I know that DevonThink 2.0 is still in development and that DevonTechnologies can’t reveal or promise anything too specific, but I was wondering if anyone had any additional information at this point about what sort of meta-data will be supported. I have developed my own system for “tagging” items in DevonThink, but am wary of committing to it fully, lest it be made obsolete by future versions of the program. Right now I cram everything (bibliographic info, keywords, image descriptions, etc.) into the comments field. But if these items are split up into multiple meta-fields, I imagine myself having to go in manually and extract the info for each item–in which case I’m probably better off waiting until the new version comes out before entering too much metadata or implementing a full-blown bibliographic and keyword system. Any advice would be much appreciated. Also is there an projected release date for the new version yet?
Funny, I was thinking of posting the same general questions (minus the release date), since I am also trying to implement a similar metadata structure on my notes and files. I look forward to the new expanses of metadata that have been hinted at by the DT crew, but I am also starting to worry about how much manual work I would have to do to reparse all the metadata currently crammed all over the place.
Anyone from the DT dep’t? Any suggestions on how to stay on the safe side?
Sorry if my release date question seems a little pushy. I simply don’t know if anything official has been announced yet. If not, then no need to answer…
I’m sure it’s true that tagging metadata can increase the value of a collection of notes and references.
And I suspect that any evolutionary developments in DT Pro won’t vitiate the value of your current tagging methods, and may well make them easier in the future.
As for myself, I long ago gave up attempts to systematically tag individual notes and references.
That’s because I quite often dump hundreds of new contents at a time into my database, and don’t have the time or inclination to go back to each and tag it. So I’m lazy. (You will find, however, threads on this forum about using a script to modify the Comment field of a batch of documents.)
But I’ve got a more general problem with tagging, for the range of literature that my interests include, dealing with environmental issues. If I’m looking at investigation of a contaminated waste site, I’ll deal with sampling issues (choosing contaminants of interest, statistical suitability of sampling locations and analytical procedures and results), soil, releases to the atmosphere, surface water runoff and groundwater concerns, and so on. Evaluation of the collected data then involves related regulatory issues, potential health effects of contaminants, risk assessments and methodologies for establishing satisfactory “cleanup” limits, and so on. Selection and design of the remedial action then requires review of potential methodologies (in situ or removal). For in situ remedies, review of case histories and alternative methodologies, which may involve a broad range of treatment or isolation technologies and, of course, cost-benefit analysis.
That involves lots of disciplines, ranging from the law to chemistry to geology to bacteriology to molecular biology to toxicology to … And I’ve never figured out a tagging or classification scheme that would cover the range of ways pieces of information will fit together for my needs.
Instead, I generally rely on DT Pro to help me pull together items for a given purpose.
-
I will use group/subgroup organization for a writing project and put my notes and perhaps replicants of interesting references in that collection.
-
I use searches to identify items of interest and often create temporary groups of replicants of the search results so that I can quickly find primary reference materials for a project in course. (Or create smart groups for the same purpose.) Sometimes I will do secondary searches on those groups or smart groups and “break out” additional grouped items for reference.
-
I make heavy use of See Also or the contextual menu option, See Selected Text to review DT Pro’s suggestions of potentially related items.
-
I make heavy use of the Option key for a selected word to see a list of items that also contain that word. I often open a document in its own window, click the Words button and scan the list of terms for other “keywords” that might help me assemble useful material.
-
I often make list outlines in my project file that link to the sections I’m writing, to particularly useful references, with notes about them, and so on. I usually use static links, but Wiki links would work also.
Although I’ve never been able to figure out a tagging system that could work for me, I can interact with DT Pro to identify and pull together a set of material that will be useful for a particular project. That’s metadata of a different sort, using my determinations of utility of search results, See Also lists, etc. to pull together for a project selected reference materials. This depends on the content of my database plus my ‘reactions’ and usefulness determinations about content that DT Pro has served up to me.
I’m using DT Pro’s lightning fast searches and AI features to do the drudge work, but I’m responsible for deciding what and how to use from the material made available. In progress of writing, I can always look for other material as needed. I’ve already got thousands of items in my collection. If I find that I’m light in reference material on a particular topic, DEVONagent can run out to the Web and gather more data that I can dump into my my database and pull into my project group.
But that’s just me. There are lots of ways to use DEVONthink Pro. 
Bill,
Thanks for this very helpful response! These past few weeks I’ve been so excited by the power of DevonThink that I wanted to try out all the features. In the end, however, this entailed a lot of fiddling with the best way to organize my data.
What impressed me about the method you sketched out so eloquently here is that the whole point of DevonThink is to free one from the time-intensive work of categorizing and labelling that is usually necessary in other databases. Perhaps the hardest thing is to get rid of old habits and preconceptions–to trust that I will be able to locate information without all the classification work up front. For a recent research and writing project I used FileMaker Pro–but the only way to get good results in that program was to spend a lot of time categorizing every new entry. With DevonThink, on the other hand, even raw web clippings will accessible when (and if) they are needed. Why then waste time fiddling with information beforehand?
One tip of yours seems especially helpful: organize data as needed for a project. Right now I find myself in between major projects. I am simply gathering information that seems interesting as I explore possibilities for my next project. With previous data management systems (both paper and computer) I had to organize this data right away (by subject, keyword, etc.). Failure to do so meant huge amounts of work down the road. But if I’m reading you correctly, I have the luxury of simply dumping interesting information into DevonThink, knowing that I can eventually use the powerful search and classification features to organize the data as my research agenda becomes more clear. If this is true, it would be an immense relief! The problem with traditional database/storage solutions is that one had to create a comprehensive filing hierarchy and/or tagging system before importing too much information. But what if I could create hierarchies and classification as needed, without worrying too much about the entire database… This seems to be what DT is offering.