This would mean that external editing would have to be disabled for imported documents, something most users wouldn’t like.
What about a separate repository of ‘linked’ comments.
So that, when you right click on some text, there is an option similar to the “Link To”, but in this case an option to “Make Comment”, which would create a link to a new comment which you edit straight away (and later, when the text is clicked, go to).
The link could be marked slightly differently to ‘normal’ links, to announce it as a ‘comment’ link.
If the text is deleted then the comment is deleted (this is the way Scrivener handles it).
Opening in an external editor shouldn’t affect this, as normal links are usually preserved when I open an .rtf in, say, Nisus.
This wouldn’t work with plain text, of course, but then links don’t really work with plain text right now.
Thanks, we’ll consider this for future releases.
1 Like
For this purposes of professional commenting it would worth for DT to suggest a proprietary format, which would allow intense dedicated commenting, connecting and annotating, and from the other hand, would be easily converted to some popular formats (like rtf or so) and back without losses. I think such a solution would suit many users.
Yes, but if the text highlighted was copied to the annotations, then any external changes would still retain what the text that was originally annotated. I’m sure there are many that edit files externally, but for those (like me) who don’t, this would be great to have.
Or, can you lock the file in the OS with a Finder Comment saying “File locked due to annotations in DEVONthink”?
This sounds very similar to an extensive script package that has been made available in this thread:
I’ve been using it pretty effectively recently, and suggest anyone brave enough to experiment with some advanced functionality try it out on a test database. Though I would love to have a native feature that does this, I’m happy for the flexibility of scripting and the willingness of users to share their ideas.
On the principle that a comment is simply manually-typed text, why not using custom meta create a field type multi-line text?
As for finding all comments quickly that could be done using a combination of a tag ‘comment’ and/or a smart folder for tag ‘comment’. And for isolating comments on a particular subject a smart folder with the talent parameters.
Position the comment relative to the position in the text could done by highlighting the position in the text.
This seems similar to annotations. It might be a beautiful way to extend existing functionality by allowing this right-click to create multiple linked annotation files (especially with the possibility of auto-inserting back links as is currently the case). This would solve a lot of the knowledge management workflow questions & problems I’ve been thinking and posting about!
But having this single annotation you can insert there all the links to any DB items or items outside DB, links to Omni Outliner topics and so on. And all them in one place, like all connections of this file in a single annotation. It maybe that having multiple annotation files will create a mess.
And if you need separate dedicated links, you may use custom meta data for it.
The idea is for “knowledge blocks” to be able to be more granular. Instead of a monolithic document with everything in it, which pushes towards a hierarchical viewpoint, more granular addressing of knowledge blocks through individual files allows data to have one-to-many relationships with related ideas.
I agree this might require a function in DT to view items together, e.g. similar to Scrivener’s “scrivenings” view, which shows selected documents together and allows editing as though they’re the same document.
1 Like
The idea is for “knowledge blocks” to be able to be more granular. Instead of a monolithic document with everything in it, which pushes towards a hierarchical viewpoint, more granular addressing of knowledge blocks through individual files allows data to have one-to-many relationships with related ideas.
Bearing in mind, this is one philosophical viewpoint, not necessarily the only one.
Of course! That’s why I’m adding it to the discourse, making the request with others for it to be considered alongside the existing hierarchical viewpoint.
See this annotation not only as a content-wise document, but as a free space to encapsulate links to other “knowledge blocks” - individual files. It may be a linear list of links, it may be a complex hierarchical text with links in a wiki style, or some “freehand creative sketch”, connecting your “granules” in this or that way, whatever you want. What am I missing?
Let’s say I use the annotation document to copy in select quotes from a Web Archive and write my own summary of an article—let’s take for example a nice, long, informative essay like this one at Aeon. As I go through the essay, the annotations file starts to get quite extensive with many disparate ideas referenced, each heading in its own direction. For example, the linked essay includes pithy quotes on migration, cinema, painting, art criticism, how art changes through its depiction in secondary media, and home.
Once I’ve made this monolithic annotations file, how then to link directly to each of these disparate ideas from other relevant articles, essays, notes or annotations? I can only link to the entire annotation file.
Excellent writing finds connections between disparate ideas. To dig deeper on them in my research often means to digress from the original topic at hand. The linked essay is ostensibly biographical piece about a seminal moment in art criticism; but if I want to dig deeply into for e.g. the links between cinema and migration, I will end up with many articles and “chunks of knowledge” that have little to do with biography or art criticism. Being able to parse out individual blocks of knowledge into addressable pieces makes it possible to link these disparate ideas.
IMHO
If you want to catch all this “multi-directionness” in its full, I’d recommend to use an outliner software. I personally prefer Omni Outliner, where you can recreate as many structures and “sense webs” as you want, all with quotes you need and backlinks to the original text (for automatic backlinks you’ll need a little script). There are many ways to do it: you may create many OO files and write links to them in annotation file (in list, in text, …), you can make one OO file with backlinks, you can make the annotation a resume with the links to OO exact topics. You can build many-to-many connections, where your essay original will be connected to many OO thematic files (links in annotation file) and every OO file will be connected to many according essay files.
The other way - is to make a PDF and use its annotations as a quoting tool with description to each quote
You may also consider a wikilinks editing feature of DT3
I’m using the WikiLinks option for now, and it works quite well. The main limitation being the “addressability” of sections within a document. I’d be fine with working with many, smaller text files - but DT3 would need another view that allows viewing and seeing multiple selected files in aggregate.
I’ve tried OO. It’s fine for quick outlines, even extensive ones but it’s still a fundamentally one-document-oriented hierarchical concept. The only thing that gets close to how memory works for me is Roam… but I much prefer DT as a secure, stable, user-oriented knowledge management system. Roam could disappear tomorrow or have a breach and then all our notes would be out in the wild… Given that I’m a documentary filmmaker I can’t trust cloud-based SaaS for notes and research unless they have an e2e encryption philosophy.
To me DT3 is the state of the art of knowledge management. It just needs a few little nudges towards making networked thinking as easy and useful as the (amazing) hierarchical techniques currently are. Being able to address chunks of knowledge in some way would allow all the rest of the power of DT to work this way
Would these goals be achieved if DT3 were to support Transclusion in its WikiLinks - just as Wikipedia does?
1 Like
Transclusion would be a dream.
For me, being able to address individual knowledge chunks is the starting point, either through better “small documents acting as one” support or being able to link within a document (to headings say, or to paragraphs as Roam does)
Transclusion would amplify that ability. Then you can transclude just the relevant piece of information - inserting/quoting/transcluding a paragraph or code snippet and not the whole document
Note: File transclusion is possible in MultiMarkdown 6 but it is not specifically supported in DEVONthink. Perhaps in the future, but Development would have to assess this further, especially as DEVONthink is not a Markdown-only ecosystem.
1 Like
I just had an “aha” moment, posted over in another thread (link below). But basically… if DT were to support both anchor tags in x-callback URLs and transclusion, it would create a powerful and standards-based solution for the networked-knowledge workflow. Both already a part of the MMD6 spec, and DT3 already supports anchor tags in MMD links—just not between documents. It could be a big win with a tweak to already-existing functionality.
Here’s the larger thought bubble: