The rise of relatively robust AI translation in recent years has introduced a new configuration to my databases. I research global media and have for years included translation notes in the “comments” field of the original documents that I clip. Thanks to various AI options (Google, DeepL, etc.) I can now easily generate full English-language translations of the original articles more quickly than I can create my own translation notes. This boon brings a slight quandary, however, as I don’t know quite how to include these (for my use) subordinate files in the database. I have considered three options, each with their shortcomings, which I will outline below. Ultimately, however, I think it would be useful to have a special link type for subordinate documents that are always associated with another file. The properties I’m looking for are as follows: (1) each document would always be displayed in the item list when the associated other document is searched and shown in DTP3, ideally with an “original” and “sub” designation made clear by nesting hierarchy (for my purposes, this could be a one-to-one or one-to-many relationship–the latter giving the option of multiple translations, etc.); (2) search would include the text of both (or all associated) documents.
Would this use case fit the needs of others?
The other three options I’ve considered are as follows:
(1) Copying translation into the original document (my current use) - Import both the original and the translation into DTP3 and then use the thumbnail tab to copy the translation into the original at the end and, finally, delete the separate translation file, The downside of this is that it will likely mess with the way DTP3 creates and uses metrics related to word usage for the article.
(2) Making a folder for each translated document and including both the original and translation in that folder. This adds an extra level of folder navigation which I prefer not to deal with at the interface level.
(3) Creating links between the files using the various current options available in DTP3 for “Document Linking.” My main concern with these is that the subordinate and always already at hand statuses don’t seem to be in effect with these options. Links must be created and then clicked on to be effective. This complicates things at both the creation and use stages.
At first blush, it appears you are referring to an annotation file (see Tools > Inspectors > Annotations & Reminders). While it doesn’t fully fit your spec, it is a document that keeps an association with another document as a special property.
The downside of this is that it will likely mess with the way DTP3 creates and uses metrics related to word usage for the article.
Are you referring to the Concordance or relevance in search?
Thanks for your swift reply, Bluefrog!
In answer to your question about word usage metrics, I meant it in general terms. I am guessing there could be pros and cons to including the original and translation in the same document, but that this could shift significantly the way that DTP3’s AI deals with “Classify” and “See Also” functions, and that concordances and word clouds for any particular document will appear quite differently when two different languages are included in a single document.
In terms of your suggestion, I’m not sure how to make an annotation file. I’m guessing (but not certain) that this would be an “attachment”? My most common use case is to have 2x PDFs, both either imported into DTP3 as PDFs or “clipped” into DTP3 and made into PDFs in that process. Is it possible to make one of these an attachment to the other, and would that (1) allow results from both to show up when I’m running a search within the database?; (2) allow me to see the attachment in DTP3’s “Item list” in addition to the original document?; and (3) allow me to open the attached (subordinate) document as a PDF in my external PDF editor (PDF Expert) for in-depth annotation? Those are the main functions I’m trying to achieve at the moment.
Thanks again for any thoughts you can offer.
You’re welcome 
Open Tools > Inspectors > Annotations & Reminders and see the Annotations section.
This is covered in the Inspectors > Annotations & Reminders section of the built-in Help and manual.
-
No, you can’t use a PDF as an annotation of another PDF. Only Markdown and rich text are supported.
-
No, the documents won’t appear in the same search unless they actually matched the search criteria.
-
You can set Preferences > General > General > Annotations to In same group if you’d like the document’s annotation file to be created in the same group.
-
You can open the PDF or annotation file in an external application, as needed. There is no change in behavior here.
Thanks again, Bluefrog!
I can see that this is a quite useful and flexible option for various texts that can be easily copied and pasted. Indeed, the .rtfd format even accommodates tables and images relatively well, so a “select all” plus “copy” from a .pdf can end up with decent results. Though this is great, there are two things that will keep the special link type I mentioned in the original request on my wish list:
- In my workflow most annotation is done in PDF Expert for a host of reasons, and that program doesn’t accept .rtfd documents. Shifting between the two programs for annotation and general use complicates the workflow.
- Though the general setting placing annotations in the same group is excellent for keeping these items together (when I move them from the global inbox to their own database, for example) it still seems to me that sub-nesting within the item list would give a clearer visual sense of the subordinate status than the current icons in the “name” field.
I recognize that these are minor and, likely, rather specific quibbles. They may be worth considering if other users note a similar need in the future.
Once again, I’m grateful for your assistance, and for all the work you to on DT!
Such as…?
Thanks again, Bluefrog!
I’ve done a lot of testing, spending serious time with Acrobat, LiquidText, Preview, and Skim on the Mac, and Acrobat, iAnnotate, and LiquidText on the iPad, and PDF Expert seemed to be* the best compromise for trying to straddle both platforms with my PDF needs. Some of the reasons are as follows:
- I appreciate the split-screen options for viewing a single or multiple documents. This is especially useful when I’m trying to read the endnotes in an academic article or, indeed, as in the case above, reading the original and translation side-by-side. Crucially, this includes the ability to annotate the same document in two different places (on the two screens) concurrently.
- I appreciate the ease with which I can navigate forward and backward between different portions of large documents (books), by pressing a single spot on the page. I often have to shift between a chapter, its “notes” section, and the references section, and the controls on PDF Expert make this simple and very fast. I appreciate that DT and DTTG do allow me to search through the indexed table of contents for PDFs that are bookmarked as such, but this requires a relatively laborious procedure for a series of events that might happen every minute or two.
- I find the annotation controls to be more accessible (and comprehensive) in PDF Expert.
- As far as I can tell, the document manipulation possibilities, including creating a table of contents, re-pagination, and using multiple forms of pagination in the same document (as in adding Roman numeral pages to a book before the Arabic numbered pages) are more robust in PDF Expert.
- The OCR capabilities of PDF Expert differ such that I end up with visually similar documents that are significantly smaller in size than those produced by DTP or Acrobat. I’m not sure why this is, but I appreciate it.
*In principle, though not fully in practice, I like the idea of the continuous experience of having PDF Expert on the Mac and the iPad. PDF Expert works well across both platforms on its own but, to my knowledge, the iPad implementation is not yet fully integrated with DTTG insofar as updates don’t cascade from PDF Expert back to DTTG consistently, especially if annotation takes place in PDF Expert when WiFi is unavailable. (See: Unable to edit DTTG 3 PDF in PDF Expert - SOLVED - #94 by jlc ; though the issue is marked as “solved,” the problem I mention still persisted the last time I tested it.) I had implemented my current workflow before realizing this problem, but I like most other aspects of the workflow, so I’m holding out in hope that this problem will be fixed.
If DevonThink had the above capabilities it would save me money, time, and effort to eliminate PDF Expert from my workflow, but at this point the navigation and manipulation capabilities are so key that I’ve basically stopped using my iPad for annotation while waiting for a fix for the above.
We’ve come some ways from the special link request, but the above, too, could be things for development to consider in future iterations.
With my thanks and best wishes, once again.