RSS

WikiTraccs 1.34.49 (Prerelease)

By Heinrich Ulbricht |
Tags:

(Pre)release Evolution

v1.34.1

  • new: enabling certificate-based authentication with application permissions via the new Cert authentication option for SharePoint
  • improved: external image endpoint health detector covers more degraded states
  • improved: confluence base addresses without scheme now automatically assume https
  • fix: Selenium proxy mode now properly applies timeouts when downloading external images, instead of trying forever
  • fix: no progress log files were written when stopping a run via Ctrl+C or Stop button

v1.34.1 → v1.34.24 (May, 19)

This is a bigger one, covering many of in-the-wild encountered challenges, as well as some new features.

  • new: “intelligent” selection of browser for Confluence login, table size optimization etc.
    • previously, Chrome was hard-coded (even if not installed) and you had to switch to Edge manually via settings
    • now WikiTraccs detects available browsers and takes that into account when auto-choosing (you can still override in settings)
  • new: Watch Tower (former “Status”)
    • insights into your settings, most importantly non-default settings
    • opinionated color-coding for non-defaults: good (green), neutral (white), degrading (yellow); also highlights settings set via config file, which could easily be forgotten
    • image
    • note: not all settings are present; again, opinionated selection of settings that made a difference in the past
  • new: attachment macro transformation is now template-based
    • the rendered attachment list (filename, size, date, link) can be customized via a Handlebars template, like many other macros
    • also, empty attachment macros are now removed by default (you can override that via the template)
  • new: first iteration of Confluence whiteboard export to draw.io (Confluence Cloud)
    • whiteboards can be exported as part of the diagnostic report for inspection (more details here; also some disclaimers as it’s a gray setting)
    • you can opt in to this export as part of the regular migration run via the “Migrate whiteboards to draw.io file” setting
    • best-effort; not all shapes are supported, but the overall results already look pretty decent
  • new: first iteration of the web part drift refiner
    • log mode is available
    • fix mode is not yet ready due to issues with pages-related SharePoint/Graph APIs
  • new: progress bar for attachment downloads on pages with many attachments
  • improved: making the “Make JSESSIONID cookie optional” setting a default for Confluence Cloud
  • improved: when providing a non-base address for Confluence (like a space or page link), WikiTraccs now shows a warning
  • improved: cache duration extended for large attachment lists, reducing redundant Confluence calls during attachment-heavy migrations
  • improved: adding more data to the diagnostic export for SharePoint pages
  • improved: page refiner shutdown; log now looks cleaner
  • improved: handling of duplicate page IDs in the SharePoint target site; a red warning will still be shown, but the migration run now continues for the affected selector; previously it did not migrate the whole selector
  • fix: SVG images without an intrinsic size could end up as 0×0 in SharePoint; they now get sensible dimensions
  • fix: log spam in the GUI log file triggered by config evaluation
  • fix: log spam when pages referenced (via attachment link) other pages with many attachments
  • fix: white space removal for macro templates removed too much from transplanted macro bodies; this has been fixed

v1.34.24 → v1.34.32

  • new: a tool to prefill the Confluence User and Group mapping list with members of named groups; also to update the email addresses of existing user accounts in that list [Confluence on-prem only]
    • this can be helpful after switching on email visibility for a Confluence site
  • new: Outgoing References inventory report (per space), listing outgoing cross-page and cross-space links [inventory mode]
    • this includes draw.io macros linking to cross-page diagram files
  • improved: pages with very large (e.g. 100K cells) or very many (e.g. 1K) tables are now explicitly flagged, marked with placeholder text (as the content would be too large for SharePoint) and the original Confluence page is attached as a native PDF and Word export
  • improved: health check for Confluence-external domains
  • fix: in inventory mode, the Discover Spaces button was only enabled if SharePoint-specific settings were configured (which are not required for inventory)
  • fix: removed the extra blank space that could appear between adjacent nested bullet sub-lists on migrated pages

v1.34.32 → v1.34.37 (May 28)

  • new: optional native Word export for migrated pages [Confluence on-prem only]
  • improved: “Update User List” tool [Confluence on-prem only]
    • you can now set a fallback SharePoint/Entra account for Confluence users that have no email address in Confluence
    • clearer progress feedback and wording
    • added documentation: Syncing and updating the user mapping list
  • fix: images inside an inline wrapper (e.g. <span>, <em>, <strong>) in list items could be invisible on the SharePoint page in view mode, only appearing after editing and republishing the page
  • note: this release bumps the local inventory database version; local data from previous inventory runs is discarded, you have to repeat the inventory runs

v1.34.37 → v1.34.40 (June 03)

  • new: multimedia macro transformation is now template-based (allowing you to remove its text note)
  • new: create-page macro transformation, based on a template; a redlink will be created
  • improved: adding character accounting to macro transformation templates
    • so far, WikiTraccs assumed that templates did “the right thing”, but errors in templates could make content silently disappear; this is why the built-in transformation templates now broadly opt in to character accounting, which makes sure that any deviations are recorded in the TextTransferredPercent metric; it’s recommended to adopt that for your custom templates as well
  • improved: Jira issue count macro resolution has been tweaked to (hopefully) run into less timeouts for heavy queries
  • improved: Scroll PDF and Word page-history exports now run in parallel, saving some time
  • fix: View File macros (View PDF / Word / Excel / PowerPoint) where the file lives on a different page, while using a specific way of linking, now resolve to the right content instead of becoming a “not found” placeholder
  • fix: data table no longer being skipped for Stiltsoft charts when the chart image could not be rendered due to cold cache
  • fix: two character-count deviations affecting the TextTransferredPercent metric
  • fix: a hard link to a page that is missing in Confluence Cloud could prevent the whole page from being migrated

v1.34.40 → v1.34.41 (June 8)

  • improved: recorded Transformation Time of pages is now closer to the actual page creation time; this is mainly done to aid duplicate detection in WikiPakk
  • fix: character count issues (title of different charts were double counted, leading to TextTransferredPercent metric being below 100)
  • fix: raw text in Aura macro body could be missing in some edge cases (caught by TextTransferredPercent metric)

v1.34.41 → v1.34.45 (June 10)

IMPORTANT: The v1.34.45 release introduced a bug in the link fixer when it fixes inline images. If a page contains multiple to-be-fixed inline images (pointing to the wrong target), those are all re-linked to only one of the target images. This will be fixed in the next iteration.

  • new: the link fixer now also repairs links in standalone Image web parts
    • both the displayed image’s source and the image’s click-target link are corrected
  • new: “Allow web driver version drift” setting (Settings → Misc)
    • lets WikiTraccs run an older browser driver against a newer browser version, skipping the strict version match
    • useful in a locked-down environment where the browser is allowed to update, but the browser driver must be configured to stay on a fixed version due to disallowed auto-updates (via the ChromeDriverVersionOverride and EdgeDriverVersionOverride config file settings); see Confluence Authentication Recommendation for an overview of authentication options
  • improved: handling of large users bases, both when starting a migration run and when using the “Sync & Update User List” tool; this makes a huge difference when going beyond a 5K Confluence user count in the user mapping list
  • improved: Confluence API rate limiting accuracy; it is now strictly enforced
  • improved: cleaner logging around the user email lookup for Confluence user accounts
  • improved: cleaner logging of “Update Created by & Modified by” run progress
  • fix: the link fixer now correctly handles inline images; previously a page-safety check misfired, preventing the update

v1.34.45 → v1.34.49 (June 11)

  • fix: when updating the Author and Editor of already-migrated pages, the update could silently fail and the page was still marked as done
    • this happened when the Entra ID app registration only had Manage permission instead of Full Control: SharePoint accepted the write but did not actually apply it
    • WikiTraccs now reads the Author and Editor back after writing and verifies the change landed
    • if it did not land, the page keeps its “needs principal update” flag and is retried instead of being treated as finished
  • fix: inline image link fixing could stamp one image’s link onto all inline images of the same web part when a web part contained several inline images; each image is now fixed individually
  • improved: in author/editor-only update runs, WikiTraccs now skips the full Confluence content scan and migration-state report that are not needed there, which speeds up the run
  • improved: the link-fixer report now lists only links that are broken or actually change; healthy images that were merely verified are no longer added as noise