Brian, you prompted me to do some experimenting, using new test messages in new test folders and with new simple annotations - so results are clear. I checked what was happening with files in the user Pmail/Mail directory each time. The results are:
- (Re your query) Moving a previously annotated message - now with the annotation absent - back to the new mail folder from a temporary folder does not re-establish links to the orphaned annotations.
- Moving a still-annotated message back to the new mail folder from a temporary folder works fine and it appears there complete with the annotation.
- Green dots, indicating the existence of annotations, work exactly as one would expect they should i.e. once a message gets an annotation the green dot exists until the message is finally deleted, including in copies and following moves between folders.
- Deleting a message with linked annotation from the new mail folder (i.e. a .CNM) or from another folder (i.e. a .PMM/.PMI) deletes any associated annotation, without leaving orphaned .PMN annotation files behind i.e. exactly as one would expect. (Note: To be clear, I did no direct editing of Pegasus files using Windows; always using Pegasus menus).
- Deleting a message that has an orphaned annotation
(i.e. a .CNM or .PMM/.PMI exists as well as a previously associated .PMN in the user's Pmail/mail folder) - leaves the associated orphaned annotation behind.
- Re-indexing a test folder after moving a message out to the new mail folder - or to any other folder - caused the message to reappear in the test folder, complete with green dot but without the original annotation. It also exists in the folder to which it has been moved, but in that case the annotation is still attached (unless re-indexing has also occurred in that folder).
- Recovering deleted space, without subsequent re-indexing, has no impact on annotations. Recovering space and then re-indexing seems to prevent a moved message reappearing, so the question of associated annotation is irrelevant.
- Whenever an annotated message is moved (without re-indexing), its long Hex .PMN filename is changed.
- The impact on annotations of moving messages between folders is the same as copying messages between folders (as one might expect).
- Searching for references to the names of annotation files throughout the user's Pmail/Mail directory (i.e. checking in file contents), brings up no relevant results. However, as a link clearly does exist originally between the message file and its annotation, which was when I checked then I think Brian's conclusion that the link is encoded somehow in the relevant files.
Overall conclusion: The potentially very useful note-taking facility of annotations is fatally flawed by (a) the fact that re-indexing breaks the link between message and annotation and (b) the absence of any facility to remake these important links and indeed the lack of information on how to identify which orphaned annotation file belongs to which message file.
As an afterthought, the reappearance of previously deleted or moved messages upon re-indexing should ideally be addressed at some stage.
<p>Brian, you prompted me to do some experimenting, using new test messages in new test folders and with new simple annotations - so results are clear. I checked what was happening with files in the user Pmail/Mail directory each time. The results are:</p><p>- (Re your query) Moving a previously annotated message - now with the annotation absent - back to the new mail folder from a temporary folder does not re-establish links to the orphaned annotations.
</p><p>-&nbsp; Moving a still-annotated message back to the new mail folder from a temporary folder works fine and it appears there complete with the annotation.</p><p>- Green dots, indicating the existence of annotations, work exactly as one would expect they should i.e. once a message gets an annotation the green dot exists until the message is finally deleted, including in copies and following moves between folders. </p><p>-&nbsp; Deleting a message with linked annotation from the new mail folder (i.e. a .CNM) or from another folder (i.e. a .PMM/.PMI) deletes any associated annotation, without leaving orphaned .PMN annotation files behind i.e. exactly as one would expect. (Note: To be clear, I did no direct editing of Pegasus files using Windows; always using Pegasus menus).</p><p>- Deleting a message that has an orphaned annotation&nbsp;
(i.e. a .CNM or .PMM/.PMI exists as well as a previously associated .PMN in the user's Pmail/mail folder) - leaves the associated orphaned annotation behind.</p><p>- Re-indexing a test folder after moving a message out to the new mail folder - or to any other folder - caused the message to reappear in the test folder, complete with green dot but without the original annotation. It also exists in the folder to which it has been moved, but in that case the annotation is still attached (unless re-indexing has also occurred in that folder).</p><p>- Recovering deleted space, without subsequent re-indexing, has no impact on annotations. Recovering space and then re-indexing seems to prevent a moved message reappearing, so the question of associated annotation is irrelevant.
</p><p>- Whenever an annotated message is moved (without re-indexing), its long Hex .PMN filename is changed.</p><p>- The impact on annotations of moving messages between folders is the same as copying messages between folders (as one might expect). </p><p>- Searching for references to the names of annotation files throughout the user's Pmail/Mail directory (i.e. checking in file contents), brings up no relevant results. However, as a link clearly does exist originally between the message file and its annotation, which was when I checked then I think Brian's conclusion that the link is encoded somehow in the relevant files.</p><p>&nbsp;</p><p>Overall conclusion: The potentially very useful note-taking facility of annotations is fatally flawed by (a) the fact that re-indexing breaks the link between message and annotation and (b) the absence of any facility to remake these important links and indeed the lack of information on how to identify which orphaned annotation file belongs to which message file. </p><p>As an afterthought, the reappearance of previously deleted or moved messages upon re-indexing should ideally be addressed at some stage.
</p>