What's new in this version:
- some false results in the spellcheck, which were due to concatenation of words with certain settings and when certain html tags run together
- selected spellcheck language not being 'remembered' sometimes
- potential horizontal scroll issue in SEO>Summary
- More clearly marks image tags with empty source. Previously 'flag missing link url' would also pick up img tags with missing src, if image check was switched on, but would confusingly mark them with the status "missing link url". Now the images are given the status "missing image src" so that they can be distinguished from missing link urls in the By Status and other views.
- Adds Expand All and Collapse All to View menu with keyboard shortcuts, and buttons for those functions to the toolbar palette. These will expand / collapse all items in the current view, if the view is expandable, eg by page, by status, Spelling by word.
- Fixes bug that may have prevented pdfs from correctly being included in the sitemap if the preference was checked or caused them to be included even if the preference was unchecked. Now this behaves as expected according to the preference.
- When scanning locally (file://) and with the 'test anchors' setting switched on, could go into a loop
- Fixes 'list' view of sitemap visualisation not appearing correctly when system in light mode
- Adds checkmark to the 'sort by' popup button above list of websites
- Fixes the issues with sorting some of the columns of the 'All links' table
- Small fix for a problem that will affect very few users, but will be a very important fix for those users. If the starting url has a path, and that path contains encoded characters (eg %20) and relative links on the page then the crawl may not proceed past the first page.
- Some improvements around the collection and counting of image urls
- Fix - if 'alt' attribute wasn't included at all within an img tag, then the image wasn't always being counted as an 'image without alt text'
- Adds 'redirect count' as an optional column to the All links table. Sorting this table will find links with the longest redirect chains. This appeared with the SEO results in version 10 and earlier. However, redirect count isn't a property of pages but link urls (different link urls may arrive at the same page via different numbers of redirects.)
- At the end of a scan, if user hasn't added a name for the configuration, "A new website" is automatically replaced with the first word or two from the starting url's meta title
- Fixes a problem that could cause repetition or even a loop with certain urls that contain percent-encoded characters
- Fixes toolbar export button which, after a search, was exporting the link urls rather than the search results
- If searching, pointless listing and checking image urls, so that functionality is forced off, even if 'check images' or 'images in sitemap' are selected. (those options will be observed if a regular scan is then run.)
- Adds double-click to SEO / Images tab for preview
- Also context menu with Preview / Copy URL / Open URL
- Fixes a problem with the "entire source / body text" switch within the Site Search dialog.
- If search is used, subsequent scans (without quitting and restarting) would incorrectly show some search results at the end of the scan. Now fixed
- Fixes an issue causing some missing anchors to not be reported
- When testing anchors, if the anchor fragment of a url isn't found on the target page, link is now coloured red in the views and treated as a bad link, rather than being coloured orange as before.
- Reinstates 'Export Image XML Sitemap' which was a feature of v10 but has been unintentionally missing from v12 to date
- Tidies up the image checking functionality internally. Images must be included in the scan if images are to be included in the sitemap xml or an image sitemap, so image checking is forced on if either of these settings are enabled. To disable image checking (for efficiency with certain operations) then all of these settings need to be switched off
- Fixes Description Length column of SEO table showing incorrect value
- Fixes empty lastmod tags appearing in sitemaps where no date is available. Tags are now correctly left out if no date is available (optional tag). NB Google says, if no date is available, don't insert date of sitemap generation
- Some fixes to the 'check anchors' functionality. Note that the status will be the http status regardless of whether the anchor is good or bad, ie '200 no error' even if the anchor doesn't appear on the target page. In the case of a problem with an anchor, a warning is created on that url which can be highlighted in orange depending on your Preferences, or shown in the warnings table in the case of Integrity Pro and Scrutiny.
- Fixes the option Preferences>Views>Display labels>'Display labels for redirected URLs'
- An efficiency improvement which may help with a slowing-down problem experienced by some users and will generally make the app more efficient for medium to large sites
- Fixes double-quotes in url (even if percent-encoded) corrupting bad links (by link) csv
- Fixes crash which happened under unlikely circumstances
Fixes some problems relating to exported report and autosaved data:
- where report summary wouldn't contain all of the information after reloading autosaved data
- where 'all links' csv in exported report would contain only bad linksz if the 'bad links only' checkbox had been checked in the interface.
- where count of 'images without alt text' would not always appear in the summary report
- Fixes pdf documents (if being checked) were being included in some SEO tests that should only apply to html pages, such as missing title / description.
- The default threshold for the 'too many links' SEO test has been raised to 1000. This is no longer a very relevant test but may still be useful to know if a page has an excessive number of links
- Similarly, default keyword density threshold increased to 6% for more relevant results
- Fixes a bug that could have caused spurious statuses for certain urls (caused by containing a percent symbol which isn't part of a percent-encoding, which is a bad practice anyway.)
- Fixes some urls with unlikely character combinations in the querystring or path to misleadingly display in tables as the domain only
- Some social and calendar 'add this' links are now listed but not checked. It's potentially not a good thing to request these urls to check them. They are major domains (eg Google, Outlook) and tend to be generated by plugins, and so more likely to give a false positive when tested than to actually be broken.
- Better handling of html5 picture tag
- Preferences for Sitemap (whether to include unique image urls, audio, video, pdf) are now all switched on by default
- Fixes sorting by 'last scanned' date in archived data manager
- In the 'by link url' table, the starting url may have shown 'appears 0 times' in the On page column if the exact url didn't appear as a link (eg if you start at mysite.com but all 'home' links point to /index.html). Technically true but inelegant. Now shows "starting url"
- Restores 'Check for updates' under Application's main menu. If enabled, an auto update check will be performed on startup and display a button in the main window. The auto update can be switched from within the manual update check window.
- Adds Tools > Autosave Data Manager, which lists autosaved data, the sites they relate to, size, and allows you to selectively delete
- Fixes a problem that could (under fairly unlikely circumstances) lead to a page incorrectly being excluded from the sitemap for being marked 'robots noindex'
- Fixes some issues experienced when updating from version 10
- Fixes a problem that led to the final status of certain urls (a redirect followed by an error) showing the 3xx redirect code rather than the final error code
- Fixes bug with the redirect table in the Link Inspector not showing the starting url on the first row but the first redirect url
- Improvements to parsing srcsets
Improves structured data functionality:
- Now allows for multiple chunks of json-ld on a page
- Now reads multiple items within @graph properly
- Displays the top-level @types from json-ld in the SEO->Meta data table
- Preference>Views>'treat blacklisted urls as bad links'
- appearance of headings 'outline' view in page inspector
- issue that could have caused twitter: meta tags to not show in SEO -> Meta data table
- Adds Twitter card visualiser and json-ld structured data viewer to Page Inspector window
- Other small fixes
- Adds 'Delete configuration' to context menu
- Fixes sorting by column in 'by status' view, plus small efficiency related to building the by status view at end of scan
- Adds a couple of context help items to the scheduling tab
- Fixes an issue that could have caused image urls after an audio or video tag to be corrupted and therefore test bad, Since 12.4.2
- relating to the 'on finish alert' preference
- relating to the insecure link and mixed content functionality. See settings > Finish Actions > Alert when links to http site are found, and Rules > When scanning a secure (https://) site... See the manual for proper use
- relating to the 'high keyword density' SEO test
- an issue which could sometimes cause scan to slow down and stop when using the rendering feature
- Many updates to the manual - found under the Help menu and website
- Improvement to 'robots noindex' search, false positives may have been seen for that
- fixes a problem that prevented the contents (src) of iFrames from being followed if the starting page consists of iframes and no other links
- Some minor udpates to the context help system. Corrected link under Help for the v12 manual. Manual completely revised.
- Adds "Save sitemap" to Finish Actions, with dialog for location and filename when you check the box. This was present in v10 but not carried to 12 because it seemed to duplicate the sitemap option in the full report. It turns out to be useful to be able to automatically save the sitemap at a known location and filename.
- Improvements to scheduling functionality, particularly for those just upgrading from v10
- dmg includes the Uninstaller. NB this is not to be used if you're upgrading to 12 from a previous version. It will remove all files relating to Scrutiny and is for use if you no longer need to use it on a computer, and is particularly aimed at the situation where there are schedules that try to run and fail because you've deleted the app. (These are best managed within Scrutiny using Tools > Schedules Overview )
- Adds support for links to audio and video files within but both markups are valid and now correctly parsed, tested and reported
- Adds 'Visit' to context (right-click) menu of website list
- Fixes problem causing css files to not be parsed for url('') images. (These image urls should be tested and reported if 'linked files' and 'images' are switched on in Options.)
- Adds "parse PDF files" and "parse DOCX files" to Options. This functionality has long existed in Scrutiny, but the checkboxes have been missing from the new v12 interface.
- Moves 'Search site' from 'Options' to Tools menu. a/ because it's not an Option, and b/ because of the above fix reducing available space on that tab with the window at minimum height.
- Fixes problem which caused starting urls with accented characters in the domain to stall.
Urls with special characters may have appeared encoded rather than decoded in one or two places including the Link Inspector 'appears on' table and redirect table.
Improvements to soft 404 functionality:
- certain social networking sites currently return a 'soft 404' which is a 200 code and a page that says "page not found" or similar. Detecting this requires the soft 404 feature to be switched on and configured. in some cases it can also require clientside rendering. External links to Twitter and Youtube are now automatically rendered in order to access and check page content (if soft404 is switched on)
- Some terms are added to the default value for the soft 404 terms field. This will only affect new users or those who haven't altered the default list.
- When a possible soft 404 is detected, a warning is created which explains which term was matched on the target page. This can be seen in the Link Inspector and in the Warnings table.
- Note that Instagram.com also requires authentication, otherwise a request for any url will receive the login page along with 200 status. Currently, attempting to authenticate with Instagram using Scrutiny's login window is hit and miss.
- Some pages won't show any content before the user has accepted cookies. It's possible that accessing the site using Scrutiny's Login window and accepting cookies will get around this, but otherwise this request for a user action will block Scrutiny from being able to detect a soft 404
- Adds context help to soft 404 settings in the Preferences window
- Removes a couple of redundant items from the View menu which were confusing
- improvements to handling of basic http authentication
- Improves sorting in Link Inspector 'Appears on' table
- Adds 'Copy URL' to the context menu for that table
- Fixes a problem preventing the 'bad links only' button from working properly in 'by page' and 'by status' views
- When Pause is pressed during scan, current connections are now stopped and returned to the check queue. Previously they were allowed to finish, which didn't cause a problem, but as the stats crept after pressing the button, it didn't look particularly 'clean'.
- Fixes message on Spelling tab 'Spelling is disabled in settings' being displayed permanently
- Adds 'live updating' to the 'By link URL' table (and only if that tab is selected)
- it is best if the Settings or another tab is selected during the scan, as the live updating is an overhead and makes the san slower and less efficient. To help with this problem, the 'live updating' happens periodically rather than with every url. But this still isn't advised for very large sites.
- Detects system appearance (dark/light mode) and makes the Summary either dark or light accordingly. (The white summary when system was in dark mode was jarring.) Exported version is still white.
- Significantly improved and more efficient parsing for meta http refresh. Now checks for delay in seconds within the content attribute, if small (<6s) will observe the redirect, otherwise will ignore.