This happens only in specific scenarios as most of the documents with annotations are visualized correctly.

Images decoded with CCITTFaxDecode filter affect the performance when showing the document.

Most PDF viewers have option (enabled by default) which allows fields to be highlighted with colors predefined in the settings. Required fields are highlighted using different color.

For example: In Adobe Acrobat Reader DC this is enabled in "Edit" -> "Preferences" -> "Forms" -> "Highlight Color" -> "Show border hover color for fields". 

Expose a button in the UI that can be used instead of Ctrl+F.

Such a button can be currently added by customizing the default toolbar. For more information, check the documentation that describes how to achieve that: https://docs.telerik.com/devtools/wpf/controls/radpdfviewer/default-ui#customizing-the-default-ui

At the moment the theme of the RadPdfViewerToolBar can be set only by setting the theme of the entire application. The individual setting (using telerik:StyleManager.Theme="Windows8") should also be supported.

When a PDF is printed all images are omitted.

Available in R1 2018 SP1 version.

This exception seems to be related to AES algorithm PaddingMode property. 

This exception seems to be caused because the stream is not decrypted correctly and the stream data is corrupted. For instance when the stream is with FlateDecode filter the exception is InvalidDataException with message "Unknown compression method".

This exception is thrown because there is field name with special characters which are incorrectly modified during the recalculation process of widget appearances. This exception is thrown in the internal RecalculateMissingAppearances method or in RecalculateWidgetAppearancesOld method.

Setting Handle property to false should result in rethrowing the exception and interrupting current RadFixedDocument rendering. This is related to Exception Handling Mechanism mentioned in this documentation article: 

https://docs.telerik.com/devtools/wpf/controls/radpdfviewer/document-model/exceptionhandling

Loading page with a lot of text takes several minutes which is unacceptable. This performance issue is caused by class GlypInfoFactory, which caches the glyph information.

Available in LIB Version 2017.3.1225.

For some fonts the issue is present. Some of them are:

KozGoPro-Medium

KozMinPro-Regular

KozGoPro-Bold

Enable the customers to directly save the document inside PdfViewer to a file or a stream.

Available in LIB Version 2017.3.1225.

Layers represent groups of optional content which can be hidden or shown.

The Ò symbol is not found in the encoding as is rendered as a ".notdef" glyph.

Available in LIB Version 2017.3.1127.

This may result in inversed transparency issue when rendering images with ImageMask.

Fix available in LIB Version 2017.3.1120.

Currently, the implementation relies on the fact that the previous cross-reference value of a stream object is always a stream object. Instead during the cross-reference recursion, we should check on each step whether the previous cross-reference item is table or stream.

Fix available in LIB Version 2017.3.1120.

Before starting to decode big image, try to estimate the needed memory and check if such memory is available using the MemoryFailPoint class. The behavior in case of failure in printing scenarios should be considered carefully.
The user should be able to turn the behavior on and off.

Currently, the Find() method searches from the beginning of the document to its end. Enable the customer to specify a particular range of pages the method should search through.

When a page is rotated its annotations are not positioned correctly.