanted to service a page from my RESTful Services plugin where Content-Type: text/plain. I found this page that professes to explain how to hook into the URL routing and after a few fits and starts I can came up with the following code for my plugin that would indeed response to my http://example.com/services URL:

wp-content/plugins/restful-services/restful-web-services.php:

1
2
3
4
5
6
7
8
9
10
11
12
13

add_action('init', 'restful_services_flush_rewrite_rules');
function restful_services_flush_rewrite_rules() {    
  global $wp_rewrite;
  $wp_rewrite->flush_rules();
}       
 
add_filter('generate_rewrite_rules', 'restful_services_add_rewrite_rules');
function restful_services_add_rewrite_rules( $wp_rewrite ) {      
  $new_rules = array(        
    'services' => 'wp-content/plugins/restful-services/rest.php',     
  );
  $wp_rewrite->rules = $new_rules + $wp_rewrite->rules;
}

The problem with the above was that it wouldn’t call “wp-content/plugins/restful-services/rest.php“; it would simply continued to call “index.php” and display the home page!  After literally hours and hours of debugging with my trusty PhpEd IDE & debugger I was able to find that the code on lines 737 & 738 of “wp-includes/query.php” told WordPress that my service was the home page! It is almost seems like the “generate_rewrite_rules” was implemented as “a good idea” yet no testing has ever been done on it because for the best I can tell it doesn’t work. (Note I’ve reformatted the code to multiple lines so that it is easier to read and does not extend past the right margin of my blog):

wp-includes/query.php:

if ( !(  $this->is_singular                           
      || $this->is_archive                           
      || $this->is_search                           
      || $this->is_feed                           
      || $this->is_trackback                           
      || $this->is_404                           
      || $this->is_admin                           
      || $this->is_comments_popup ) )   
       $this->is_home = true;

I could possibly hack it to get past this by setting one of those to “true”, but none of them are really appropriate; there is nothing there quite like an “is_service” instance variable. Setting something like “this->is_singular” or “this->is_feed” might work but it could manifest incompatibility problems with other plugins or future versions of WordPress. Frankly it was rather disappointing to discover this because it tells me that WordPress has hard-coded all the potential scenarios and doesn’t really have a way around it. Seems to me there should really be a hook here and the type of pages should be allowed to be expanded by plugins rather than be hardcoded as only one of ’singular’, ‘archive’, ’search’, ‘feed’, … and ‘home.’

Anyway, where this manifests itself is “wp-includes/template-loader.php” file which I have included in it’s entirety below.  It is on lines 24 and 25 where the template loader loaded the home page because it’s not possible to specify otherwise:

wp-includes/template-loader.php:

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77

/**
 * Loads the correct template based on the visitor's url
 * @package WordPress
 */
if ( defined('WP_USE_THEMES') && constant('WP_USE_THEMES') ) {
  do_action('template_redirect');
  $is_home = is_home() ;
  if ( is_robots() ) {
    do_action('do_robots');
    return;
  } else if ( is_feed() ) {
    do_feed();
    return;
  } else if ( is_trackback() ) {
    include(ABSPATH . 'wp-trackback.php');
    return;
  } else if ( is_404() && $template = get_404_template() ) {
    include($template);
    return;
  } else if ( is_search() && $template = get_search_template() ) {
    include($template);
    return;
  } else if ( is_home() && $template = get_home_template() ) {
    include($template);
    return;
  } else if ( is_attachment() && $template = get_attachment_template() ) {
    remove_filter('the_content', 'prepend_attachment');
    include($template);
    return;
  } else if ( is_single() && $template = get_single_template() ) {
    include($template);
    return;
  } else if ( is_page() && $template = get_page_template() ) {
    include($template);
    return;
  } else if ( is_category() && $template = get_category_template()) {
    include($template);
    return;
  } else if ( is_tag() && $template = get_tag_template()) {
    include($template);
    return;
  } else if ( is_tax() && $template = get_taxonomy_template()) {
    include($template);
    return;
  } else if ( is_author() && $template = get_author_template() ) {
    include($template);
    return;
  } else if ( is_date() && $template = get_date_template() ) {
    include($template);
    return;
  } else if ( is_archive() && $template = get_archive_template() ) {
    include($template);
    return;
  } else if ( is_comments_popup() && $template = get_comments_popup_template() ) {
    include($template);
    return;
  } else if ( is_paged() && $template = get_paged_template() ) {
    include($template);
    return;
  } else if ( file_exists(TEMPLATEPATH . "/index.php") ) {
    include(TEMPLATEPATH . "/index.php");
    return;
  }
} else {
  // Process feeds and trackbacks even if not using themes.
  if ( is_robots() ) {
    do_action('do_robots');
    return;
  } else if ( is_feed() ) {
    do_feed();
    return;
  } else if ( is_trackback() ) {
    include(ABSPATH . 'wp-trackback.php');
    return;
  }
}

Still another problem in this puzzle is the $wp->send_headers() method shown being called here on line 293 of “wp-includes/classes.php”:

wp-includes/classes.php:

290
291
292
293
294
295
296
297
298

function main($query_args = '') {   
  $this->init();   
  $this->parse_request($query_args);   
  $this->send_headers(); 
  $this->query_posts();  
  $this->handle_404();   
  $this->register_globals();   
  do_action_ref_array('wp', array(&$this));   
}

The problem with the $wp->send_headers(), also from “wp-includes/classes.php”, is that it seems to have the option of either serving an HTML content type on line 183 and 185, or a content type based on a feed (the content types for the feeds are set in their respective “wp-includes/feed-*.php” files) but no custom content types as far as I can determine as there seems to be no way to override calling this function or the logic path contained within:

wp-includes/classes.php:

175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227

function send_headers() {
  @header('X-Pingback: '. get_bloginfo('pingback_url'));
  if ( is_user_logged_in() )
    nocache_headers();
  if ( !empty($this->query_vars['error']) && '404' == $this->query_vars['error'] ) {
    status_header( 404 );
    if ( !is_user_logged_in() )
      nocache_headers();
    @header('Content-Type: ' . get_option('html_type') . '; charset=' . get_option('blog_charset'));
  } else if ( empty($this->query_vars['feed']) ) {
    @header('Content-Type: ' . get_option('html_type') . '; charset=' . get_option('blog_charset'));
  } else {
    // We're showing a feed, so WP is indeed the only thing that last changed
    if ( !empty($this->query_vars['withcomments'])
      || ( empty($this->query_vars['withoutcomments'])
        && ( !empty($this->query_vars['p'])
          || !empty($this->query_vars['name'])
          || !empty($this->query_vars['page_id'])
          || !empty($this->query_vars['pagename'])
          || !empty($this->query_vars['attachment'])
          || !empty($this->query_vars['attachment_id'])
        )
      )
    )
      $wp_last_modified = mysql2date('D, d M Y H:i:s', get_lastcommentmodified('GMT'), 0).' GMT';
    else
      $wp_last_modified = mysql2date('D, d M Y H:i:s', get_lastpostmodified('GMT'), 0).' GMT';
    $wp_etag = '"' . md5($wp_last_modified) . '"';
    @header("Last-Modified: $wp_last_modified");
    @header("ETag: $wp_etag");
 
    // Support for Conditional GET
    if (isset($_SERVER['HTTP_IF_NONE_MATCH']))
      $client_etag = stripslashes(stripslashes($_SERVER['HTTP_IF_NONE_MATCH']));
    else $client_etag = false;
 
    $client_last_modified = empty($_SERVER['HTTP_IF_MODIFIED_SINCE']) ? '' : trim($_SERVER['HTTP_IF_MODIFIED_SINCE']);
    // If string is empty, return 0. If not, attempt to parse into a timestamp
    $client_modified_timestamp = $client_last_modified ? strtotime($client_last_modified) : 0;
 
    // Make a timestamp for our most recent modification...
    $wp_modified_timestamp = strtotime($wp_last_modified);
 
    if ( ($client_last_modified && $client_etag) ?
         (($client_modified_timestamp >= $wp_modified_timestamp) && ($client_etag == $wp_etag)) :
         (($client_modified_timestamp >= $wp_modified_timestamp) || ($client_etag == $wp_etag)) ) {
      status_header( 304 );
      exit;
    }
  }
 
  do_action_ref_array('send_headers', array(&$this));
}

Still, I was able to come up with a solution although it is so very hackish.  My solution was to hook the “template_redirect” action on line 7 of “wp-includes/template-loader.php” (see code from that file above.) Though it seems to works thus far, my solution just feels wrong for the following reasons:

1. It ignores the fact that WordPress continues to think that my web service URL is the home page,
2. It first lets “$wp->send_headers()” set the content type before it overrides it,
3. It uses an “exit” rather than a return to keep WordPress from serving up the home page template, and
4. It doesn’t use the routing mechanism apparent built into WordPress (see “null” on line 30 of “wp-content/plugins/restful-services/restful-web-services.php” below, I assume it should have been the URL of the .php file I plan to execute but WordPress doesn’t see to use what I put there.)

The function “restful_web_services_exec_service()” is what is called to execute the appropriate web service:

wp-content/plugins/restful-services/restful-web-services.php:

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55

/*
Plugin Name: RESTful Web Services
Plugin URI: http://mikeschinkel.com/wordpress/restful-web-services/
Description: This plugin enables REST-based API web services for WordPress.
Author: Mike Schinkel
Version: 0.1
Author URI: http://mikeschinkel.com/
*/
 
define('RESTFUL_WEB_SERVICES_DIR', dirname(__FILE__));
define('RESTFUL_WEB_SERVICES_URL_PATTERN','services(/?.*)?');
$abspath = trim(str_replace('\\','/',ABSPATH),'/');
$rest_services_dir = str_replace('\\','/',RESTFUL_WEB_SERVICES_DIR);
$rest_services_path = trim(str_replace($abspath,'',$rest_services_dir),'/');
define('RESTFUL_WEB_SERVICES_PATH', $rest_services_path);
 
// NOTE, See: http://codex.wordpress.org/Custom_Queries#Permalinks_for_Custom_Archives
add_action('init', 'restful_web_services_flush_rewrite_rules');
add_filter('generate_rewrite_rules', 'restful_web_services_add_rewrite_rules');
add_action('template_redirect', 'restful_web_services_exec_service');
 
add_action('init', 'restful_web_services_flush_rewrite_rules');
function restful_web_services_flush_rewrite_rules() {
 global $wp_rewrite;
 $wp_rewrite->flush_rules();
}
 
add_filter('generate_rewrite_rules', 'restful_web_services_add_rewrite_rules');
function restful_web_services_add_rewrite_rules( $wp_rewrite ) {
  $new_rules = array(
    RESTFUL_WEB_SERVICES_URL_PATTERN => null,
  );
  $wp_rewrite->rules = $new_rules + $wp_rewrite->rules;
}
 
function restful_web_services_exec_service() {
  global $wp;
  if ($wp->matched_rule==RESTFUL_WEB_SERVICES_URL_PATTERN) {
    if ($wp->request == 'services') {
      header('Content-Type: text/plain');
      print 'TODO: Generate a list of RESTful Web Services here for this WordPress Blog.';
    } else {
      list($dummy,$service_name) = explode('/',$wp->request);
      if (file_exists($service_php = (RESTFUL_WEB_SERVICES_DIR . '/services/' . $service_name . '.php'))) {
        include_once $service_php;
      } else {
        header('Content-Type: text/plain');
        status_header(404);
        print '404 - Service not found.';
      }
    }
    exit;
  }
}

You’ll note that my function “restful_web_services_exec_service()” is very bare-bones at the moment serving only a plain text message “TODO:” for the path http://example.com/services, and assuming that any path http://example.com/services/{service} will execute a same-named .php file in the services subdirectory, i.e. for http://example.com/services/latest-post it will look for “wp-content/plugins/restful-web-services/services/lastest-post.php” and then delegate all the work to that .php file.

wp-content/plugins/restful-services/services/latest-vidclip.php:

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

/*
Filename: latest-post.php
Service Name: Latest Post
*/
global $wp_query;
$post = $wp_query->post;
$file = get_attached_file($post->ID);
if (empty($file)) {
  list($file)= get_enclosed($post->ID);
}
$charset = get_option('blog_charset');
header('Content-Type: text/xml; charset=' . get_option('blog_charset'), true);
$link = get_permalink($post->ID);
$html = < <<POST
<?xml version="1.0" encoding="$charset" version="1.0">
<post id="{$post->ID}">
  <title>{$post->title}</title>
  <video>$file</video>
  <link>$link</link> 
</post> 
POST;
print $html;

Here is an example output returned by calling http://example.com/services/latest-post:

1
2
3
4
5
6

< ?xml version="1.0" encoding="UTF-8"?> 
<post id="2">
  <title>Sample Post #1</title>
  <video>http://videos example.org/video1.flv</video>
  <link />http://example.org/sample-post-1/ 
</post>

While bare-bones, that will meet my needs for the moment since I really only have the need for one service that responds to the HTTP GET verb. However, I can see where I might end up fleshing this out and create lots of helper functions that would streamline the creation of web services for RESTful access to the entirety of the WordPress database. If I do so I’ll be happy to donate this plugin to the community.  In the mean time if you’d like to use this for your own use feel free but caveat emptor. However, if you’d like to use this code to create a plugin to contribute to the community, please contact me for collaboration.  Of course if you’d like to retain me to fully flesh it out for you, I’m always open to that too! :-)

Finally, if anyone who knows the WordPress API better than I do can tell me where I erred in my analysis, and I really hope I did, please let me know so I can architect this thing better. On the other hand, if I was spot-on in my analysis maybe this will help the WordPress team understand what needs to be done so they can empower WordPress to generate any arbitrary content type without having to resort to hacks or bypassing the WordPress index.php startup code. 

Enjoy.

FCKeditor ShowBlocks Feature

The other day I wrote about how I was using Dean’s FCKeditor Plugin for Wordpress to solve the problems with WordPress’ default TinyMCE editor eating my hand-coded HTML tags.  Well Joe Banks left a great and detailed comment giving several tips including one about the ShowBlocks button. Unfortunately the default comment system in WordPress ate part of his comment so I couldn’t see what he was talking about and went exploring. 

Well it turns out he was right; ShowBlocks is awesome!  It shows the <p> tags, the <div> tags, and more. Check the screen shot above to see a larger version.

Joe Banks forgot to include his blog URL on his comment (or at least I think he forgot), but I was able to dig it up via a quick google.

This month at the Atlanta Web Entrepreneurs meetup group I organize we hosted two sharp email marketing professionals: Sandi Karchmer Solow of I Send Your Email and Ben Chestnut, co-founder of Mail Chimp, a successful Atlanta-based Email Service Provider. Sandi presented Email Marking 101 to the group, and Ben regaled us with his story of how MailChimp came to be.

Sandi gave us a really great base level of overview of the email marketing landscape and explained how its critical to correctly opt-in your subscribers and to give them exactly what they asked for, and only what they asked for. Otherwise you loose trust and the fallout is worse than anything you could gain. Oh, and Sandi was a real trooper to speak this month because she’s about seven month’s pregnant. So good luck to her and her soon-to-be-newborn.

As for MailChimp, evidently it was a side project that Ben and his partner’s web consulting company implemented to keep a client who wanted them to manage his email broadcast from hassling them, but they didn’t fully embrace it as their primary offering until many years later. And the month after they fully embraced it their revenue exceeded every prior month’s revenue they’d seen life-to-date for their business! Ben told us how MailChimp has a focus on simplicity and when we reviewing his prices we found MailChimp to be very price competitive, especially for email lists of less than 100 which they send for free!

Now most marketers have heard of ExactTarget before but many may not have heard of MailChimp, and based on MailChimp’s low pricing, it simple-to-use interface and its fun and irrerevent name, many people might think that MailChimp is only for businesses with tiny email lists. But most in the audience including myself were shocked to learn that they have successfully delivered some of the largest email broadcasts in the industry!  Ben told us about a major software launch announcements where they sent out millions of of emails in just about 30 minutes! (Ben said the client asked never to be named but believe me, it was major!)

What was especially interesting was when member/attendee Jason Prance mentioned during Q&A that he’d been using both MailChimp, for personal projects, and ExactTarget for a 100,000 name work mailing list, and that he loved the former and really disliked the latter. He then said if he had his druthers he’d be using MailChimp for work but couldn’t switch without re-opting in and loosing probably half his subscribers. To this Ben replied that all he’d need to do is provided his ExactTarget reports showing them being a responsible emailer and then he could easily move his 100k list to MailChimp. Sold!

Anyhoo we had a great time, enjoyed learning about email marketing, and look forward to future Atlanta Web Entrepreneur meetups. Oh, and I want to thank both Sandi and Ben for taking the time to make such a memorable evening for us.  It really is great to have such nice people who are willing to help their peers and who are offering such worldclass services so reasonably priced, all here in our hometown of Atlanta GA. Go Atlanta!

Visit Flickr to see all photos I took for this event.

P.S.: This was NOT a paid endorsement for MailChimp. We invited Ben to speak about MailChimp because one of our members that we really respect recommended him very highly. Plus Ben turned out to be a really great guy and there were actually several members in attendance who already use his service and love it. Evidently, MailChimp really kicks ass!

Photo by Nemo’s great uncle

Several months back we had a meeting on Leveraging Mobile Apps for your Web-based Business at Atlanta Web Entrepreneurs and on the topic of mobile push marketing the consensus of the attendees was overwhelmingly “Don’t you even think about it, or I’ll end up ramming my mobile device far, far up someplace you’d really it rather not be.” Or something like that. Soon after I made a comment on Douglas Karr’s blog  on his post about bluetooth proximity marketing saying exactly that.

Well I just noticed that Michael Katz of AdMarkTech.Com posted a response entitled Do We Really Know What We Want? where he takes the position that bluetooth proximity marketing is less intrusive that television advertising because when watching TV you are not "in the shopping zone."  He then posits that:

"Maybe Bluetooth proximity marketing seems more intrusive than other forms of mainstream advertising because we’re just not conditioned to be approached while we shop."

He then claims that (with emphasis mine):

 "bona fide Bluetooth proximity marketing campaigns, similar to contextual vertical advertising platforms, are no different to television commercials"

And he summarizes with:

"The capabilities of two way push, receiving surroundings based information, on-the-spot offers and the ability to instantly reject or to receive based on Bluetooth activation preference, Bluetooth proximity marketing is actually less intrusive and potentially more helpful than television advertising."

Maybe.  But television advertising is not a fair comparison to bluetooth proximity advertising. First the TV viewer (or radio listener or magazine/newspaper reader or website viewer) makes a proactive choice to engage in a medium that they know is supported by advertising and that is part of an implicit pact where people gain content in exchange for attention. In the case of walking by a store, a person may or may not be in the frame of mind where they am ready to perform an attention exchange for an advertisement, and if they are not it is highly intrusive. Sure people could potentially be "conditioned" to accept it (shades of Apple’s 1984 campaign?) and he may be right, but I doubt it for the next reason.

There is a natural limit to advertising that can be inserted into content before people will no longer accept the exchange; only so much ad time on a TV or radio show, only so many square inches in a magazine or newspaper, and only so many square pixels close to valuable content available on a web page.  In the case of bluetooth proximity marketing it’s likely a person walking down a mall will be inundated with ads; literally tens if not hundreds in a short period, and that is far more than anyone can process.

In my opinion that’s a major reason why people hate email spam; that it overwhelms them. If we could somehow configure the number of commercial emails we’d get per day, and we could have them filtered by our preferences I think we’d be happy to get "unsolicited" commercial email. But there is no moderation on on spam, and without moderation on bluetooth proximity marketing it will just be another form of spam, only even more offensive because the devices are smaller.

Can there be moderation on bluetooth proximity marketing?  Ads on TV and radio, for example, come in series. Bluetooth proximity marketing would comes in parallel as does spam, and with nothing to moderate it. One possible way to moderate would be for governments to set up regulatory agencies to manage and meter access to said marketing but that sounds like a cure worse than the disease. Another potential way would be gatekeeping companies to control who can broadcast bluetooth proximity marketing messages to their subscribers and who cannot, and people would subscribe based on who does the best job of filtering. But I think a lot of visionary people in very powerful and competing positions would have to come together to make such a network of gatekeeper possible, and I just don’t see it happening.

Bluetooth proximity marketing is a typical example of how most people spend too much time on what they want and too little time creating value. Michael’s post title (Do We Really Know What We Want?) is a great example of that.  It sounds to me that Michael is asking "How can we convince people to want what we want?" That is wrong-headed, we should instead be thinking "What do they want, and how can we make a business to give it to them?"

Here’s a product I know people will really want instead; something that blocks all bluetooth proximity ads from ever reaching their mobile device.  Now that is a product of tomorrow that will be in VERY HIGH demand!

But rather than blog my full thoughts on Twitter right now I just want to ponder the following; if a web page can have a URL a.k.a. Uniform Resource Locator (or for w3c purists, a URI), shouldn’t we establish a Uniform Person Locators, for people, or "UPLs" for short? 

I’ve actually pondered this question numerous times in the past as I contemplated one of my many web entrepreneurial ideas.  In one context I’m lucky; there are very, very few people in this world whose name is "Mike Schinkel", and the few I’m aware of are not at all active on the web so I pretty much "own" the "MikeSchinkel" as my uniform person identifier.  I’ve so blanked the use of "MikeSchinkel" on the web that there is little change of anyone else actually wanting to use it lest they be confused with me (that is, unless they wanted to create that confusion. But that is another subject.)

But what about my friend David Cohen?  From him I understand that he has the opposite situation.  For him there are hundreds if not thousands of other David Cohens (to link but a few.) Poor guy; how does he get his name known for being him? But I digress.

Yes we can all have our own domain names, like I have http://mikeschinkel.com and my friend has http://davidcohen.com (can you believe he got that domain, with all that competition?!?), but so many people don’t have their own domain and for some people it takes more knowledge or effort then they are willing or able to invest. On the other hand getting a Twitter account is free to create and free to maintain and only requires 5 minutes and then occasional access to an Internet-connected computer which in the USA and most non-3rd world countries can be had at the library or an Internet cafe. Once a twitter account is created, that’s it; no one else can stake that claim. 

Now, thanks to some rather saavy engineers at Twitter and the brilliance of the underlying technology of the web (i.e. the URL) we now have what could potentially become the Uniform Person Locator, at least within the subset of people are active on the web.  With someone’s Twitter URL you have a direct way to "locate" them. Minimally you can follow their updates, but for most Twitter users you can relatively easily contact them; just tweet them and most will reply. Just as importantly you can use someone’s Twitter user name and unambiguously refer to them by that unique identifier which behaves for people just like URLs behave for "resources" (i.e. web pages, PDF files, graphic files, .ZIP files, etc.)

Will Twitter ever make it’s way to effectively being the "Uniform Person Locator."  Probably not for all people, but at least for the subset of people on Twitter, it is a really interesting piece of infrastructure to consider.  And if you don’t yet have a Twitter account, now’s the time to get one.

P.S. So can you guess my Twitter user name?  "MikeSchinkel", of course.   David Cohen’s Twitter user name?  "davidscohen"; Unlike with his domain name, "davidcohen" had already been taken on Twitter (by a duffus that’s not even using it!) before my friend David Cohen could grab it. Ah, such is life when resources are scarce.