We are glad to announce the immediate availability of Help+Manual 9.5 and an updated Translation Assistant.
Help+Manual 9.5 comes with OCR
The most obvious new feature in version 9.5 is optical character recognition (OCR). If you ask yourself What’s the benefit of OCR? then we are in the same boat. We, too, have been reluctant to implement OCR, but a couple of user arguments eventually convinced us. There are two scenarios where OCR comes in handy: (a) you are documenting a software and need to capture text and labels directly from the screen. (b) You have inserted a screenshot into a topic and need to describe it in detail, repeating the labels and text from the picture. Not everyone has experience in touch-typing and that’s where OCR is helpful.
We have implemented OCR into the screen capture tool (to capture text anywhere from the screen) and added it to the image tools in the editor context menu (to extract text from an image already inserted).
The OCR function uses the Windows 10/11 built-in OCR engine. Though this OCR engine is not the absolutely best one out there, it saves us from distributing hundreds of megabytes of OCRÂ recognition data. The setup of H&M is still about the same size as before. The only downside: it does not work on earlier versions of Windows.
OCR in Help+Manual, Screen Capture Tool
OCR in Help+Manual, Picture Context Menu
eWriter output: Preview with MS Edge
You may have read about the eWriter Version 5 Beta. Help+Manual 9.5 uses this version for preview. When you create an .ewriter file, Help+Manual uses its internal version to display it.
Depending on the system it runs on, this eWriter viewer uses Microsoft Edge as the display engine. Note that “Microsoft Edge installed” is not the same as “Microsoft Edge Runtime installed”. On Windows 11 computers, the runtime is already pre-installed, many Windows 10 systems are catching up.
Miscellaneous Changes and Bug Fixes
- New Project: support for FastHelp import implemented
- HTML Export: ARIA attributes added to images with captions: role=”figure” aria-labelledby=”labelid”. To enable the export of the “labelledby” attribute, the image must have a dedicated ID.
- ePUB Export: in ePUB, .SNIPX images are no longer converted to SVG, because several ePUB viewers cannot display SVGs correctly. ePUB export now uses PNG instead.
- PDF/Print: when printing with 2 or more columns, the temporary switch to a landscape page caused the landscape print width to be wrong with a right margin that was too large.
- Keyboard shortcut CTRL+W implemented to close the current tab
- Drag & drop of web browser links into topic editor created a link with target “https” only, the actual web address was not resolved.
- Paste from SnipSVG: when pasting images from SnipSVG with numbering tags, the numbering table used the wrong symbols for tags > 10.
Help+Manual Translation Assistant 9.5
- No functional changes
- Application manifest enables long Windows 10 path names
All updates are available from our download page.