Optimal combination of DT and BE?

A place for users to ask each other questions, make suggestions, and discuss Bookends.
Post Reply
jzents
Posts: 50
Joined: Fri Oct 26, 2007 11:04 am
Location: Edinburg, TX

Optimal combination of DT and BE?

Post by jzents » Mon Sep 01, 2014 10:46 pm

This is not a question about future possible interactions between the two programs, nor is it a feature request. Rather, I am asking those of you that already use both DEVONThink and Bookends together what you have found to be the optimal way of parceling out your work between the two applications. Do you duplicate your PDFs in both bookends and in DEVONThink? Do you mainly keep your PDFs in DEVONThink? Where do you keep your annotations on a particular book or journal article? Do you mainly keep them in Bookends and then duplicate them to DEVONThink or do you mainly keep them in DEVONThink? Thank you in advance for sharing and your time! Cheers!
Jeff Zents
South Texas College

terrydev
Posts: 15
Joined: Sun Aug 31, 2014 4:23 am

Re: Optimal combination of DT and BE?

Post by terrydev » Tue Sep 02, 2014 11:27 am

Ph.D in the Humanities candidate here and this is my workflow, fwiw.

I divide the workflow into two separate sections.

1. Managing the pdfs/epubs/whatevers and citations
2. Processing the information

I use BE to manage the pdfs and the citations. When I acquire a source I import it to BE and make sure the citation is correct. In this process the pdfs are moved to a folder within my PhD folder (that's my BE attachments folder) and renamed for Author, Year) I have the attachments folder within the Ph.D folder for back up purposes and also it allows me to index the folder with DTPO. So, one set of files, managed by BE but indexed by two apps, each with their own special characteristics. When I import a source I keyword it. As DTPO can display it's keywords in a separate window (Tools -> Show Groups and Tags), I can see those in any other app I use, so there is a continuity - I'm not using one keyword in BE and another in DTPO for the same job.

Obviously I use BE to cite as I write and so on.

Processing the information for me means when I read and annotate a source. The first thing I is create a Folder (or Group) in DTPO. I then replicate the pdf to that folder (as it's already indexed by DTPO). Then I start to read - most often I use Preview for pdfs - but I also use Skim, DTPO itself and BE on occasion - and Murasaki for epubs. (This as it's the only reader that doesn't use a Library, and so can access the BE attachments folder).

Here is where the fun starts:

If I find a passage I want to quote and annotate I have a Keyboard Maestro workflow that I invoke. This runs as follows:

1. I select the passage I want to quote and invoke the KM workflow. This does the following:

a: It copies the passage.
b: Highlights it
c: pops up and dialogue that asks for
i. A name for the note
ii. The page number
iii. My comments on the passage
iv. (up to..) four keywords to tag it
v. The source citation
d: I click ok and it switches to DTPO
e: creates a new markdown file there in the group
f: names it with the page number and the name from the dialogue
g: pastes in the quotation
h: pastes in my comments
i: pastes the citation
j: adds the keywords
k: saves

A couple of things: I name the note carefully. I bear in mind that it could be three years before I scan this title again. So, if I'm annotating Goldilocks and the Three Bears I won't name the note 'Porridge!' but instead something like 'She tries the porridge'.

So in the list view in DTPO I get a name like

@45 She tries the porridge.

I prefix with the page number as it means that I can list the notes in sequence in the view.

Four tags? Could be 5, could be 3. If it's more than 4 I can add manually later.

The Citation only needs to be added once - KM remembers it until you change it again, but the cite goes in every note. Why? Because if I search on keywords then I can find notes but have no idea where they are from.

I use markdown as it's text, and that's the most absolutely portable thing ever.

In the end I have a searchable database of all my sources and all my annotations of all their content.

I don't use the annotation feature in BE mostly because my system predates it rather than anything else. DTPO has a feature to create an annotation from a BE citation but I don't use that as it creates a single file, and bearing in mind I might have hundreds of notes on a book a single file is useless. For books I often use sub-folders for individual chapters.

What I don't like about it? It has an amount of friction. I have to input 7 bits of data into that dialogue - but I've not found a less coarse way to do that. I also feel a need to double check the KM workflow to make sure it has behaved correctly. I have found on occasion that it glitches. But with two screens I see the source and DTPO side by side.

Hope that helps.

jzents
Posts: 50
Joined: Fri Oct 26, 2007 11:04 am
Location: Edinburg, TX

Re: Optimal combination of DT and BE?

Post by jzents » Tue Sep 02, 2014 4:00 pm

terrydev wrote:Ph.D in the Humanities candidate here and this is my workflow, fwiw.

I divide the workflow into two separate sections.

1. Managing the pdfs/epubs/whatevers and citations
2. Processing the information

I use BE to manage the pdfs and the citations. When I acquire a source I import it to BE and make sure the citation is correct. In this process the pdfs are moved to a folder within my PhD folder (that's my BE attachments folder) and renamed for Author, Year) I have the attachments folder within the Ph.D folder for back up purposes and also it allows me to index the folder with DTPO. So, one set of files, managed by BE but indexed by two apps, each with their own special characteristics. When I import a source I keyword it. As DTPO can display it's keywords in a separate window (Tools -> Show Groups and Tags), I can see those in any other app I use, so there is a continuity - I'm not using one keyword in BE and another in DTPO for the same job.

Obviously I use BE to cite as I write and so on.

Processing the information for me means when I read and annotate a source. The first thing I is create a Folder (or Group) in DTPO. I then replicate the pdf to that folder (as it's already indexed by DTPO). Then I start to read - most often I use Preview for pdfs - but I also use Skim, DTPO itself and BE on occasion - and Murasaki for epubs. (This as it's the only reader that doesn't use a Library, and so can access the BE attachments folder).

Here is where the fun starts:

If I find a passage I want to quote and annotate I have a Keyboard Maestro workflow that I invoke. This runs as follows:

1. I select the passage I want to quote and invoke the KM workflow. This does the following:

a: It copies the passage.
b: Highlights it
c: pops up and dialogue that asks for
i. A name for the note
ii. The page number
iii. My comments on the passage
iv. (up to..) four keywords to tag it
v. The source citation
d: I click ok and it switches to DTPO
e: creates a new markdown file there in the group
f: names it with the page number and the name from the dialogue
g: pastes in the quotation
h: pastes in my comments
i: pastes the citation
j: adds the keywords
k: saves

A couple of things: I name the note carefully. I bear in mind that it could be three years before I scan this title again. So, if I'm annotating Goldilocks and the Three Bears I won't name the note 'Porridge!' but instead something like 'She tries the porridge'.

So in the list view in DTPO I get a name like

@45 She tries the porridge.

I prefix with the page number as it means that I can list the notes in sequence in the view.

Four tags? Could be 5, could be 3. If it's more than 4 I can add manually later.

The Citation only needs to be added once - KM remembers it until you change it again, but the cite goes in every note. Why? Because if I search on keywords then I can find notes but have no idea where they are from.

I use markdown as it's text, and that's the most absolutely portable thing ever.

In the end I have a searchable database of all my sources and all my annotations of all their content.

I don't use the annotation feature in BE mostly because my system predates it rather than anything else. DTPO has a feature to create an annotation from a BE citation but I don't use that as it creates a single file, and bearing in mind I might have hundreds of notes on a book a single file is useless. For books I often use sub-folders for individual chapters.

What I don't like about it? It has an amount of friction. I have to input 7 bits of data into that dialogue - but I've not found a less coarse way to do that. I also feel a need to double check the KM workflow to make sure it has behaved correctly. I have found on occasion that it glitches. But with two screens I see the source and DTPO side by side.

Hope that helps.
Wow, that is very detailed and helpful in that it gives one (plus hints on others) very detailed approach. You have given me lots to think about. I really appreciate the time you put in to tell me this. I will have to consider all of the points you have made. Thank you for you time and thoughts. I will also have to look into KM. Cheers!
Jeff Zents
South Texas College

jtth
Posts: 5
Joined: Tue Sep 16, 2014 12:37 pm

Re: Optimal combination of DT and BE?

Post by jtth » Tue Sep 16, 2014 12:45 pm

Is there any way you could share that workflow?

terrydev
Posts: 15
Joined: Sun Aug 31, 2014 4:23 am

Re: Optimal combination of DT and BE?

Post by terrydev » Tue Sep 16, 2014 4:29 pm


gnoli
Posts: 32
Joined: Wed Feb 28, 2007 4:51 pm
Location: Ravenna, Italy

Re: Optimal combination of DT and BE?

Post by gnoli » Fri Sep 26, 2014 2:41 pm

I used a workflow similar, even though without macros etc. and all went very well. But now my attachment folder is HUGE (more than 10.000 items scanned in many many years), so I was obliged to abandon this workflow, because DTPO cannot manage such an amount of data. Now I am searching a solution and I would be grateful for any suggestion ....

terrydev
Posts: 15
Joined: Sun Aug 31, 2014 4:23 am

Re: Optimal combination of DT and BE?

Post by terrydev » Fri Sep 26, 2014 2:56 pm

You can have unlimited DTP databases. I have 5 for instance. But IIRC the number of items in a db doesn't really matter, it's the number of words indexed. Bit that's more one to ask on the DevonThink forums :D

drsoram
Posts: 30
Joined: Fri May 02, 2014 2:34 am

Re: Optimal combination of DT and BE?

Post by drsoram » Tue Nov 04, 2014 3:13 am

Thank you for the KM macro!

Soram

cinepou
Posts: 15
Joined: Thu Aug 27, 2015 10:26 am

Re: Optimal combination of DT and BE?

Post by cinepou » Sat Aug 29, 2015 5:45 pm

Late to the conversation, but I have to say I'm in awe of terrydev's workflow, and will try to emulate it. Thanks to the KM macro he's posted in this thread, this shouldn't be too difficult.

I've been trying to accomplish something very similar these past couple of days on the DTP trial version (without macros a rather tedious process). Happening upon this simple and elegant solution has definitely swayed my decision to make the purchase. I think I can speak for all fledgling PhDs, in the humanities and elsewhere, when I say: this should be stickied or something!

However, to make it work, I'd also have to get Keyboard Maestro. Before I go ahead and do just that, one question (with a view to my limited budget): has anybody managed to successfully implement this workflow in the current version of DTP?

Any pointers are appreciated.

cinepou
Posts: 15
Joined: Thu Aug 27, 2015 10:26 am

Re: Optimal combination of DT and BE?

Post by cinepou » Sun Aug 30, 2015 6:44 pm

I just realised there's a trial version of Keyboard Maestro. I'm a little embarrassed to ask, but after a week of getting into both BE and DTP I'm lacking the willpower to learn the intricacies of yet another app: would anyone kindly give me some quick advice on how to implement terrydev's KM script? :oops:

cinepou
Posts: 15
Joined: Thu Aug 27, 2015 10:26 am

Re: Optimal combination of DT and BE?

Post by cinepou » Sun Aug 30, 2015 7:42 pm

I managed to make it work after all, sort of. I've successfully imported the script to KM, and when I invoke the script via the hotkey I assigned to it the dialogue window does indeed pop up. But after I've filled in the various fields (title, page number, comment etc.) the script, rather than creating a new markdown file in the DTP group that encloses the highlighted PDF, initiates a rather weird sequence of commands: it sucks the open DTP window into the dock, and when I maximise it again, the highlighted text appears as the suggested name of the PDF (i.e. if I were to press enter now, the PDF would be named after the highlighted passage).

New request (thanks for bearing with me): anybody know how to fix this? I think I've followed all of terrydev's instructions, but of course I might have overlooked something...

Post Reply