Forum Posts Feed
Journaling, a simplistic view
< Next Topic | Back to topic list | Previous Topic >
Pages: < 1 2 3 4 5 6 7 > Last ›
Posted by MadaboutDana
Jul 22, 2018 at 01:21 PM
Okay, that I didn’t know - thanks for this rather valuable information! I shall check it out!
Dellu wrote:
>NO, you don’t need to run Devonthink to get your files. You can open the
>package (“show package contents”) in the Finder and find your files
>organized by file type.
>
>Devonthink database is far from proprietary.
>The same is true with MacJournal. All the files can be rescued using
>“show package contents” in finder.
>
Posted by Paul Korm
Jul 22, 2018 at 07:59 PM
DayOne entries are stored in JSON. That’s not really “proprietary”—no more than XML is “proprietary”. JSON and XML are formatted based on configurable, public, well-document and open source standards, both of which are easy to read, though JSON more readable than XML. It is true that it takes effort to transform JSON or XML to what most people would consider readable text, but tools exist to do that. Also, DayOne offers the options to export some or all entries as PDF, HTML or plain text.
Having spent years discussing disasters with DEVONthink users, I strongly recommend to NOT open the package and muck around with the internal folders. That can easily lead to lost data and in the extreme a pretty bashed up database. Believe me, people who say “never me’” frequently end up doing that to themselves. And there’s no reason to do that. To get a file, open the client, and drag it to the filesystem somewhere. People way, well maybe DEVONthink will stop working someday. Really? The likelihood that the app will just drop dead someday is, IMO, slim to none. And, if it did, then the option to trash a database by mucking around inside the package will then no longer matter. So—client first, mucking as last resort.
Dellu wrote:
>Other genuine proprietary packages like Day One don’t have that option.
Posted by Amontillado
Jul 23, 2018 at 01:11 AM
As a break glass in case of emergency option, direct file access is nice.
I wouldn’t modify anything, but you can copy the files from within DEVONThink without specialized tools.
Scrivener is really nice in that regard. Their format is simple enough to reverse engineer, but you don’t have to. Full developer documentation is available on request.
Posted by Amontillado
Jul 23, 2018 at 05:17 PM
A possibly interesting sidelight - DEVONThink definitely doesn’t hide the real location of the files in its databases.
This morning I pondered the idea of using DT as a Scrivener alternative. I could put a bunch of document files in a group, set the sort to “unsorted”, and order them as needed.
I discovered if I select all the files in a group and hit command-C, it copies the full, real, path to the files to the clipboard.
In the case of replicants, you still get the full path to the real file (or directory, in the case of a package stored in DT).
I could use Nisus to write chapters, set them in the proper order with drag-and-drop, and cut and paste the file names to a script that would run pandoc.
My thinking is this sort of manipulation isn’t a problem, if you are confident in your methods, the implications of what you’re doing, and you don’t blame me if things go pear-shaped.
Backups are a good thing, too. ‘Course, Time Machine riffles through DT databases and accesses all the files. At some point, you have to trust something outside of DT.
Posted by Dellu
Jul 23, 2018 at 07:09 PM
Paul Korm wrote:
>DayOne entries are stored in JSON.
I know it exports in Json format. But, I didn’t know it stores the database in Json. thank you Paul.