Technorati Makes Breaking Change to Multi-Word Tag Syntax: Hyphens Replace Plus Signs – Epic #Fail

---

Update 11/15/2009: Multi-word tags with “+” word separators and search feeds now seem to be working. The following capture resulted on 11/15/2009 from a SQL Azure Database (SQL+Azure+Database) tag:

 

Update 11/13/2009: @TechnoEric said “[W]e really should get the old tags working with our new system; you may not need to change.” in an 11/9/2009 tweet.

Update 11/12/2009: Brandon Turner of the Windows Live Writer (WLW) team responded to my tweet regarding updating WLW as follows in an 11/9/2009 email:

Sadly, we don’t have a work around in our code right now, if this is a problem for you, then you might want to look into a 3^rd party plugin.

It’s strange they want to use dash instead of a plus sign, when all other blog tag aggregators seem to still use a plus sign.  This is probably because URL encoding requires spaces be convert to a + sign or %20. 

None the less, I have filed a bug to look into the issue.  We’ll see what we can do. 

---

While testing Technorati tags generated by Windows Live Writer in my recent blog posts, I discovered that multi-word tags, such as Windows Azure (Windows+Azure) were returning hundreds of links to posts apparently tagged Windows.

Similarly, posts tagged SQL Azure (SQL+Azure) or SQL Azure Database (SQL+Azure+Database) returned links to posts having no conceivable relationship to Microsoft’s new cloud-based relational database (click image for full-size capture):

 

Following is a transcript of my Twitter conversation with @TechnoEric about this problem:

1. From @rogerjenn 11/8/2009: @technorati: Tag-based search with phrases (word1+word2) ignore the second and higher terms. Try any link on http://bit.ly/2wAQGC. #Fail.
2. From @rogerjenn 11/8/2009: @technorati: The http://technorati.com/tag/W... URL returns mostly Windows 7 posts, neglecting the Azure restriction. Epic #Fail.
3. From @TechnoEric 11/9/2009: @rogerjenn @technorati:for multi-word tags try dash: http://technorati.com/tag/W... Not perfect, but better.
4. From @rogerjenn 11/8/2009: RT @TechnoEric: Dash for multi-word tags is better, but requires manually editing tags created by Windows Live Writer. Why the change? #Fail

Substituting hyphens (-) for plus signs (+) does, indeed, improve the results, as indicated by SQL Azure (SQL-Azure) or SQL Azure Database (SQL-Azure-Database) links:

As I mentioned in my reply to @TechnoEric, this unannounced (and unwarranted, in my opinion) change requires manually editing the Technorati tags generated by Windows Live Writer, my preferred blog editor. However, this gratuitous change renders non-functional about 10,000 tags in my earlier blog posts.

I’ve contacted wlw-team@microsoft.com to request the option of changing the space separator to and from plus signs in new posts. I’ll update this post, if and when I hear from @TechnoEric or the Windows Live Writer team. I’ve temporarily reverted to Google Alerts for cloud-related blog posts.

Technorati TagName Articles Pages Are Missing the RSS Feed Icon: Another #Fail

Another issue I encountered after Technorati overhauled their site was the loss of an RSS feed icon on the page(s) returned by tag URLs. Previously, I was able to add feeds to my reader that returned links to updated tag-specific pages, which Technorati calls “search feeds.” Clicking Internet Explorer 8’s RSS feed icon returns an “All Items” feed, which, of course, is totally useless:

There are several recent @TechnoEric tweets that mention “Technorati search feeds went away w/redesign. Back soon. You're getting original articles feed” or the like.

Technorati has invested thousands, perhaps millions, in an attempt to upgrade its service. The foolishness described in this post indicates a lack of attention by Technorati to the needs of the majority of the site’s contributors.

Technorati Tags: Technorati,Tags,Tagging,Search Feeds

Does Anyone But Me Miss Technorati’s RSS Feeds for Specific Tags?

I’ve subscribed to Technorati’s tag-based RSS 2.0 feeds for more than a year. Tag-based feeds provided me links to new posts tagged LINQ, Entity Framework, LINQ to XML, Azure, Cloud Computing and the like. These links often are in addition to those from blogs with feeds to which I subscribe.

On about 1:30 AM on 12/17/2008, IE 8 Beta 2’s feed reader suddenly started returning error messages for popular tag feeds, such as those for LINQ:

Firefox 3.0.4 returned the infamous “Doh! The Technorati Monster escaped again” page.

RSS Bandit displayed this IE error message:

The following tags were not closed: rss, channel. Error processing resource 'http://feeds.technorati.com/tag/LINQ'.

Here’s Fiddler2’s capture of a current request for recent LINQ posts:

GET /tag/LINQ HTTP/1.1
Accept: */*
A-IM: feed
Accept-Language: en-us
User-Agent: Windows-RSS-Platform/2.0 (MSIE 8.0; Windows NT 6.0)
UA-CPU: x86
Accept-Encoding: gzip, deflate
Connection: Keep-Alive
Host: feeds.technorati.com
Pragma: no-cache
Cookie: tvisitor=66.123.163.242.1229281145878383; TECHNORATI_MEMBER=497505af64ce1835caa9fa1ece590cb7; __qca=492dc786-a1a11-5e3a1-d9b67; __utma=39112614.974384128884774300.1229550691.1230230604.1230305278.9; __utmz=39112614.1230230604.8.5.utmcsr=oakleafblog.blogspot.com|utmccn=(referral)|utmcmd=referral|utmcct=/; s_vi=[CS]v1|494FD9BB00003618-A0208390000043A[CE]; VWCUK200=L122208/Q22273_3696_1056_122208_1_033109_94744x91717x122208x1x1

and following is the malformed response with XML formatting added:

HTTP/1.1 200 OK
Date: Tue, 30 Dec 2008 20:29:33 GMT
Server: Apache
Cache-Control: max-age=1800
Expires: Tue, 30 Dec 2008 20:59:33 GMT
Content-Type: application/rss+xml
Vary: Accept-Encoding,User-Agent
Content-Length: 529

<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:tapi="http://api.technorati.com/dtd/tapi-002.xml">
  <channel>
    <title></title>
    <link></link>
    <description></description>
    <pubDate>Tue, 30 Dec 2008 12:29:34 -0800</pubDate>
    <generator>Technorati v1.0</generator>
    <image>
      <url>http://static.technorati.com/pix/logos/logo_sm.gif</url>
      <title>Technorati logo</title>
      <link></link>
    </image>
    <webMaster>support@technorati.com (Technorati Support)</webMaster>
    <docs>http://blogs.law.harvard.edu/tech/rss</docs>
    <ttl>60</ttl>

It’s not surprising that all three feed readers returned error messages, but the different response to Firefox requests is strange. Some requests from IE 8 B2 and RSS Bandit, such as for SQL Data Services tags (GET /tag/SQL+Data+Services HTTP/1.1) return the “Technorati Monster” page.

Technorati Won’t Acknowledge the Problem or Respond to Requests to Fix It

I’ve sent numerous messages to support@technorati.com which Technorati ignores. They don’t acknowledge the messages nor do they respond to them, whether sent by email or Twitter.

Technorati Stopped Indexing the OakLeaf Blog 12 Days Ago and No Longer Resumes Indexing When Requested

Starting in early September 2008, Technorati periodically stops indexing the OakLeaf Systems blog. Manual pings don’t solve the problem. In the past, they have resumed indexing this blog for brief periods after I send an email to support, but then stop indexing after a few days.

Scott Watermarsysk replied to my Twitter complaint about Technorati’s earlier failure to index the OakLeaf Systems Blog:

Now they haven’t indexed the blog in 12 days despite several requests, which they won’t acknowledge:

Here’s a reverse chronological list of my posts about earlier problems with Technorati:

• Technorati Stops Spidering the OakLeaf Blog Again! (10/24/2008)
• Technorati Troubles Continue (9/19/2008)
• Technorati Ranks Mike Ormond’s Blog #8 with Authority of 11,764 (9/14/2008)
• Technorati’s Problem with IE7’s Feed Reader (9/12/2008)
• Technorati Contracts Twitteritis (9/12/2008)
• More Technorati Troubles (9/11/2008)
• Technorati Fails to Index OakLeaf Systems Blog Posts (9/2/2008)

It appears to me as if Scott is right: No one has told Technorati that they’re dead.

What’s worse is that it appears to me that I’m the only one who cares. Searches for Technorati in Google’s Blogger Help Group doesn’t return any indication of tag RSS problems detected.

Update 12/31/2008: Technorati Responds

Technorati indexed this blog on 12/30/2008. Tag feeds aren’t up. I’ll update this post when and if they appear.

Technorati Tags: Technorati,Blogger

Technorati Stops Spidering the OakLeaf Blog Again!

For reasons unrevealed to me by Technorati, they’ve ceased pinging this blog for about the tenth time in the past three to four months.

Here’s the result on 10/22/2008 after manually requesting a ping on 10/20/2008:

Today is at least the sixth time I’ve had to send e-mail to Technorati’s “technical support” folks to restore their spidering of this site. They don’t reply to the request, but usually I see a recent post in most (but not all) of the appropriate Technorati tag groups.

Crazy!

Click here for details of my previous Technorati problems.

Update 10/23/2008 7:30 AM PDT: I didn’t receive the promised “Support Ticket,” but Technorati did ping the site “4 hours ago” and tag-based feeds are working again. It’s a PITA to manually request "technical support” to spider the site, but perhaps I’ll write a small program to automatically send Technorati a message each Monday or Tuesday.

As mentioned in my earlier posts, Technorati attempts to filter “splogs” from spidering operations. If they’re using an algorithm based on the ratio of the number of outbound links to post length, the OakLeaf blog should have a much lower splog index than, for example, Scott Guthrie’s Link Listing series.

Update 10/24/2008 3:00 PM PDT: It’s clear that Technorati is embargoing the OakLeaf Systems blog. My site hasn’t been spidered since 10/23/2008 AM, despite a manual ping on this morning. Sending another message via email and Twitter.

Technorati Tags: Technorati

Technorati Troubles Continue

Update 9/19/2008 12:0024 noon: ALLMost of my tag feeds are now working:

• http://feeds.technorati.com/tag/SQL+Server+Data+Services
• http://feeds.technorati.com/tag/Entity+Framework
• http://feeds.technorati.com/tag/Live+Mesh
• http://feeds.technorati.com/tag/LINQ
• http://feeds.technorati.com/tag/ASP.NET+Dynamic+Data
• http://feeds.technorati.com/tag/ADO.NET+Data+Services

This wasn’t:

• http://feeds.technorati.com/tag/ADO.NET+Data+Services

Some tag feeds didn’t come back until I replaced encoded spaces (%20) with + or removed query string extensions added by Technorati. The tags with spaces and query strings had worked previously.

End Update

Rob Diana’s Technorati: How the Mighty Have Fallen Mashable post of 9/15/2008 begins with this observation:

Late last week, Technorati was down for over 12 hours. I checked Friendfeed and Twitter, and there was barely a mention of it. I was surprised that Technorati being down garnered so little conversation. It seemed like Technorati went down, and nobody cared. The most recent blog entry only mentions an outage from the previous week, but the Technorati Twitter account does mention “A major bout with bots.”

One reason that Rob didn’t see much Twitter activity regarding the outage was too narrow a search term (technorati+down). A wider search on technorati alone would have uncovered my 15 tweets about the various problems I’ve been having with Technorati’s services over the past few weeks, to wit:

• Technorati stopped indexing my posts
• Technorati didn’t recognize my manual pings
• Technical service didn’t respond to my inquiries
• Eventually Technorati started indexing some of my posts
• Those posts that were indexed weren’t included in all tag pages specified
• RSS feeds for tag pages have been down since about 8/25/2008

Yesterday (Monday 9/15/2008), Technorati wrote the following on Twitter:

After our outage last week, we had to turn off Technorati Feeds and API. These will be coming online again on Monday. Sorry.

Yesterday was Monday, today is Tuesday, and tag feeds are still down. I’m surprised that no one but me appears to have blogged about that missing feature.

Rob asks:

When did people stop caring about Technorati? Is it the lack of new features? Or, is the decline just due to the fact that people do not care about what it offers any more? I find the lack of interest to be odd. Bloggers live by the numbers. They spend plenty of time checking the daily traffic statistics and subscriber levels. Technorati authority used to be one of the statistics that was checked often. So, what happened? The problem is that nothing happened.

And concludes that Technorati has stagnated at about 2.7 million unique visitors per month for the past year.

Instead of serving their blogger constituency, Technorati has made dubious acquisitions, such as Personal Bee and Blogcritics. Truly, nothing has or is happening at Technorati.

Following are links to my recent complaints about Technorati operations and customer service:

• Technorati Ranks Mike Ormond’s Blog #8 with Authority of 11,764 (9/14/2008)
• Technorati’s Problem with IE7’s Feed Reader (9/12/2008)
• Technorati Contracts Twitteritis (9/12/2008)
• More Technorati Troubles (9/11/2008)
• Technorati Fails to Index OakLeaf Systems Blog Posts (9/2/2008)

Technorati tags: Technorati, Technorati Troubles, Technorati Problems, LINQ, Entity Framework

Technorati Ranks Mike Ormond’s Blog #8 with Authority of 11,764

I consider myself lucky to maintain a Technorati Authority of 76 to 80 for the OakLeaf Systems blog. (It once hit a high of 130 when I posted a controversial topic.) 

So I was surprised when I accidentally clicked on Technorati’s synopsis of a recent post about the Entity Framework by Microsoft UK’s Mike Ormond, as shown here:

 

I didn’t notice the Authority rating of either entry for Mike’s post in the preceding blurbs, but I did when I saw Mike’s rated #8 out of the approximately 2,000,000 blogs Technorati spiders, as shown here:

To give you an idea of some technology blogs competing in this authority range, TechCrunch is #2 with an authority of 22,152 and Gizmodo is #3 with an authority of 21,123. Scott Guthrie (ScottGu) ranks 739 with an authority of 1,855 (but his posting frequency has dropped considerably in the last month or so).

That’s not to say Mike’s blog isn’t great; it’s on my Feeds list to check every four hours.

However, Mike says its real Authority is 39 83 (see comment). The point is that rating a group blogsite (e.g., http://blogs.msdn.com) by the sum of its members authorities is nuts.

Issue: The preceding links appear to appear and disappear randomly and for no apparent reason. Several of my recent posts with Entity Framework tags still don’t appear in the list and the Entity Framework post frequency today is obviously incorrect:

I posted at least 10 items with an Entity+Framework tag during the period 8/20/2008 and 9/10/2008. It appears that Technorati was encountering spidering problems before their initial outage in August.

Note: My A Suggestion for Technorati's Blog Authority Ranking Algorithm post of 5/8/2007 suggests an improvement for Technorati’s ranking algorithm. However, it would make better sense for Technorati to concentrate on consistently listing new posts by tag and eliminating aggregate ranking of group blogsites before considering a more accurate ranking algorithm.

Technorati tags: Technorati, Technorati Troubles, Technorati Problems, Entity Framework, LINQ

Technorati’s Problem with IE7’s Feed Reader

I have the IE7 and IE8 B2 feed reader set to capture updates to tags of interest to this blog’s readers, such as LINQ, Entity Framework, ADO.NET Data Services, ASP.NET Dynamic Data, etc. The Feeds feature works as expected for all sites except Technorati.

Update 9/13/2008 8:40 AM PDT: Technorati tag feeds no longer validate. See end of post.

The Technorati folks sent me the following tweet:

@rogerjenn: feeds validate. What's the issue? http://tinyurl.com/4m76ag about 16 hours ago from web in reply to rogerjenn

Here’s the story: On or about 8/28/2008, I ceased receiving updates for these tags. When attempting to open the LINQ tag page, for example, I receive the following error message from IE7 under Vista SP1:

Here’s the less informative error message from IE8 B2 on a different machine under Vista SP1:

I agree that the Technorati feed validates, but I assume there are many folks using IE7’s Feed feature who might want it to work with Technorati tags. It probably would be a better customer service policy to test the the problem rather than respond with a curt “What’s the issue?” question. I’ve explained the issue in two previous messages to the Technorati tech service (?)group.

Update 9/13/2008 8:40 AM PDT: Technorati tag feeds no longer validate because they return 0 bytes:

Click here to run your own test with Technorati’s test URL: http://feedvalidator.org/check.cgi?url=http%3A%2F%2Ffeeds.technorati.com%2Fsearch%2Fpolitics

Technorati tags: Technorati, Technorati Troubles, Technorati Problems, LINQ

Technorati Contracts Twitteritis

The Technorati folks made the following announcement on Twitter last night:

Technorati web site went down around 9:40pm Pacific time. We're working on it.

When I attempted to open the LINQ tag page this morning, I received this message:

Does Technorati have it in for me?

Update 9/13/2008: At about 9:00 PM PDT on 9/12/2008, Twitter posted the following tweet:

It's been a busy last 16 hours and we're not quite there yet. Getting closer to fully restoring the web site. Thanks for your patience. about 17 hours ago from web

Followed by this announcement at 10:38 PDT the same night:

Technorati is back. A major bout with bots. This round, Technorati. Grueling and thanks to the team.

However, tag feeds still aren’t working.

Update 9/15/2008: Mashable’s Rob Diana posted Technorati: How the Mighty Have Fallen, which remarks about how little ruckus was raised when Technorati was down for over 12 hours. However, their search didn’t pick up my 15 Twitter posts with Technorati problems as the subject. Click here for all five posts about Technorati troubles.

Tag feeds still aren’t working.

Technorati tags: Technorati, Technorati Troubles, Technorati Problems, LINQ

More Technorati Troubles

Technorati is having more problems with blog indexing and customer service than I reported in my Technorati Fails to Index OakLeaf Systems Blog Posts post of 9/2/2008. Here’s the current status:

Update 9/11/2008 2:30 PM PDT. Apparently the above post caught Technorati’s attention. The latest posts are now appearing and the Authority was updated (-2 points.)

However, there’s a 20-day gap between today’s two posts and the third on the Blogs / OakLeaf Systems page and RSS feed updates still aren’t appearing in IE 7 or IE 8 beta 2’s Feeds window.

• Technorati is finally recognizing manual OakLeaf Systems pings after a lapse of more than 20 days.

• Technorati still isn’t indexing all OakLeaf posts on a timely basis and has acknowledged a serious indexing problem as of September 9. Their Blogs / OakLeaf Systems page reports the last entry was Upgrading Databound Projects to Entity Framework V1 RTM Doesn’t Expose DataSource’s 1:Many Associations of 8/14/2008, three days before the first report of a search result updating problem noted below.

• The last OakLeaf item in the Entity Framework tag list as of today was the preceding post.

• The last OakLeaf item in the LINQ tag list was LINQ and Entity Framework Posts for 9/2/2008+. Why it doesn’t appear on the Blogs / OakLeaf Systems page and in the LINQ tag list is a mystery. The missing LINQ and Entity Framework Posts for 9/8/2008+ undoubtedly is due to the reported indexing problems.

• RSS feeds for all Technorati tags that I follow now fail to update in IE 7 and IE 8 Beta 2 due to an error with a malformed document (an extra character preceding the XML declaration.) IE also refuses to add a new tag feed. (This problem started about 9/8/2008.)

• support@technorati.com did not acknowledge my initial and follow-up messages about these problems.

• The reply to my last two messages of yesterday to support@technorati.com was: “RT could not load a valid user, and RT's configuration does not allow for the creation of a new user for your email.” (I.e., Technorati no longer provides technical service.)

Recent Indexing Problems: September 2008

Indexing operations took another hit, especially for Word Press and Blogger users, as reported by Janice Myint in her Delays in Indexing post of September 8:

Late last week part of our indexing system underwent a brief outage. During that time a number of blogs experienced a halt in indexing. Since then, fixes have been implemented and indexing is occurring again. However, because of the outage, a backlog has built up in our spider queues. The spider queues keep track of pings as they come in, so no pings were lost, we just have to process them.

We are catching up as quickly as we can, but in the meantime, please expect delays in indexing, particularly for those on the Blogspot and Wordpress blog platforms. We appreciate your patience as we catch up and thank you for your understanding. [Emphasis added.]

Initial Indexing Problems: August 2008

Their indexing operations began to fail in mid-August, as reported in Dorion Carroll’s Search updates not showing in results post of 8/17/2008, which when I started having serious problems with Technorati’s tracking of OakLeaf articles. Dorian said:

Technorati is experiencing a problem with our search result updating infrastructure. We continue to crawl and save data, however, post search results are stale and temporarily stuck at about 3pm Pacific Fri. Aug 15. Link results (reactions) are stuck at Thu. Aug 14.

We have identified the root cause and are actively working on the issue. We expect to have the system caught up during the evening hours.

No data is being lost, but the most recent posts and reactions are not reflected in results at the moment.

UPDATE:

We have restored our post and tag search results. Link (reactions) results are catching up. We expect the system to be fully restored late this evening.

Fiddling with Acquisitions While Indexing Burns

Richard Jalinchandra said in his Technorati Acquires Blogcritics and Welcomes the Blogcritics Community post of 8/26/2008, in the middle of this mess:

Why did we [acquire Blogcritics]? It just made sense – as we’ve stated more times than you probably care to hear, our mission is to help bloggers and the people who read blogs. Blogcritics shares this mission, executed in their own unique way by providing a large stage for bloggers to express themselves while giving readers a great array of high quality blog content. [Emphasis added.]

Technorati currently rates a FAIL on its “mission,” and needs to put its house in order before adding more advertising real estate. As I mentioned in my earlier post on this topic:

IMO, Technorati needs to focus more energy and resource on it’s core competency as a technical blog search facilitator and ranking service. Its first priority must be to fix its indexing service.

Technorati’s second priority must be to improve the responsiveness of its technical support.

Technorati tags: Technorati, Technorati Troubles, Technorati Problems, LINQ, Entity Framework

Technorati Fails to Index OakLeaf Systems Blog Posts

Commencing about two years ago, Technorati began periodically ceasing to index this blog, even after manually pinging http://oakleafblog.blogspot.com. Initially, indexing would stop about every six months but would be reinstated immediately after sending a message to tech service.

Recently, the indexing drops increased to about two-week intervals. Again, messages to tech service restored indexing. However, Technorati’s last automated mail acknowledgment included the following paragraphs:

Thank you so much for taking the time to drop us a line. Please check our Support FAQs located at http://support.technorati.com/ which may answer some of your questions. Please also review our Blog Quality Guidelines at http://support.technorati.com/guidelines/ which we've compiled to help you understand what criteria we use for indexing blogs.

Note that in the past, simply pinging Technorati initiated an indexing of a site. This open door policy meant a lot of splogs and non-blogs would enter the system and masquerade as blogs. In order to better filter out spam as well as inadvertent pings to Technorati, we have implemented review measures to evaluate a site prior to indexing it as a blog.

It occurred to me that Technorati might be using an automated “review measure” to determine whether the OakLeaf blog conforms to their requirements, to wit:

• Publish feeds that are discoverable by using feed discovery
• Publish full content (not summaries or excerpts) in an Atom 1.0 feed
• Sign in and ping Technorati directly with your claimed blog URL
• Write posts with original content. One of the biggest differences between blogs and websites is that blogging is about being part of the online conversation. We’d like to hear your opinions and comments in your blog posts.
• Write often. The more you write, the more you are a part of the conversation.
• Cite your sources. Provide attribution by quoting and linking to the source.

It appears to me that all OakLeaf blog posts meet the preceding requirements. Further, the blog’s Technorati Authority rating is 76, and varies between about 72 and 79. An authority rating of 76 gives a rank of 85,203 in a universe of more than 2,000,000 blogs followed by Technorati. The site includes a number of Technorati widgets.

Note: The Blogs / OakLeaf Systems page shows my most recent post to be LINQ and Entity Framework Posts for 8/4/2008+, while the Entity Framework tag topic show my Comparison of Code and T-SQL Queries for LINQ to SQL and Entity Framework Databound Forms of 8/4/2008 as the most recent entry.)

Technorati is attempting to expand it’s reach by becoming a content producer (Technorati Acquires BlogCritics, Gets Into Content Game), launching an advertising network (Technorati Launches Blog Ad Network, Technorati Media) and adding a business focus (Technorati Confirms New Financing; New Business Focus Coming).

IMO, Technorati needs to focus more energy and resource on it’s core competency as a technical blog search facilitator and ranking service. Its first priority must be to fix its indexing service.

Update 9/5/2008: It’s been 48 hours since Technorati sent it’s automated message that contained the following paragraph:

If you are reporting a problem you may be having you will be contacted by a support technician once we have had a chance to review your message. Note: We've been experiencing a backlog in support and are working hard to address everyone. [Emphasis added.]

Technorati’s second priority must be to improve the responsiveness of its technical support.

 

Technorati tags: Technorati, Entity Framework, LINQ