My DevonThink set-up

I use DevonThink(DT) for my personal knowledge management. DT is extremely powerful and can be used in so many ways. I thought I would offer my simple set up as an example of how I use DevonThink and hope people find it useful.

A good place to start is how I divide my data between databases.

Deciding on databases

It has taken a bit of trial and error to get my databases right, but data is easy to move about in DT on the road to getting it right. The key elements on deciding how to split my information were:

  • Do I want to sync the data?
  • Is it an ‘area’ of work where I need everything together?
  • Does the data share a common tagging schema?

Databases I have include:

  • Work (used for all work stuff)
  • Birds (holds lots of notes about my bird watching hobby - the feathered ones)
  • Newsletter (I write a monthly village newsletter)
  • Village Hall (I am on the management committee)
  • Reference (reference materials and personal stuff like car insurance documents, purchase receipts, vacation info, manuals, guides)
  • Thoughts (collection of quotes, snippets, and ideas)
  • Cookbook (collection of recipes)

Yes, I could have all in one giant database broken down into major groups as above, but the tagging schemas are different in several databases (see below) and the schemas make sense when in that database. Groups in the databases are generally only two or three deep, and I try to make group names understandable and as unique as possible. It stops (for me) lots of groups with the same name being confusing in a search list.

Do I want to sync the data?

Yes, I know I could sync everything everywhere, but why, when I don’t need it on my iPhone or iPad? There is plenty of room, but why sync gigabytes of data when I don’t need to? I don’t sync Work to my devices as it contains very confidential information. It could be safely encrypted etc. but it is far safer not to have it on my iPhone in the first place. I want Cookbook and Reference on my iPhone for ideas when shopping and insurance details if I have a car accident. Newsletter stays on my MacBook as that is the only place I work on it.

Is it an ‘area’ of work?

An ‘area’ of work is what I am going to be doing when I open my MacBook, drink a coffee, and get going. Work pretty much speaks for itself. When working on the Newsletter, I only need all the newsletter information, like past issues, details for publishing, ideas for future issues, etc. You might argue that Birds could go into Reference, but I have a specific tag schema relating to bird classification and it gives me a classification view from the tag list without lots of other tags messing it up. Another consideration on grouping of data is if I want to use replicants (they only work within the same database)

Does the data share a common tagging schema?

I am a bit OCD and hate messy tag lists. Tagging in Work relates to the type of document (proposal, report, agenda etc.) and its status (final, draft, in review, waiting on someone). Tagging in the Newsletter is the month and the information’s source. For my Cookbook, the schema is ingredients. I don’t want to have pumpkin next to proposal in my tag list, which is why they sit in different databases. For this reason, I do not unify tags but have them listed in each database only. Another reason for not unifying tags is both Village Hall and Work have an agenda tag. Unifying, I got two agenda tags, and I was never sure which was which. When working in a database, it is easy to remember the particular tag schema for that database. Not keeping to a tag schema ends up being ‘garbage in, garbage out’ and having difficulty finding anything based on tags. Was it tagged vacation or holiday?

Workflow

Now I have my databases, how do I get information into them and process it?

The Global Inbox is my clearing house

All new information, whether from my iPad in DevonThink to Go (DTTG) or DT on my MacBook, gets sent, shared or dragged into the Global Inbox. Smart rules based on a tag then move it into the Inbox of the appropriate database for processing when I am working in that ‘area’. I can add these simple tags in the sorter when clipping, when saving in finder, when sharing to DTTG, or in the Global inbox itself.

Simple Smart rules

Simple smart rules look for the simple tag and move the item into the correct database, deleting the tag as soon as it is moved. So tagging an item newsletter moves it to the Newsletter Inbox. Similarly, cookbook into Cookbook.

The local DB Inbox

When I am working in an ‘area’ and have that database (and usually the Reference DB) open, I will further process anything in the database’s Inbox. Workspaces can be your friend here for opening what databases are needed (or just have the lot open). If you put a database in favourites and it is closed, clicking on the favourite will open it.

Items in the database’s inbox might be further tagged, moved into the right group (using the move shortcut ctrl-command-M or using ’see also & classify), or discarded to stop the garbage pile getting too big. On some databases, like Work, I have a few additional smart rules that try to auto tag based on the file name and/or contents when the item arrives in the database inbox. This allows for targeted rules appropriate to that database. For example; if the name contains ‘plan’ it is tagged plan. I did get excited with trying to automate as much as possible based on content, but got a lot of ‘false positives’. A report from someone might reference a ‘proposal’ and got tagged proposal when it wasn’t. Sometimes, human oversight is the fastest way to do it right.

There are lots of examples on their forum where people use rules to rename bank statements and such things when scanned.

My end of year review

Over the Christmas period, with a beer in hand, I will go through some of my databases (particularly Reference) and decide what can go. I don’t need the car insurance documents from three years ago or the manual for the old fridge. Doing the occasional review is good for keeping the garbage in check. It makes searching more efficient if the results list is not filled with irrelevant stuff. I think of my databases like my house. I don’t want them to look like the worst hoarders home on TV, with piles of old newspapers and junk everywhere.

In Summary

Overall, my workflow is simple. When I have an item and save or share it, I add a tag for the ‘area’/database it needs to end up in. When I am next working in that ‘area’ the item(s) are sitting in that database’s inbox, waiting for final review and processing. The final review stage doesn’t take any time, ensures everything is correct and, importantly, makes me think “do I really need this?”