I never suggested that.
You could easily make something with AI. Or lots of other basic consumer software. Or just sketch it by hand.
It’s a whole lot easier to understand and consider a UI suggestion if you show what it will look like.
That in turn might encourage alternate suggested mockups. And that in turn may lead to discussion about the pros/cons of each.
For what it’s worth these are AI mockups courtesy of Claude:
https://claude.ai/public/artifacts/0c7b6adc-5b80-4e92-98b4-be3311effa02
The easiest way is to listen to what is already on the table.
See either my points. Or the long sketch provided by @anton .
What is happening instead:
a) silence on those points (largely, in particular by team)
b) lots of meta-comments from people w/ no visible interest in the topic of the thread, or its 2nd part (UI/UX around annotation)
c) raising the “ante” for ordinary users – those present, and all those reading along – w/o any signalling of real interest in what is said.
This is not discussion. This is a complex pattern of shielding and deflection. And Ingroup formation at its finest, btw.
Now, with the last turn (occult insider signalling; censoring arguments) it has even degraded lower.
I really have no interest in this pretense of “discussion”, really masking exclusion dynamics which are very obvious – and now even expressed explicitly here.
(– For people interested: alongside my own reading along as someone knowing the mechanics of discourse and positioning in communication, I have an interesting “meta-analysis” of this thread, done with some non-partisan prompts and focussing on communication dynamics in general and forum dynamics in particular. An interesting – and depressing – read. But reason enough to step away, and into more healthy and genuinely interested communication spaces).
Thanks for your proposal, though.
I hope it will not keep or distract anyone from replying to @Anton in earnest (in case of interest in the real topic of the thread, or more exactly it´s second “UI/UX”-part.)
Appreciate this constructive signal.
I know some – even some present in thread – are likely more sceptical as to (direct) use of “AI” for such things.
I can relate to using such tools, if done with a view helpful in such a complex process as UI/UX-architecturing (– which for a reason is a science, profession and trade).
So, curious how this might be taken up as part of the process.
Then,some notes of caution as to the expectations re. mockups from user side, esp. in such early stage discussion which haven´t even progressed to the fundamental stage of problem identification/definition. Not to talk of team processes, and fit/adaptation to these in such complex processes as UI-revision (which IMU is not even on the table here).
Still curious to see whether this solicits constructive engagement…
PS – after studying your mockups. I think this is already right in the direction discussed. Not sure people/team share the take – but it certainly reflects some central things which has been said/raised here. – One thing not in there, IMU: if one treats all annotation-(aside-)note aspects as truly part of one functional complex, there is one thing that additionally belongs in such UI-musing: the scenario where one finds the aside-note independently of the “parent”-document. – So, in your version this should/could also go into the Document Inspector as well as in the File List. (To give an esspecially interesting scenario/context: annotated images/media! Here, it is very likely to first find the annotation (not the parent) bec DT-search is obviously text-focussed. Then, especially in this context (images/media), such additional textual aside-annotations make a lot of sense, bec it (indirectly) feeds the document into the DT-core processes (like “see also”, concordance etc.)
Can I suggest not overthinking this?
I don’t think we need a detailed academic discussion of the theory of UI design.
If any of the concepts would be of interest to you then say so.
If you have a tweak to suggest then go ahead and show us your alternative.
If you think the suggestions are off base, then give us a couple samples of what you would suggest instead. I am quite sure that AI mock ups or mock ups from some other software or handwritten sketches would all be considered equally.
All fairly practical and very simple. For sure a picture is worth more than 10 posts of theory.
I don´t see a note of cuation as to the value of mockups – as they suddenly seem such a central criterium/tool of choice (for some) – as overthinking 
. It was nowhere “academic” (IMO), but related to usefulness in real processes of UI/UX feedback (that is the link, if you refer to that).
Also forums normally do not function by ‘I show you my mock up – show me yours´. It´s possible, but not really the one (prescriptive) way…
– … and if you really looked at the link you found plenty of arguments pointing to the fact that visualization before (shared) analysis does not neccessarily “make things easier” in the process, or that an image positing a solution is “saying more” than a one sentence statement of pain points.
Though, if it´s best medium of expression for you: fine & well!
Then, on the non-”theoretical” side (topic): personally I have said everything I have to say as addendum to your visualization of parts of what has been said (see the “PS” – the much longer part of that last post). Why not continue there?
– Or let others speak – as I do not see this as personal onus at all…
… so, before starting everyone doing mockups, let´s maybe first see what concrete thoughts/responses these, based on prior input, generate
I’m afraid this is what I see as well. It happens all too often in the DEVONthink forums. No need to go into detail, but reading from the concrete critique of the interface to the cosmic speculation about implementation of suggestions, we can see what’s really going on. Disappointing, again.
Did you mean to say all forums, or even the internet in general?
On the other hand, it’s probably safe to say that the DT folks believe the current UI to offer the best balance among the many tradeoffs involved in the design. Offering an alternative solution to the problem (1) gives them an outside perspective that might not have been considered, and (2) gives the person making the proposal an opportunity to consider some of those tradeoffs themselves.
– 0) We don´t know, yet what the team thinks on the implementation in light of the criticisms/points-for-thought raised. A lot of other (proecessual, expectation-related) things were spelled out extensively, though
• 1) The outside perspective *is there*. It´s in the points I made (which can be taken up, added to, contradicted etc. pp.), it´s in the very systematic sketch that @Anton made, and all of this seemed to have been clear enough, so that @rkaplan could (with whatever he added in the process) turn it easily into visuals (via AI); so the alternative visuals are also there by now (I even commented on them on one substantial aspect, otherwise stating they represent very well a lot of what has been said).
So, I don´t see what´s missing on that side.
• 2.) see 1.
I think the most easy and natural way of dealing with things, and geting out of view here (with too many assumptions, projections, and comments as to what others should do…; or with secondary loops about expectations that might be involved like a “ghost in the bottle” etc.):
that simply everyone interested in the topic of the (potentially) powerful annotation functions and their accessability/coherence/transparency in the UI, simply states what he/she likes/dislikes, or what he/she sees as improvable in terms of their implementation. Or not.
Nothing more, nothing less. Like a normal forum.
– Like: what´s your (or anybody talking here in this thread on “annotations” in DT) position on @rkaplan visualization of the points raised earlier? What´s your position on Do you share, see ways of modifying, appreciate or reject @anton ‘s systematization/proposal (including my points)? What´s your interest/stake in the topic. Simple. And simply that. Like a normal person to person talk. Each one simply acknowledging what the other brings to the communication.
No assuming what else is going on. Or reading foreign minds. But talking points raised, and words and ideas actually uttered.
My position is that I don’t plan to open DEVONthink until after New Years, as I’m on vacation. If I decide I have anything to say, I’ll do it then.
Best holiday wishes to you and yours.
I depend heavily on Annotation Files for my work in which I read very many scientific journal articles. Having a tightly linked set of notes and key points unique to my way of thinking and priorities (and not replicated by AI) is of great assistance to me and allows me to much more easily write and reference the position papers that I write based on these. It is probably the most important feature that DEVONthink offers for my use case.
Thanks for sharing your experience and use case! Are these journals for work, study, or pleasure?
Great to hear! And how is your experience of the UI?
I use it for work as a freelance self employed consultant. Once free from the shackles of corporate IT and Windows, I got a Mac and shortly after started using DEVONthink (around 2011). I haven’t looked back since.
I have been using DT for 10 years or more. There was somewhat of a learning curve but now I am very comfortable with the UI and generally happy with it. Behind Mail and Safari it is the app I use the most, so its UI is pretty much second. nature to me at this point. I have a 14” MacBook Pro and things are a little crowded on its screen, but most of the time it is docked to two 27” monitors and DT is very nice to use when you have the screen real estate to spread out. The UI is still serviceable when the MBP is undocked, however.
I see 
Thx!
Given the illustrative thread patterns it seems save to assume that the issues/challenges around the fine asides-annotation feature/capabilities and it´s rather suboptimal UI-implementation (especially from perspective of less technically-inclined users) will neither be acknowledged nor addressed in any forseeable future.
So, for people still interested in having this powerful feature set disposable in some more compact and intuitive form for daily/habitual use (and not from the coders or tinkerers department), I post my workaround to at least have some kind “solution” as of closure for this “communication”.
It follows the hints laid out by FrankT, which means it relies on 3rd party apps and some (modest) tinkering investment, plus one probably critical ‘insight’:
One can use Keyboard Maestro (or some pendant), and put the critical parts of asides-annotations – creating, writing, finding, reading – into two shortcuts:
So, using "Click button : “Annotationen” after "shift focus to panel “Info-Panel - Annotationen” (via Tools-Inspectors-Annotations&Reminders = “^ 3”)
This to me is the only way to get the whole process around aside-annotations somewhat streamlined and sane, as it puts the core of it (creating, writing, finding, reading) behind two – additional – shortcuts.
It of course is only a workaround (not a real “solution”), especially as it doesn´t address the visual scattering of this functional complex in the UI.
But it is what it is, and here we are.
Fare well all you non-techies,-coders, and -tinkerers out there!
And enjoy the power of annotation in DT! 
//
– Particular note here to @anton , @Amontillado , and @Kevin (maybe also also helpful for @ethanw6258 and/or any lurkers here, which are into annotations and UI/flow improvements )
Woah — this thread really took off! I haven’t had the chance to read everything yet since I’m on holiday (and Merry Christmas to those who celebrate 
).
Speaking only for myself: I believe it’s healthy to express opinions about software. When feedback is constructive, there’s real value in sharing it.
At the same time, I fully recognize that the development team has spent a lot of time thinking through design decisions, and that features are built the way they are for good reasons. I also understand that it’s practically impossible to create software that fits everyone’s workflows and preferences equally well.
When I ask questions or share thoughts, my intention is genuinely to understand why things work the way they do, and—where it feels appropriate—to suggest alternative perspectives rather than make demands. I don’t necessarily expect acknowledgment (even though, of course, it’s always appreciated). For example, in the case of the “user stories” I posted earlier, the goal was simply to provide a basis for discussion rather than to present anything final or prescriptive.
I do find it a bit unfortunate that the conversation has shifted toward meta-discussion. In the end, we’re all here because of a shared fascination with DEVONthink, and I believe most people are participating with positive intentions. From a user’s perspective, open and respectful discussions can also help newcomers feel welcome and better understand the app, rather than reinforcing an “insiders only” atmosphere that sometimes emerges in technical communities.
Personally, my experience on this forum—and in exchanges with the support team—has been overwhelmingly positive. I’m still learning my way around DEVONthink, and I value these discussions because they help me become a better user, and hopefully help others along the way as well.
I’m speaking here of only the DEVONthink forums (fora?)
