Questions about rethinking Annotation, Tags, and Notes

My main point is not that the scripts should work together. The point is that DT does not offer a more practical way to extract pdf content efficiently into note files, and linking them more accurately…

The process borders primitive, cutting and pasting info from sources into RTF files, or whatever other format (and yes, we use page links, but that’s not enough). Now, the script eliminates this step to a good extent, but that’s not what we expect from a reputable software like DT. We expect solutions in the form of better, more powerful organizational tools. At this time devoted customers are producing and accomplishing more through scripts than what DT has done in the latest version regarding note taking. If package production were a possibility (like those produced for Sublime Text), that would be quite an improvement. But DT does not allow that, so a segment of users find themselves between a hard place and a rock.

A script is an external automation, not a standard inherent feature of DT. We expect better note taking tools for more seamless data gathering, collecting, and connecting it in a network suitable for present and future work. Bottom line, we want the product to excel. Staff, please take it as constructive criticism. If DT offered more specific tools for note taking people would not be mass producing scripts like we see happening. Take scripts as example and a starting point.

@korm, thanks for your response. I will try to be more specific. First a clarification. What is inadequate and primitive is DT’s native note taking tools, not the contributing scripts. Here are basic steps that would move some of us, me included, in the right direction. For each of these it would be nice to have 3 ways to run them: shortcut, right click, and toolbar icon.

  1. One file for more than one quotation. We have this already in the script QuoteHighlight&Annotate (Script 1) but it would be nice if we could customize output under Preferences. Furthermore, the feature should not break down like Script 1 does when used in conjunction with “Make annotation files for each citation plus notes+tags” (Script 2).

  2. One file for a single quote. Script 2 fits here, but it should be used in conjuction with other scripts/note taking features such as 1 above. Again, output options could be customized in Preferences, because after all only a fraction of us know how to produce a script. Preferences could have a special tab for Note Taking where one could choose the fields that can be output into the RTF note file.

Another method of extracting PDF content would be to mark the selected quotation/material with a vertical line, which would in turn generate a “node” and resulting RTF note for annotation (yes, qualitative software has them but I am not saying DT should fully emulate them). Tags would then be applied at the level of nodes, a deeper level than files.

Both scripts and may be other features should be integrated into the menu as native features of DT. I hope these examples have helped to illustrate how DT can implement better note taking functionality.

and

Perhaps you could do sketch (even a hand drawn one) of how the UI might work in the scenario you are describing. I can’t see an easy way to distinguish between a primary annotation file and annotations files that relate only to a specific portion of text. Remember the URL field can only link to one document or group - this is the annotation document or annotation.

If you have many quotations that you want to link into a single file you can merge them afterwards simply by selecting and merging all the text files in the linked tag group. This has the advantage that your annotations can also be placed in the order you want them in the final text document before you merge them.

The next version of the “Make annotation files for each citation plus notes+tags” script will adopt the template model of DT’s existing annotation script so you will be able to choose what fields you want in the annotation.

Actually you can emulate something like this by highlighting the text, drawing a link box around it, and pasting the the url of the annotation into the url link of the link box. Its clumsy so wont work for anything other than the occasional annotation. Unfortunately I can’t see any way to automate it because there is no programmatic access to the pdf annotation tools.

As a matter of course I tend to highlight or mark with an adjacent red line any passage which I have annotated. Because I preface my annotations with a page number its generally no more than two clicks to go directly to the appropriate annotation from the page of the pdf (One click on the url link to open the tag group and a second click to open the annotation itself) I also keep my passage highlight colour and my annotation highlight colour the same for any particular passage. This also helps with quick visual linking.

On different annotation styles

I think the virtue of DT’s approach is the ability to adopt a style of annotation which works for you (see for example this thread - [url]"Highlights" app - export annotations to DEVONthink, etc.] ). Recently for example, and because the type of document being annotated, is not susceptible of easy contextual annotation, I have been clipping whole series of pdf pages as annotations. This way I can more easily have a summary of the relevant marked up pages in context.

There is a lot written on the web about how people use DT different annotation techniques in in their writing. I found this article by Steven Berlin Johnson particularly good http://www.stevenberlinjohnson.com/movabletype/archives/000230.html.

I also cannot commend Alex Strick’s recent post highly enough especially if you use a lot of kindle reference material - http://www.alexstrick.com/a-different-place/2014/10/note-taking-jujitsu-or-how-i-make-sense-of-what-i-read

Frederiko

1 Like

Frederiko, it would not be a good idea for me to sketch what the UI of an implemented script under DT would look like because I have no clue how one would do it; I only know that such functionality would be very useful. I also “can’t see an easy way to distinguish between a primary annotation file and annotation files that relate only to a specific portion of the text.” That is precisely the question that needs to be pondered. A solution would open new vistas, so that note files could be linked in additional, new ways. For example, URL links a file, but nodes (parts of a document) could be interlinked. Imagine that…

Yes, tag groups can be used to merge several quotations into one file. But node tagging, the ability to tag paragraphs, sentences etc… when needed, is even a more granular, deeper form of control than the one asked about in the original question above.

Looking forward to the new version of Make Annotation Files for Each Citation plus Notes+Tags. Wondering how I can be notified when ready… Thank you for pointing out to me the Annotation app and some other links.

Korm, if DT keeps implementing good PDF markup (rectangle, different collors, circle, square, arrows) why not extend the same zeal toward thinking about new tools for note taking as illustrated by the two scripts referred to above? It need not be relegated to a standalone application… we would encourage DT to think in that direction.

Yes, Script 1 (QuoteHighlight&Annotate script) and 2 (Make annotation files for each citation plus notes+tags) could be implemented in a Menu, “Annotation” menu for example. To use both in conjuction involves using both of them which is now impossible. Let me explain: if one uses script 2 after using script 1, script 1 breaks down at the last point of insertion when you come back to it to add further quoted material. At that point the script 1 no longer works (because the URL changes to script 2 I suppose).

One could think about a list of options to include in a menu for script 1 and 2, but for now we just want both of them to work in conjunction with each other, Once that’s taken care of users can think of ideas about how best to implement them in DT. Make the functionality of these scripts native in DT, and users will appreciate it.

Yes, nodes are lots of fun. Atlas.ti, NVivo and others specialize in them. No, a node is not a separate file, but a fragment (a paragraph, sentence, clause, word, letter, whatever you choose) that can be tagged for a myriad of purposes. There is a hierarchical tree for node tags, much the same way DT does with doc tags. Imagine a tight association between the two… yes DT dreamland. Yes, a window opens showing your abacadabra nodes.

This deep control in the service of academic research is not beyond DT’s reach and scope, believe me.

[quote=“Nhaps” …nodes are lots of fun. Atlas.ti, NVivo and others specialize in them. No, a node is not a separate file, but a fragment (a paragraph, sentence, clause, word, letter, whatever you choose) that can be tagged for a myriad of purposes. There is a hierarchical tree for node tags, much the same way DT does with doc tags. Imagine a tight association between the two… yes DT dreamland. Yes, a window opens showing your abacadabra nodes.

This deep control in the service of academic research is not beyond DT’s reach and scope, believe me.[/quote]
Ten years ago or so there were some relatively popular programs that included the potential of using them to create “nodes” such as you want, although it would have required work to further develop that potential. They are no longer popular. Very few people found such potential useful to them. Other applications such as DEVONthink have supplanted them for document/information management needs. DEVONthink has been a successful application precisely because it is a flexible general purpose approach to working with the information content of document collections that meets many needs with the included tool set, and is capable of extension to meet many other specialized needs.

Here’s my take: If the future of academic research becomes rooted in the need to parse computer-searchable nodes in existing collections of text, academic research is a sterile dead end. No academic researcher can compete with developing versions of IBM’s Watson.

I’ve long been in awe of an academic researcher named James R. Partington. He was a chemist and historian who made important contributions in both fields, especially the history of science. He documented and analyzed the development of scientific concepts and discoveries in many languages and cultures. He did this without computer assistance, based on extensive research, comprehension and understanding, and thinking. He was a marvelous writer. He used footnotes extensively in his many books and papers. Although his library was stacked with references, those footnotes were often created from memory, without need to rummage through a stack of books or paper–and they were correct and relevant. He made major contributions to knowledge and influenced generations who succeeded him. IBM’s Watson is nowhere near being able to emulate Partington’s work.

It’s my belief that Partington would have liked DEVONthink, if only to compact his collection of references. And he would have dismissed QDA software as trivial, because it has little use for the kinds of analysis he made across time and cultures. He didn’t merely look at words. He looked at concepts, which vary widely in the ways they can be expressed. He was gifted in grasping concepts wherever they might be found. Most of us (me included) are not so gifted. :slight_smile:

I’m so glad this post is getting into a deeper discussion about new approaches and technical possibilities with annotations… So many great questions, ideas, suggestions here.
Ok, @korm asked for specific suggestions about the kind of annotation function we’d ideally like DTP to create. Based on my experience, and tinkering w/ the various scripts, here’s my take…

I’d like some such an annotation document to first be created though a blend of (1) the DTP Annotation template – through its creation of basic parameters the title and ID of source document (as well as the comment icon that it graphs into the source file), time and date in which the annotation document was created (I don’t need the “user” part of the template is necessary) and (2) the “Make annotation files for each citation plus notes+tags” script [url]Make an Annotation with Links, Notes, Tags v2] which, in its most recent version, includes a brilliant interface via Pashua for creating an annotation note, and setting up a useful Tagging taxonomy that relates to the annotation file, and flexibility for placing the note in a particular area.

And then

The Annotation document would have the functions of the “QuoteHighlight&Annotate script” [url]QuoteHighlight&Annotate script] so that DTP users are able to copy multiple quotes and easily link to them in the source documents. (I think the current way of doing this in DTP’s annotation template is far too clunky, and the script really excels at this.) As I said, @Frederiko has been helping me w/ annotation scripts, and he said there might be a way to add / create more of the Text links for that script, so that it actively highlights the quoted/selected text when you click on the link (again, he credited @korm for it’s creation). This is an incredibly valuable feature that enables DTP uses to save much time by quickly zipping to that exact part of the text, and flashing it with a quick highlight. Also, as I said in my previous post, it seems that the “QuoteHighlight&Annotate script” handles UTF-8 conversion problems better than the “…citation plus notes+tags” script.

The only dilemma I haven’t solved w/ is: in general, I’d like to have one annotation linked to one source document, so that I can have all of my notes and quotes located in the same space, but I can also see that it would sometimes be useful to create a separate and distinct note per annotated section (which was the original purpose behind the creation of the “Make annotation files for each citation plus notes+tags” script). In my case, I could imagine doing this for follow up research or new source or if I wanted to create a sidebar issue. So, I suppose my dilemma is that I’d like to ideal have the ability to create would be great to still have the ability to create such a separate and distinct annotation note per an annotated section while also being able to somehow link it to the main, annotation document. That might be a bridge too far, but you asked what we’d ideally like annotation scripts / functions to look like, and there you go…

I hope this has been helpful. I’m happy to provide more info if anyone would like…

Thank you again!

@Bill, there is no appeal to make DT turn back the clock or detract from its main core objective of being a general, powerful information manager. But DT’s flexibility need not be compromised by the implementation of more specific tools for note taking/annotation. These can be hidden in menus so as to not disturb those that don’t need these tools. The fact that developers and other enthusiasts are producing scripts is because there is a need for this functionality.

In my view academic research does not depend on parsing searchable nodes for conceptual reasoning. But conceptual thinking is facilitated by the gathering of related nodes together, and rearranging these nodes in different patterns within the network of sources we keep in DT.

Thank you for enlightening us about Partington. But I have a hard time understanding why you equate my node tagging (intra-document) suggestion for DT as a stumbling block to effective research and thinking. QDA node tagging functionality provides a canvas to paint whatever concepts you are working on. Nodes (intra-document tagging) serve as anchor points to organize semantic relationships.

Of course Partington would have liked DT, just like all of us here. But that does not mean we have to be stuck in the 80’s with only inter-document tagging available to us. Contrary to your opinion, QDA node tagging is trivial only to those who use words independently from concepts related to them. The fact that one word or expression can convey several different concepts is no reason to discard intra-document tagging (and linked RTF notes) at all. :slight_smile:

@ jprint714 I’ve often shared your dilemma. Long before the appearance of the Annotation template, I was in the habit of creating a rich text note linked to a referenced document, to hold notes about it.

Sometimes I found it useful to create separate “spinoff” notes that went into more detail about an item, and/or might refer to other notes or references. It was possible to do this, although a bit of extra work is entailed to enter appropriate links, as well as to indicate in the primary note a searchable “cue” that such additional notes exist and links to them. A bit of grunt work, but doable.

The Annotation template provided automation assistance in creation of the principal note document and linking to the referenced document, as well as a link to the Annotation note from the referenced document.

The Annotation template didn’t change my habit of sometimes spinning off additional notes linked manually to and from the Annotation note and possibly to other notes and references as well. That still involves some grunt work.

I haven’t heavily used Frederiko’s annotation script yet, but see it as a possible way in my workflows to reduce some grunt work, when used as a means to “explode” individual items outside the Annotation note. There are number of ways in which one can searchably indicate the existence of such additional annotation notes, some of which could be customized in the script. Still other possibilities will probably appear in the next generation of DEVONthink.

@ Nhaps: I mentioned J. R. Partington as an exemplar academic researcher because (although he did not have the benefit of computer assistance) he was a prolific writer and his influence as a teacher and researcher has long outlasted him. He was of course a genius, fluent in many languages and a workaholic. Above all, he could think “outside the box” and illuminate topics from new perspectives. His accomplishments used the tools available to him: wide-ranging knowledge, print, pen and paper and a remarkable intelligence.

DEVONthink has a niche market. Although there are many thousands of Mac owners who have substantial information collections and feel a need to use software for that purpose (and we do well among such users), millions of Mac owners do not feel that need (however we may wish they did).

DEVONthink is already a complex database system, especially in the Pro and Pro Office editions. It provides a lot of tools and commands. That makes it useful to many people, some of whom accomplish extraordinary things, and some of whom, like korm and Frederiko (to mention only two of many), provide assistance to other users by developing scripts that enhance DEVONthink, or provide tips on how DEVONthink meets specialized needs…

Ironically, it is the complexity–in terms of the number of tools, commands and scripting capability–that almost guarantees that DEVONthink will not become a mass market product. Many people will download a trial copy, take a look at the user manual and decide the learning curve is too high for them. I try to get them to start simply, to satisfy a simple need and then experiment to see what other features might be useful, a stage at a time.

QDA software is an even smaller niche market, and has its own niches. Much QDA software is hardwired for cases in which a researcher tightly controls the environment and vocabulary to be analyzed, such as responses in a forum group conducted by an advertising or marketing agency, or a psychology experiment. That’s not a direction in which DEVONthink is likely to go. But we’ve received comments from researchers who use both DEVONthink and specialized QDA software.

The next generation of DEVONthink will in some respects appear simpler, yet provide still more power. After more than 12 years of experience and the attention of the developers to user suggestions, I expect that it will please current users and am hopeful that it will attract a larger audience. I’ve been a heavy user of DEVONthink myself, since 2002.

Having dropped a hint that there’s a next generation coming along, it’s our policy not to predict a release date. To borrow from an old commercial, no wine before its time. :slight_smile:

@ Bill, thank you for your posts, always useful. I agree with you that Partington was one of a kind thinker. I still believe, however, that intra-document tagging can take note-taking to a higher level in DT. I sincerely advocate that intra-tagged notes inside a file (which DT does not offer) in conjunction with inter-tagged ones (DT offers) can increase the chances of innovative thinking. Right now people can only tag documents to arrange them in different configurations. Imagine raising the level of precision with the ability to tag paragraphs and any other portion of a file!

If intra-tagging functionality sounds too specialized then tuck it away in a menu so those that do not want to see it are spared. But let dedicated DT users benefit from necessary research features under the hood. We must not stifle DT’s ability to provide differente strokes for different folks. DT provides a haven for several niches.

DT is complex because the UI is not streamlined, and that’s why it scares people away. But appealing and user-friendly software can be nevertheless powerful, and that’s what I hope for the next release as you alluded to.

Particular attention should be given to the excellent request of jpring714 above for the new version, as well as tweaking of the scripts of korm and frederick in the meantime.

If and only if DT’s UI is revamped and simplified—without excluding any features we enjoy—will the company acquire a large following (and intra-tagging will contribute to it). Let us not relegate intra-tagging capability to the genre of QDA software because it is a basic feature of research. Intra-tagging is the glue that binds lexical and semantic analysis together. Paragraphs can have x number of tags which relate them to different as well as complementary concepts. This is extremely necessary for qualitative research in social sciences, psychology, history, theology, etc…

If the next generation of DT will appear simpler and more powerful, then it should seriously consider free-flow intra-tagging functionality. But don’t take my word for it, just look at what folks came up with in this link: http://www.tagnotate.com. And by the way, this dedicated intra-tagging app was created for the iPad only. For the very niche of people looking for a combination of simplicity and power for note taking (look at their screenshots).

Here is the first question from their FAQ:

They report a deskop version of the app is coming soon.

On the one hand it’s a shame that we don’t have such a thing yet (with the exception of the intra-tagging offered by circus ponies notebook), but on the other, what a nice opportunity lies ahead… :stuck_out_tongue:

This looks very interesting. Have you used it, by any chance? If so - could you clarify three points please?

A.) Can you import an already annotated document (from Goodreader, for example), and add tags to those annotations after the fact - or must the annotations be ‘created’ inside TagNotate?
B.) Can the user set the highlight colours with a colour wheel?
C.) Upon export - are the tags picked up in DTPO?

@Bill_DeVille, per your quasi-announcement about an upcoming DTP release that will (hopefully) include new, better annotation templates and your shared dilemma of whether to “create separate ‘spinoff’ notes that went into more detail about an item,” may I suggest… Please consider the suggestions I detailed in a new, improved annotation template (in answers to @korm’s question, and based on users’ hard work to create alternative models) – and maybe include a way to link / unlink “separate ‘spinoff’ notes” to the main annotation template document (should that be possible).

Just a thought…

I hope this update comes soon! Thanks…

Hi Cassady, I found out about TagNotate only two days ago, and it will take some time for me to see it more thoroughly. It would be nice if you could join me in buying and testing it ($2.99) and share relevant insights in this thread on how to better handle note taking. So far I can see that intra-tagging functionality is certainly promising, and this intersects with the aim of the two scripts above: improving the level of control regarding notes.

I’ll probably take the plunge on it in next week or so, and pop up something here - so sure!

Already emailed the developers with some of my queries - will also pop something up if/when I get a reply.

Wow!! Thanks for mentioning this app. And they’re working on a desktop version. It looks very interesting. It’s on sale now as well!

TagNotate is pretty simple to get up and run and looks like it has promise beyond the confines of the iPad app. The UI is very pleasant, easy to understand, and enjoyable to use. Though, for now the app’s value remains mainly within the proprietary app. On the upside, the authors have prepared the way for integration into other apps. From a tagging/annotation workflow perspective – this app makes the linkage very easy. The abilty to AND / OR tags into a kind of on-the-fly smart group is useful, and then the annotations can be collected and exported. The tags are included as string in an additional line of text at the bottom of the annotation.

Some comments:

[size=125]Pro[/size]

  • Each notation is stored in that PDF’s metadata in the “Author” element. Each notation has a UID (not sure if the domain is unique to that file or genuinely unique), and the “tags” are stored in a set. 3rd parties could take advantage of this feature. Unfortunately, DEVONthink (as of 2.8.2 on Yosemite) does not display the contents the PDF-specific metadata, including the “Author” element, for PDF’s notated with TagNotate. The contents of “Author” can be viewed in Preview, Acrobat, etc., on Yosemite. Example:

{"uid":"D62F3673-689C-4873-BB42-FB12AA1F7CEB","tags":["Cicero","Ovid","Cato","Virgil"]}

I did not yet test to see if TagNotate overwrites existing contents in “Author”, which would be a problem especially for academic users.

  • Simple interface – better design than many iPad PDF programs.
  • Clipboard is a good feature for collecting annotation info to copy into other documents.

[size=125]Con / Suggested Improvements[/size]

  • I dislike the privacy notice and the fact that so-called non-identifiable information is sent to the mothership with no opt-out. Why the authors are collecting the tags that users create is mysterious. This is 2014. Opt-outs are essential. Over here, I work with numerous customer-sensitive documents and would be unable to use TagNotate because of it’s non-conformance with common corporate and governmental security policies. It’s a bizarre design choice that is not essential for supporting the app’s mission.
  • Not updated for iOS 8 – this matters because two of the strongest features – UID-identifiable annotations and the clipboard mechanism – would play very well in the iOS ecosystem. Also, there’s been no update since May 2014. This is forever in iOS land and possibly a red flag. I discount the promise of a desktop version – lots of developers promise new platforms or versions for years on end. (Ahem, we know who :cry: )
  • Annotation types are limited to highlights, free-form drawing, text, and notes. More of a personal preference, but the absence of boxes and stamps is odd. OTOH, TagNotate recognizes these latter types and does not seem to mess them up.
  • No apparent use of call-back URLs – which would be important for tighter integration of TagNotate with other apps on the iPad – including DTTG. The bones of this feature (the UID) is present.
  • Clipboard copies of annoations + tags is plain text. Would be a nice-to-have if the export were in Markdown format with a link back to the annotation.
  • It would be useful to have tag sets, in three senses:
    [*]if I’m tagging news articles I want a different set of tags than if I’m tagging information theory articles
  • I want to import/export a tag set for use elsewhere – perhaps as an OPML list
  • Hierarchical tags
    [/*:m]

Korm, thanks for the time spent in reviewing TagNotate. I am glad that DT staff is taking an interest in intra-tagging functionality. I have contacted the TagNotate developer and I hope that he will be able to join us in this thread.

Yes, agreed – thanks @korm for your review!

While we’re on the topic of using external apps to make annotation, tagging, and notes work better, why not use the Highlights app that @korm suggested in this post: [url]"Highlights" app - export annotations to DEVONthink, etc.] ?

Or does it lack some feature that doesn’t address what we’re seeking?

By the way, I traded a few emails w/ one of the folks at Highlighter, and he said they’ll look into a way to include links to the PDF source files for future editions (which he promised will be out soon).

If Highlighters is a viable options, maybe there’s some way to integrate it w/ the kinds of tagging options we’re also seeking.

Thoughts?

jprint714, here is my take on:

Highlights app for mac. It overlapps with Skim which means tagging is out of the picture for now, but it has advantages. I asked Highlights developer to include page link on the righ panel so we can quickly locate a quotation in context back on the pdf on the left side. To my surprise they accepted it, and it’s slated for next release.

TagNotate for iPad. I will be waiting for a response from the developer of TagNotate to korm’s review above. The issue at stake here is the potential of intra-document tagging. While this is not yet materialized we should continue encouraging DT developers and volunteers to improve on the scripts are you requested above. Truth be said, both apps have complementary features, and a promising future. But TagNotate is more innovative.

@Nhaps, thanks for your reply… I haven’t tried TagNote yet, but I’ll look into it.

For now, process involves:

-Saving docs and article on webpages as PDFs
-Saving those PDFs in DropBox folders (organized by topics)
-Syncing my DropBox folders within iPad’s GoodReader app
-Reading and marking up my PDFs in GoodReader
*Eventually I route the marked up PDFs back to DTP where I…well, where I’d like to create tags for the marked up PDFs, create annotated files that would include quotes and links to the marked up bits, and then (eventually)…compile memos that consist of quoted and linked bits from annotated files.

As I understand it, if I take my annotated PDFs from GoodReader, and open them through the Highlights app, Highlights will auto-generate a Markdown document based on of selected annotations and notes I created in GoodReader. And…they say they might soon be able to create a version that includes a Markdown document with links to those annotations.

So, what does TagNote have that Highlights doesn’t offer – or what do they have that might be a better solution to my process? (By the way, I’m not tethered to my process – it’s just the best system I think I divined – so I’m happy to consider an alternate process.)

Thanks!

@jprint714. My workflow differs from yours, I don’t use GoodReader, and I am still learning how to use TagNotate.

But here is the main advantage that Highlights does not have: search and filter your global (all pdfs) or single (pdf) annotations by tags (underlining, highlighting, comments). TagNotate’s forte is smart listing your annotations on-the-fly by tags. You can search by page number, by tag or tags. By clicking the quotations you are taken immediately to the location in the pdf and back to the annotation again. Instructions are easy, accessible at the bottom.

Highlight is faster because underlining generates the quotation automatically, but gives you only bright that nasty bright yellow. TagNotate is slower to create but gives you more options plus one or more tags for each annotation.

What does this do for your research? Well, for now it offers a tighter integration between the whole pdf and your annotations. It equips you to see the whole article from different angles through the selection of tags. In this process you come to get a better handle on the material.

How do you transfer your notes to mac? Well, there is a clipboard option within TagNotate, which you can use to copy to Notes through iCloud or any other way. Tags are included as text file along with quotations. Or you can save the PDF to dropbox and the all annotations will be visible in Preview and Skim. All this for $2.99? You gotta be kidding me, get outa here. :smiley:

From Skim you can explode to RTF files in DevonThink, or combine into one RTF. As korm said, all tags are located in the “author” metadata field and unfortunately DT cannot read them yet. So for now I am using TagNotate to digest my pdfs through the benefit of intradocument tagging. And if I need to paste a quotation it’s right there on DT. (The only problem then is basically having to reproduce the tags in Devonthink, for each exploded RTF, which I am not doing btw)