In this lesson, I show you how to make annotations in Highlights App, export the annotations to Finder or DEVONThink in markdown, import the annotations into Tinderbox, and curate them. I also show you how to leverage Zotero for citation mapping.
I explain how I use four annotation colors (yellow, orange, red, and blue), strikethroughs, underlines, images, and personal notes to differentiate my annotations/highlights and capture my own ideas. I then show how, using Tinderbox prototypes and action code, I parse the annotations and can start making sense out of them.
NOTE: I’m still refining this process, please feel to reach out with you have any questions or suggestions on how to make it better. Also, I ran into a couple of issues that I need to speak to the Eastgate and Highlights team about. The Highlights app does not split the files when exporting to the finder, and I’m having trouble parsing the un-split file in Tinderbox. I’ll update you once I’ve figured this out.
00:00 - Introduction
00:12 - Introduction to the 4Cs
00:57 - The Apps: Highlights, Tinderbox, etc.
03:07 - My use of DEVONThink
03:28 - Annotating in Highlights App
05:37 - Setting req. Highlights App preferences
06:12 - Export out of Highlights App to Finder
07:29 - Importing annotations into Tinderbox
09:40 - Prototype types and “The Stamp”
10:41 - Applying the Stamp
12:27 - Linking back to the PDF
12:37 - Linking to website
12:53 - Wrapping up
14:13 - Please be sure to like the video
14:16 - Don’t forget to subscribe
Hi all, there are a few issues with this work flow that I could use help with to make it more efficient.
Thanks, please do give me your feedback. Also, if you can help me refine a few of the workflow elements, especially because you know DEVONThink, that would be great. Few points.
As noted in the video, having trouble parsing the Finder export process.
Highlights does not split the file on the Finder export, that is problem 1. Problem 2 is Tinderbox is not recognizing *** as a delimiter for the attributes and thus I can’t explode the single file. I’ve reached out to both Highlights and @eastgate on this topic.
When you export times into DT3 they get tagged, which if you do this often can create a huge tag list and from a UI point of view can get VERY annoying. I think there’s a “tagging” strategy that can/should be implemented within the workflow to improve this, I’ve just not figure it out yet. Also, this could help with the image path issue I raise in the video.
Next steps in the 4C process
This specific set of tasks only focuses on the collection and the smallest part of curation (aka annotation hygiene). It does not yet zero in on more refined curration. I think we can a deeper conversation about that.
Anyway, thanks. Looking forward to the conversation
@eastgate solved the explode problem I was having. I neglected to consider that Explod is using regex and asterisk are characters that need to be escaped, so to explode the combined highlights file I’d need to use *** and not ***. This does pose another issue though. My stamp is optimized for the markdown format going out of the split notes, not the combined note. We’ll need to modify the stamp for the combined now. Easily doable, with some testing.
This is a fantastic video, Michael! You have addressed a topic that I have been struggling with for the past few months (how to make use of highlights from reference documents in Tinderbox).
In your opinion is Zotero a necessary part of the workflow, or can DEVONthink do the job? I keep all of my reference material in DEVONthink and have had good results linking the documents to the notes I take in Tinderbox. I’m not familiar with Zotero and would like to consider using it.
Zotero? Absolutely note. Personally, I find the user of a dedicated reference manager experience useful. I guess DEVONthink could play this role, but I’ve never been able to use it effectively as a reference manager.
Key point: a reference manager, like Zotero, is used for not only managing the citation but dynamically creating in-line citations and bibliographies. While DEVONThink is a file repository. They are two very different things. When using both I get the best of both worlds. As shown in the video I can link back to DEVONThink and Highlights, as you note above. However, by integration with the Zotero (and Pandoc) I can then automatically leverage the $CitationKey to generate the inline citations and bibliography, not just for one item shown above, but for hundreds (my larges doc is 200 printed pages with a few hundred citations, all processed automatically).
Finally, you don’t need DEVONThink or Zotero, you just need Highlights (or any PDF annotator that support markdown export). Granted, as I noted above, there will be needed modification to the stamp if you go this route.
Believer! I used to be a PDF Export only pdf annotator, but with Michaels workflow and the ability to have separate, markdown DEVON files using Highlighter Pro for my annotations, I have made the switch.
Thank you Michael! This is awesome.
My workflow differs in one aspect: I use Tinderbox as my Zettelkasten and for me the most important criteria to a Zettel is: it has to be written in my own words.
So I use Highlights too for the annotations (since the export function of MarginNote3 is not usable at all) on my iPad. Before I annotate I add all PDFs to DevonThink. Tried this with a reference folder (indexing) but this will give me problems syncing the content with DevonThink to Go - so I add them into DT. Highlights allows to open the PDFs out of DevonThink To Go on the iPad. That’s important because this works in the opposite direction too. All my annotations automatically appear in DevonThink.
Then I go through the annotations and create my Zettel in Tinderbox - merging all ideas I like to see as a Zettel from the annotation into a note in Tinderbox. I add a reference from DT into the Tinderbox note. So I can open the pdf on the relevant page out of Tinderbox.
Also - after finishing the work with the PDF I export the file again and add it to my Bookends entry. This way I have the annotations redundant in my reference manger and in DevonThink. This last step would work much better if indexing a folder (including the sync to my iPad) would work in DevonThink. Having only one PDF to work on is better then having two copies at different places.
Wouldn’t be easier for you to just export annotations using Devonthink’s Summarise Highlights option (sheet)? This way one of your notes’ attribute will get link to relevant page, so you don’t have to input them manually.
not really - since I don’t need the annotations themselves in Tinderbox but my own ideas linked to those annotations - sometimes it is a 1:1 relation, sometimes a 1:n…
Also, as has been pointed out, the linkbacks with the Highlights method goes back to the actual PDF, not the annotation.
The cleaner markdown makes writing the regex easier and gives me more confidence that I’ll ultimately end up with what I want. Also, the DEVONThink markdown does not include the color as a piece of text that I can convert into an attribute.
but why to export annotations to Markdown files? The way I do it is to export them as a “Sheet”(TSV or CSV file- I forgot which one was it), move it do Desktop first and then to TBX. What I get are notes with attributes containing highlighted text, my note, and link back to annotation in Devonthink. This method has one drawback, namely Devonthink export doesn’t recognise colour of highlighted text. A workaround would be to add tags (like “#quote”) while highlighting and run an agent in TBX.
Great question. The answer is, because you need to work within the context of what you have. Highlights today exports a reliable and consistent markdown format, so I use that. It does NOT export CSV or TSV. Also, there are problems with CSV or TSV. If there are special chargers, e.g. commas semi-colons, invisible characters in the CSV or TSV they can easily trip up the Tinderbox parser and make a mess of things. For now, for me, markdown is much or reliable and consistent. I get what I expect.
What I meant was Devonthink’s Summarise Highlights option called “Sheet”. I have to agree that TSV/CSV file format has its own limitations. If markdown works for you, that’s great! I just wanted to share how I do it hoping that someone may find it useful.
Oh, don’t get me wrong, I love your CSV/TSV suggestion, given the right context it is a perfect solution. Unfortunately, in the context of getting the annotating text out of highlights, it will not work. Actually, my initial approach and ask of the Highlights team was to get a CSV/TSV export, but that request has not materialized yet, so went with markdown. We’re on the same page.
Coming to this late, but just wanted to say that I’ve put together a pretty good approach to annotating with MarginNote, rendering the exports via OmniOutliner (with help of an AppleScript), and then importing them into Tinderbox. I also Devonthink Pro, but it’s not essential to this process. Anyway, it’s not a flawless workflow (mostly because of the slow pace of improvements on the MarginNote side), and I’m still working out the bugs, but it does provide another approach to integrating annotated notes with Tinderbox. Happy to further discuss it if you’d like…
To be clear, I use OmniOutliner – with help of an AppleScript – to render and prepare my MarginNote-created annotated files for Tinderbox. Setting up that process took a long time to figure out…
