Release Notes for version 0.9.4.24

This is a stability and bug fix release. Introduces also some UX fixes to disambiguate URLs with same path in scenarios of multi (sub)domain/protocol.
With this release, minimal required .net framework version is now 4.x (used to be 3.5). Notice that it also runs on XP SP3, so there are no minimum OS requirement changes.

User Experience / Usability:

• Suggestions grids, everywhere Path and Canonical Path were shown, they have been by default substituted with URL and Canonical URL, to better disambiguate cases of cross-protocol or cross-(sub)domain.
• Table View: added column "Canonical URL" (URL decoded); changed default visibility of columns based on users' feedback.
• Suggestion Trees: protocol shown for page URL and canonical URL.
• All over the application the page property "Authority" (name coming from the official URI RFC: host name and port number, the latter only if different than default) has been replaced with "Authority Name" which in VSS lingo is the same but prefixed with the protocol scheme.
  This permits to disambiguate cases where both HTTP and HTTPS versions are crawled. Applies to all languages.

Various:

• Setup: .NET Framework now detected earlier.
• Prerequisites change: At least .NET Framework 4.0 is now required to run the application.

Fixes:

• URL Suggestions: reports for duplicated canonicals didn't take into account when there was a difference only in the protocol two canonical paths.
• URL Suggestions: report "duplicated URL (case-insensitive) didn't take into account when there was a difference only in the protocol between two crawled URLs.
• When a page canonical URL had the same path of the crawled path, but a different (sub)domain and/or protocol, the program didn't distinguish.
  The fix solves the issues for future crawls, while for older ones the original protocol/domain information is unfortunately lost.
  Note: the fix requires an update in the project format.
• Cross protocol support, same pages accessible with both http and https versions where fetched only with the first version met, somehow impairing the cross-protocol feature in many cases.
• Custom Filters "links to domain" could have resolved incorrectly relative URLs belonging to a subdomain different the start URL.
  This in some cases - depending on the queried domain - could have produced an erroneous subset of extracted pages.
• Sitemap Creation window: the page counters in the upper-left grid reported an incorrect number of pages (it comprised the number of detected folders).
• Fixed UI regression: with 0.9.2 Page Links grid got accidentally moved 30px to the left, partially impairing its usability.
• Sitemaps window didn't properly support cross-protocol cases and didn't distinguish the two URL versions (HTTP/HTTPS)
• Crawling, UI/UX issue fixed: when only one node was added under the (sub)domain node, the latter stayed collapsed.
  Happened for example when only the home page was crawled, or when an unexpected response code in robots.txt prevented further exploration. The latter was particularly annoying as the collapsed node prevented understanding at first sight where the issue was.
• Custom Filters "links to domain" could have resolved incorrectly relative URLs belonging to a subdomain different the start URL.
  This in some cases - depending on the queried domain - could have produced an erroneous subset of extracted pages.
• Cross protocol support, same pages accessible with both http and https versions where fetched only with the first version met, somehow impairing the cross-protocol feature in many cases.
• Crawling, UI/UX issue fixed: when only one node was added under the (sub)domain node, the latter stayed collapsed.
  Happened for example when only the home page was crawled, or when an unexpected response code in robots.txt prevented further exploration. The latter was particularly annoying as the collapsed node prevented understanding at first sight where the issue was.