“Oh, this is fun! Thanks for all the hard work. I know how much time this stuff can take. I really appreciate you putting so much effort in.”
Thank you Isaiah for your long answer: it’s really nice to have so much listening with the long post you wrote and also to better understand the stakes of your development. Oh, I forgot, I want to apologize about the quality of my writings: I am French and I use Google to perform the translation. Any excuses for any misinterpretation.
“Even if I don’t use some of this – I still love that you were thinking about this stuff and I take every bit of this in consideration. In other words, this info will absolutely become part of future designs – but probably mixed together with a dozen other great ideas in non-obvious ways.”
I am perfectly aware of that. My proposal is motivated only by the desire to think about a problem of user interface and as I still work a little under RapidWeaver and only with the Stacks (and I groan to always have to unfold / fold to have the best situation “geographical” of my interventions), I launched, without a second thought, for fun! … It’s up to you to do what you want. And in fact, it is a real challenge to have to create an “offline” editing interface drastically different from the “inline” preview … and if I had a wish, it would be that RapidWeaver offers one day the possibility of being able to edit, in PREVIEW MODE, the contents TEXTES or IMAGES … it would be very comfortable for the users … but it is another debate.
“One thing I should probably point out: building user interfaces in native MacOS apps is non-trivial. Even in a UI like this that is largely composed of CSS, it can still take months of work to align the native components with the non-native web-based components.”
I imagine, while absolutely not competent to appreciate the difficulty of your work. I stayed at the creation of interfaces, 2D vector Illustrator, imported into Flash to add interactivity, timelines, some pointers and animation! … Yes, it’s from another age, it was twenty years ago now! …
“Most of the UI work for Stacks was completed over six months ago. I’ve still been refining some toolbars and button colors, especially for dark mode – but changes as dramatic as you’ve proposed would definitely take months – so I’ll probably just have to file these suggestions away for the next big UI overhaul in Stacks 5. But… you never know… sometimes I can sneak one or two things.”
I did not know you were working on an update to Stacks - I can not wait to see what improvements you will make. Is there a way for me to participate in beta testing?
“OK, feedback. Here’s what I love:
2 - I love that you tried to keep it real and used a titlebar height that’s very close to the real height (you’re about 2px larger).”
In fact, I imposed myself to be as close as possible to the existing so as not to confuse the habits of historical users. On my model, the vertical spaces are identical to the current version of Stacks: I just added a pixel or two to the selected capsule, a story that visually it is more present compared to others. I also tried to colorize it in black so that it is even more pregnant! … It was not especially “sexy”! …
“OK, some critiques:
1 - So… 28px is the max – with at least 1px of top and bottom padding.”
I had not imagined that this height constraint was paramount in the development of your software. But maybe you could lower this constraint by the fact that the user would not have to unfold ALL these STACKS, and therefore gained height, because he would now have more information directly readable on each one. they, FOLDED!
“5 - Placing the X away from other controls is very important. It’s a destructive action so doing all we can to help the user avoid accidental clicks is a good thing.”
I confess I never understood why a function as important as the DELETE did not automatically invoke the display of a confirmation dialog! … An option that could be deleted in the Preferences of RW.
“7 - The title edit button. I kind of think this one might be better as a setting in the info panel or directly clicking on the title. Stacks 4 will put it in the info panel. Mostly because direct manipulation would mean a second click on the titlebar – which already has the meaning of editing the content.”
I was going to offer you this solution of the double click! … Moreover, there could be more opportunities with the contextual click … but maybe the development in CSS makes it difficult to put implemented.
“7 - The title edit button. I kind of think this one might be better as a setting in the info panel or directly clicking on the title.”
Where, again, a simple double-click! … More natural.
“7 - Stacks 4 will put it in the info panel. Mostly because direct manipulation would mean a second click on the titlebar – which already has the meaning of editing the content. I’d rather not over-laden the gestures with too many things if I can help it. I also think there’s room for adding other user-customizations (like color – but even more stuff) that might be challenging to squeeze onto the titlebar.
Lastly, and this is the big one…”
If I had to argue about a relative progress in the complexity of the user interface, I would say that habits are taken relatively quickly. And then, RapidWeaver and Stacks are old software today! … 
And there is also the possibility to offer two modes of interface: beginner and confirmed.
“Lastly, and this is the big one…
All of my original designs for Stacks 3 looked almost exactly like your first proposal (with circular buttons). And it wasn’t merely design. I spent months implementing it and preening the details as well. I went all the way through creating the real UI and sending it out to testers. I’m afraid the results were not very good. And that’s putting it mildly. In fact they were so bad that I had to trash the whole idea, delay the project 3 months, and start again building out different toolbar ideas.”
;-(
“To be honest, I was totally crushed. It seemed like such a great design. It looked amazing. It created all kinds of new opportunities for directly manipulating each stack. And it eliminated so many toolbar buttons, and placed so many things exactly where the user needed them. But alas – no matter what I tried, most of the buttons were just too foreign and people could not figure them out. I had to change course.
To be honest, I debated telling you this at all. I know how finding this info out made me feel and it was not good. I tried so hard to force this design on users in a bunch of different ways. I really wanted to make it work. Not to mention that I had no backup plan – and it meant a huge redesign task and delaying the project. And I was frustrated, sad, and disheartened when I couldn’t make it work. And maybe, being a bit honest for a minute, I might have been a bit disappointed in users – I wanted them to love this – I wanted them to try harder to get it – but I just couldn’t make it happen. For me this was crushing – and I really hope that you don’t feel that way hearing this too.”
Sorry for you in front of so much misunderstanding and frustration: it’s a shame. Yet there are many examples that demonstrate the opposite. Remember the reaction of users with Firefox and Safari when their programmers have decided to use the same text field for displaying the url AND entering text for search! … It may be necessary insist, perhaps, and again. But here I am not in your place, unfortunately. Good luck for the rest of your work.
I’m at least going into the Stacks 4 beta with the edit button entirely removed – you just have to double click a stack to edit it. 
Removing any UI doesn’t always work out – wish me luck. I think this one has a good shot, but I have a backup plan if people hate it.
And I think it still causes significant confusion for novice users. But… I understand your point. Sometimes it’s important to have strong opinions about your software and disregard the complaints of users – users never like change.
sorry just a style thing. it’s not terrible— it’s just my personal preference.