• Import to SGML is not directly supported by Arbortext Import/Export. To import SGML files, create an XML version of the target document type and use an ACL postimport hook to save as SGML

• Arbortext Import/Export cannot import graphics as entity references.

• Processing instructions cannot be directly imported.

• MIF embedded graphics created using the FrameMaker drawing tools are not imported.

• In rare circumstances, certain embedded images in MIF files cannot be converted from their original format, instead appearing in the imported file as embedded Microsoft Word documents. The embedded images are identified in the FrameMaker MIF files as "OLE2" images, which is the same OLE compound document format as Microsoft Word files. A possible source for this scenario may occur when the images were originally created in Microsoft Word and then pasted into the FrameMaker document. These embedded images must be recreated in a different format before conversion or be manually recreated after the conversion.

• Imported documents are not guaranteed to be contextually valid documents.

• Arbortext Import has no concept of an XML template, which uses Arbortext Editor's caret location as the starting point. This means any boilerplate data must be encapsulated in one or more MapObjects.

• Viewing the HTML versions of the source document (.doc, .rtf, and .mif) requires Internet Explorer 6.0. Without Internet Explorer 6.0, Arbortext Import and the Arbortext Import Workbench will still operate properly, but the HTML rendering of the source document will not function.

• Arbortext Import imports RTF versions 1.7 and 1.8 (although specific features of any version may not be mappable or importable). The hundreds of third-party applications that produce RTF may create RTF files with unexpected content. Arbortext Import may not be able to import all of these variations of RTF.

• Microsoft Word text boxes are not fully supported and may import in unexpected manners due to the nature of their anchor point/XY location relations.

◦ Top-down order may vary from the apparent document order as anchor points are often hidden and unknown.

◦ Layers of text boxes can obscure each other and can produce unexpected results.

◦ Certain fonts may be processed incorrectly.

◦ Freeform mixtures of text and images within text boxes, shapes, and groups may result in lost data or out-of-order data.

• Arbortext Import can only process system files. Ensure that permissions are set properly if a database repository is involved in your import process.

• Arbortext Import does not support importing Hebrew, Arabic, and Thai content.