Sitebeam 5.0.0

This is a major upgrade and was made publicly available on 31st August 2011.

Big long technical list of changes

  • Added support for Presentations. Most tests now create slides as well as reports, and can be viewed in a browser and exported to PowerPoint.
  • Added core support for language detection. All tests now work in all languages except for Spelling (which works in many more languages, but not all of them) and Readability (which is English only).
  • Tests which are clearly of no value are now automatically hidden from the navigation (they still appear in the list of overall scores, so you can access them if you wish). For example, if there are no Broken Links, the Broken Links test is automatically hidden.
  • Adding websites now has a completely new interface designed to capture more information up-front, with the minimum of work from the user. It is no longer possible to add many websites at once, as more information about each website is now required in advance.
  • Running reports now has a completely new interface, and there are now 12 default report types.
  • Added ability to select and delete multiple sites, reports and dashboards in one go.
  • Added smarter percentage rounding. Decimal places are now avoided unless they are useful, e.g. 24.7% is now 25%, but 99.9% remains 99.9%, as to round to 100% would be misrepresentative.
  • The Readability test now intentionally excludes pages which look like legal (e.g. terms and conditions) from the example extracts provided, wherever possible. This detection is based on analysis of each page’s title tag, and whilst imperfect will exclude most common legal pages, resulting in better examples of poor readability.
  • Readability test now uses SMOG testing, instead of Fletch Kincaid.
  • For Readability analysis, numbers are now converted into the most appropriate internal verbalisation for analysis. E.g. 101 = “one hundred and one”, while 01332 224800 is oh one three three two, two two four, eight zero zero. This means some numbers are far easier to read, like “fifty”, whilst slight variations are harder (forty-nine).
  • For Readability, page footers are automatically detected and excluded wherever possible.
  • Noscript blocks are excluded from readability testing.
  • Individual ‘hard to understand’ extracts and reading ages are shown for each page, where possible.
  • <footer> and DIVs that look like footers are now excluded from body text. This fixes legal text found in most sites, particularly for readability.
  • Added support for HTML5 tags, implicit tag closing to HTML parser.
  • Added better plain text extraction based on more intelligent parsing of obscure tags, such as <dt>, implicitly closed <td>s (in HTML5).
  • Added vastly superior sentence splitting, to handle nested punctuation (e.g Simon said “hello.”), variable Unicode punctuation marks, non-delimiting periods and more.
  • Completely rewrote sentence and word splitting for multibyte support, superior intelligence.
  • Mass migration of ereg to preg throughout (faster, required for PHP 6 support).
  • Fixed major flaws in non-UTF-8 parsing.
  • Added ability for tests to ‘require’ other tests to run, e.g. Spelling requires the Language test.
  • Optimised to make remote DB transfers faster.
  • Links no longer count as having implicit whitespace around them for the plaintext analysis. This will reduce most problems but create others, depending on the site.
  • Spell checker now checks hyphenated words as a whole (e.g. “Wi-fi”) and as separate components.
  • Added following Bing locales: Argentina, Australia, Austria, Belgium (Dutch), Belgium (French), Brazil, Chile, Denmark, Finland, Hong Kong, India, Ireland, Japan, Korea, Latin America, Mexico, New Zealand, Norway, China, Portugal, Philippines, Russia, Singapore, Spain, Sweden, Switzerland (Italian), Switzerland (French), Switzerland (German), Taiwan, US (Spanish)
  • Reduced simultaneous HTTP pool downloads to 4 (was 5).
  • Made Javascript parser understand escaped characters in strings, e.g. “/” is “/”. Particularly important for Flash test.
  • Flash parser now filters out more stupid URLs, e.g. “/.swf”
  • Added support for CrazyEgg analytics / to cookie test.
  • Search Engine Placement is now hidden if no keywords are specified.
  • We now use a copy of the latest official W3C validator for all validation, particularly HTML 5.
  • Added tabs to appropriate tests.
  • Removed superfluous detail from SERP test.
  • Added support for comparison URLs
  • Added comparison URLs to Popularity test
  • Incoming links is now scored by MozRank (was Google, Yahoo).
  • Incoming links now shows comparison sites
  • Incoming links now shows sites linking as well as total links
  • Reduce severity of Redirections test
  • Added new Languages test.
  • Added new Facebook test.
  • Added new Social interest test.
  • Totally rewrote Twitter test, splitting partly into new Social interest test.
  • Added vastly more competitive analysis to Search Ranking test.
  • SE ranking checks top 30 results only.
  • SE ranking scores results more harshly.
  • Rewrote all SE code to be faster, more stable, handle MB better (preg fixes, fewer calls).
  • SE test now shows links to all SE results regardless of whether matches were found.
  • Reduced various Marketing Score penalties.
  • Fixed bug where multiline dates wouldn’t detect in Freshness test.
  • Amount of content now scored on logarithmic scale, entirely from word count (i.e. pages without content are now irrelevant).
  • Changed Freshness test to have much more lenient scoring with moderately up-to-date sites.
  • Fixed bug where Yahoo/Bing report not found when can’t return any results.
  • Renamed Cookie Law test ePrivacy Law test.
  • Added support for HTML5 detection to ePrivacy Law test.
  • Made ePrivacy test make more intelligent guesses for unknown cookies.
  • Added recommendations to ePrivacy Law test.
  • Removed Whois from Domain name test.
  • Renamed Domain name test Typosquatting test.
  • Added smarter Typosquatting calculation (now guesses phonetic typos, such as “example.com” -> “xample.com”, “youtube”->”yuutube” etc)
  • Substantially reduced weight of redirections test (much less relevant with all modern browsers taking care of this automatically).
  • Content score now includes Social interest as a measure of quality.
  • No longer heavily penalise sites for only having 1-3 pages (previous assumption was these were broken, we now assume they’re intentionally small).
  • Now grade Readability on a logarithmic scale.
  • Completely re-wrote Twitter test. Now faster and focused on competitive assessment.
  • Completely re-wrote Content Keywords test. Now fully supports all languages.
  • Completely re-wrote Spelling test. Now indicates location of spelling issues (e.g. in meta, title, alt text), handles multiple languages.
  • Spelling no longer affects the Overall summary score.
  • Key points are now split into 2 columns, positive and negative points.
  • Added support for iPhone and iPad access (in browser).
  • Added many sections to Help to explain common problems.
  • Added new diagnostic feature under Account: Test URL.
  • Fixed bug where database transfers wouldn’t show their progress correctly in the progress meter.
  • Fixed bug with amount of content test not realising a site was too small if fewer pages were *tested* and fewer pages were found than 20.
  • Fixed issues in parsing obscurely formatted META refresh headers.
  • Fixed bug where broken links test would not test non-HTML links (e.g. links to PDFs).
  • Fixed many IE specific issues, including presentation of forms, and allowing users to press Enter to submit their login form.

Was this article helpful? Contact our support team if you have a question.