Bibliographic information and work flow

DEVONthink Feedback
1

In my research I downloads dozens of articles from journals. Many are not searchable PDF’s and I have appreciated DT Pro’s ability to OCR these puppies and nest them into collections. I am better organized thanks to DT Pro.

However, I am overborne by the work of retyping the bibliographic information. Because I use DT Pro both as a repository for my articles as well as a way to think, outline, and write I see clearly that the thinking side of DT far outshines the repository side.

I have been drawn to iPaper and Papers ability to automatically download the meta-data of author, title, journal etc. from Pubmed and other databases. I am ready to jump ship for the collecting and repository side of my research. I already rely on EndNote to download bibliographic data and wish DT Pro had some way to handle bibliographic data.

I had thought it would be possible to create a sheet with this information in DT for all of my articles but this would be more labor.

Maybe I have missed something but there is no reference to Bibliography in the user forums. I find this pretty curious and must admit that I purchased DT Pro with the intention of finally organizing the gigs of research papers I have collected over the years.

What is the recommended workflow for keeping Bibliographic information in DT?

2

This is a very interesting subject.

Personally, I use BibDesk as a bibliography manager. It has a lot of very handy features, as autofile, search into Skim annotations, etc. Papers is also very cool. But in my opinion, it has the drawback of being too centered into the scientific papers workflow (very interesting in ‘hard’ sciences) but not very useful sometimes in social sciences and humanities (we still work with books, non indexed articles and stuff).

The workflow solution that I recommend is the following : Use Bibdesk or Papers for your papers library. They are specialized solutions to the problem of organizing and dealing with scientific papers. You will never have the same features in DEVONThink (DT).

What DT can do is index your papers library folder(Bibdesk or Papers). I use DT to find relationships (thanks to its AI) between papers. You don’t need the fields of your database because what DT does is to find content relationships. It is very useful. With Bibdesk, what I do is to export my papers database as RTF (using a custom and beautified –I’m a visual designer also– RTF export template) and keep it updated into DT. That will help DT to find relationships between things.

What is cool in DT is that you don’t really need to replicate the structure of your bibliographic database, you just need to have the content indexed in a appropriate semantic manner. Generally Bibdesk or Papers allow you to store your papers in the Finder in a semantic manner. So you index that library and then you could use replicates to “think” into outlines or projects (even if for me, replicating is a pain because that horrendous red color of replicates in DT hurt my eyes. I think it is the bigger UI error into the big list of UI errors and oddities of DT).

By far, the workflow works very well from Bibdesk/Papers to DT. One thing missing (we still wait for DT2) is a way to do the contrary: let Bibdesk or Papers have links to your DT documents or notes. This could be done by simply having an URL for each item of your DT database (applications that have this feature : Yojimbo, Journler, Together, Omnifocus, Mail.app, Papers, etc.). Unfortunately, DT isn’t yet a very good citizen into the Mac OSX 2008 city of applications.

As for the writing part of the workflow, I recommend using Scrivener. Apart its already known features, Scrivener supports “multimarkdown”. Multimarkdown is an extension of the already known markdown format, but it will let you write LaTEX using a more intuitive markdown. So, if you have a Bibtex bibliography, you could always integrate it with Scrivener. This is a very nice feature.

I hope this will help you.

3

Though they are not particularly beautiful :slight_smile: , DTP handles .bib files as sheets, so that you could export your Papers (the application) library as a .bib file and import it, or simply index your BibDesk library file.

(actually, its handling of ‘Annote’ field is downright ugly; radii0, could you share your template? )

4

Thank you for the thoughtful and complete replies. I must admit to getting old and lazy. I remember 5foot long boxes of index cards in the reference room of the graduate library when I was a pup. And computers with room size appliances that you stood in line for and handed over a stack of punched cards to run in a batch later in the day.

I will likely continue working with Endnote, now on version X1 for the Mac. It has the benefit of a large enterprise keeping the filters current and because I am largely University based, most academics I know use it. (With the exception of the furry Unix/Latex hobbits I admire but lack the technical facility to emulate.) Endnote allows me to associate files with the entries and I will continue to use DTP to index articles, ocr, and manage their inter-relations.

Thanks for the Scrivener tip. It is a lovely environment and I am just finding my way.

But for future development I don’t see that it would be that much of a stretch for DTP to have a functional front end like Papers. Those are some enterprising lads and you have to admire how honest they are presenting the programs and libraries they have emulated and cobbled together.

Bruce