Hi Jim: I want to thank you & everyone else that responded & shared their valued wisdom. It is greatly appreciated!
I plan to use DEVONthink & not to follow the crowd. I ordered Taking Smart Notes with DEVONthink by Kourosh Dini & Take Control of DEVONthink 3 to start my journal.
Thanks again to everyone that shared their thoughts & wisdom!
3 Likes
Youāre welcome!
We are fortunate to have a helpful group of people on our forums, people with a broad range of experience and perspectives to be drawn from. 
4 Likes
I can imagine! Backup, backup and then backup. Sorry but we have all done it, or not done it. I am very harsh on colleagues and so on when āsodās lawā kicks in. That one time you forgetā¦
2 Likes
Great we all have very idiosyncratic usages. That is why some of us are not able to be very helpful sometimes. There are features on DEVONthink 3 I only find out about after years and never use even then.
3 Likes
Considering the cost of data migration, consistency and coherence, Devonthink Pro is the best choice. It has extremely high freedom and scalability, without proprietary formats to restrict data in and out. More importantly, the buyout model gives me long-term use guarantees. Of course, Devonthink Pro cannot optimize all problems, so I will use many other tools as processing entry points, and then index the generated files to Devonthink Pro. Devonthink Pro can meet your needs, and its advantages will be more obvious when your data volume increases.
4 Likes
Philosophically and tugging thoughtfully on my virtual beard, linking is extremely important.
In my personal use, tags do the heavy lifting.
4 Likes
I now run the risk of incurring the anger of some people.
Linking notes is a tool, like a screw driver. Great with screws but pretty useless with nails.
For example; I have a cookbook database and it makes sense to cross link my markdown recipies for sides that go with mains etc. (I use tags for ingredients to aid finding a recipe for that pesky zucchini left in the fridge). Similarly, I have a work database full of word docx, xlsx, pdfs, markdown meeting notes, etc. linking here makes no sense (though extensive tagging, groups, auto filing do).
FYI: I maintain everything in Devonthink. I use standard recipe templates and the filing power of DT as well as lots of its other strengths. However, my cookbook is indexed to an Obsidian vault as I have a simple uncluttered Obsidian preview on my iPad for when I am actually cooking.
Use a screw driver with a screw and a hammer with a nail.
Follow Bluefrogās and others sage advice:
Decide for yourself based on how you think and work with your data.
4 Likes
Does that provide any advantage over searching for zucchini?
Yes and no. Searching returns everything zucchini. With tagging, I can be selective about what I want to include and it is quick to see the list of all my preferred zucchini recipes under a tag in the side bar. Just the way my brain works as historically, I came from the dark days of early Windows when only folders were used for organisation. I moved to the Mac bright side since.
Devonthink will do almost everything. You may need to season the mix with Keyboard Maestro or add additional applications like spreadsheets or mind maps, but with a will you wonāt be blocked.
Iāve lost data with other systems. As best I can tell, and I do things like weekly verify and file checksum tests, Iāve never lost anything in Devonthink. Itās a battleship!
3 Likes
I think one day when I am bored and feel like starting a war, I will open a poll in the forum about how people sort their files, ask whether Johnny Decimal is better than PARA, then sit back and let chaos reign.
Like many wise people on this forum before me, Iāve just taken the bits I liked of the systems I read about and made it mine. I donāt mind showing my system, but youād have to think and work like me for it to be of value. We might think the same, but itās ok if we donāt!
(And I think the PARA method is a gimmick that doesnāt reflect modern work for a lot of people 
But I know other people feel differentlyā¦)
14 Likes
I agree with you. I see The PARA Method all over on social media & everyone is giddy over the system. I tried it, but it was not a fit. I kept thinking to myself, what am I missing?
. Everyone craves PARA but not me! 
Well, I do not want to search and I do not want to create (many) links.
I want to find.
Therefore, for each file I refer to, I have an identifier that is as unique as needed, but short as possible. Ideally, I get exactly one search result with some predefined searches.
I just had a look at PARA. I think I use only P and R. P for notes (projects indexed by Devonthink) and R for documents (resource database in Devonthink).
1 Like
I use the unique identifier (UUID) assigned to each record in Devonthink
Example: C676D7BC-8CD7-406F-9A82-C1D440930AA9
3 Likes
Sure, but I was thinking of patterns that I can remember.
The uuid is more or less the link.
By the way: Is the UUID v4 and completely random (which I assume) or does it include a timestamp?
This has been an interesting thread about organizational workflows. I suppose a lot of apps or workflows with a cluster of apps can do just about āeverything,ā depending on the hoops you want to jump through to make it happen or the compromises you are willing to make. I think the answer to your question might be āYesā in this forum and in the forums of other products (Google stuff, Microsoft stuff, etc.). I donāt see a necessity to use just one app, but I find it convenient to rely on DT for most of my note-taking and organization, leaving some work for apps that specialize in one area or another (Word for formatting things, PowerPoint for presentations, BBEdit for editing text files, etc.).
For example, DT and Obsidian overlap with some features, but Obsidian has really nice ones DT doesnāt have, and DT has really nice ones missing from Obsidian. Why not use both? DT is agnostic in this regard and generally plays well with other products out there, letting users adapt it to their individual workflows. Obsidian is something Iāve played around with before and I appreciate that they appear to have great security. Logsec? Looks promising, but I donāt understand how secure it isāI cannot find it clearly stated anywhere that there is end to end zero-knowledge encryption, so maybe it is not there. Same for Tana. For me, if an app isnāt committed to zero-knowledge end to end encryption, I will try to find an alternative.
As far as organizational systems go, I keep trying and trying to find something better than the one I use, but nothing quite seems to be an improvement (the Noguchi filing system + quasi-bullet journal + relying on search to find things). They all promise something great, but I havenāt become a convert yet. I just keep plugging along with my workflow, which has somehow gotten me this far. PARA? Maybe it works for some folks. It doesnāt seem like a solution to my problems. CODE? I donāt know it. Iāll take a look around.
5 Likes
Thank you for sharing your thoughts & wisdom!
I hit on something this morning.
The idea is I wanted to view a markdown file in DT as a map of contents in source mode and as an aggregate document in preview.
If you copy an item link, you can paste it a couple of different ways.
Command-v in a DT markdown document pastes the link something like:
[Jeff Martin](x-devonthink-item://313366DC-7F8F-4C08-A8E7-B51A57D3F68C)
From either source or preview mode, you get a link you can click on. I wanted a way to see it as either a clickable link or a transclusion.
Enter BBEdit and Keyboard Maestro.
I created a BBEdit Text Factory that would transform that clickable version of a document link to something like:
Jeff Martin:
{{x-devonthink-item://313366DC-7F8F-4C08-A8E7-B51A57D3F68C}}
Cool. In source mode, I can click on the link and navigate to the cross reference. In preview mode, I see the file as a transclusion. Map of contents in one view, dynamic document in another.
If thatās of interest, hereās the nitty-gritty.
The BBEdit Text Factory is a single step, a Replace All operation.
In the BBEdit Options for that step, I set the Grep option. The Search for field contains this bit of reasonably incomprehensible grepish:
^\[([^\]]*)\]\(([^\)]*)\)
The Replace with field contains this bit of sheer poetry:
\1:\n{{\2}}\n
Hereās what that looks like in BBEdit:
Next, I created a macro in Keyboard Maestro:
While Iām editing a markdown document in DT I can right click on another markdown document and choose copy item link.
Then Option-Command-t will paste the other documentās name and link as the documentās name followed by its link in double curly braces. In preview mode, itās now an aggregate document. In source mode, itās a map of cross references.
In other words, Devonthink for everything, particularly with a few helpers like KM and BBEdit.
1 Like
Basically, youāre running a regex replace via BBEdit. Which you could also do via a simple JavaScript script in DT or even KM.
In DT:
(() => {
const app = Application("DEVONthink");
app.selectedRecords().filter(r => r.type() === "markdown").forEach( r => {
r.plainText = r.plainText().replaceAll(/^\[([^\]]*)\]\(([^\)]*)\)
/g,"$1:\n{{$2}}\n");
})
})()
This script works on the record(s) currently selected in DT, ignoring those that are not MDs, and changes their content. It uses the same RE as you did in BBEdit. Iād probably amend that to use non-greedy matches (?*) instead of negative character classes, but thatās just a preference.
Note I didnāt test the script, just scribbled it down. Please run it only on copies of your data until you know it does what you want.
Not saying that BBEdit is bad or anything, it just seems to be a bit convoluted to start up an editor only to run a regex replace.
6 Likes
Nice! Many thanks.
Iāll try that. One thing Iād like to adapt into your script would be to act on only selected text, or on the clipboard content.
I use a KM macro to create markdown files with a link to the file itself at the beginning of the file.
That sounds quite useless, but if you are viewing an aggregate of several documents transcluded into one, itās handy to be able to navigate to any component document for updates.
I really need to get friendlier with Javascript in DT.