I think it useful to understand ‘note’ is a very loose term here. One users note’s might be a few sentences of un-styled text, another’s might be a bit of text and a lot of user attribute values, yet another’s might be the equivalent of several pages of styled text with images/documents inserted in the text. All these behave differently in scale and also depending on which view type(s) the user chooses. It’s simply not possible to say “this works up to N notes” or “view type X only works well up to Y notes”.
A lot of people use Tinderbox in tandem with DEVONthink. Why? DEVONthink works well as an ‘everything bucket’. That is not a term of disparagement. Rather it describes a tool intended to track lots of items and to cope with the scale involved. Tinderbox, by original design, was intended for personal notes. Back in 2001 that would have been small sohrt notes with little.minimal text styling. In response to use request, there is now much more complex style support and the ability to add images.files to notes. But all this extra complexity is stored in the note.
So, any issues of performance and scale are a user choice. If you want large notes with lots of styling, I’d use multiple Tinderbox documents files (and not open lots at the same time). If your note-taking is more restrained, you may have less need for multiple Tinderbox documents.
With such a multi-facetted tool, the hardest answer for your fellow users here to answer is “How do I use this?” There will be a best way for you but as we don’t know the overall count of your notes and their size/complexity, it’s hard to suggest what that might be.
Cost and value are separate things. The sticker price is the cost. Whether that is value for you is a personal judgment. Tinderbox certainly isn’t a 99¢ AppStore impulse buy. At the same token, the fact that there are those here who’ve used the app for many years (and this (presumably) kept engaged and updating it should give you confidence that the app can deliver useful value to its users.
Tinderbox’s ability to visualise relationships isn’t a case of tipping all your info into a bucket and being given a pretty picture. Rather it provides a set of tools (nuanced in function once you get familiar with them) for exploration of your notes/information/data.
This above is the crucial point. Tinderbox won’t do your work for you, it will do your work with you. It has few preset paradigms, apart from (very fundamental) things like different views, note attributes and prototype inheritances. The rest is largely up to you. This can mean a bit of a lost-at-sea feeling at first, but once it clicks its vast utility and flexibility can open up possibilities that prebaked solutions often can’t.
The learning curve is steep, but in my opinion, well worth it. It’s aided by this forum, @mwra’s website and the Getting Started and Actions and Dashboards documents in the app help menu.
(As for MacBook vs Windows, I use both OSs. I strongly feel that on a laptop, macOS is by far the superior, largely because of Spaces. You can have multiple desktops running and swipe between them with ease. Windows has a version of this, but it’s just nowhere near as good, again in my opinion. There are things I prefer in Windows, but if I had to choose between all-in on one or the other, macOS would be my choice. Tinderbox and DEVONthink are 2 of the many reasons.)
Actually, modern Tinderbox should not experience performance problems on modern machines, even with thousands of notes.
If you do experience problems, get in touch.
If you have hundreds of thousands of notes for a single project, Tinderbox might not be a good fit. But of course there’s no known way to display an information visualization of that many notes in which individual notes play a meaningful role.
this makes me happy. i’m concerned that some of my files may become unwieldy.
each file is discrete, yes? or…can we link files? could i create a master file called latin.tbx and create containers such as verbs, nouns, adjectives, etc, and then in each container link the appropriate files: verbs.tbx, nouns.tbx, adjectives,tbx and so on?
i don’t think i’d need or want to embed the files, just link them so i can access them when necessary without creating 100,000 notes in one file?
I’ll give you an example. I have parsed the “Online Dictionary for Library and Information Science” which in its original format wold total to around 800 pages of simple dictionary like book format. In Tinderbox the book contains 8733 words (1 container per word) with the description of each word residing in the text pane. The maximum depth of the outline hierarchy container is 1 (with two top containers for organizing purposes). I have no actions, a couple prototypes, and 23 aliases.
I didn’t even consider creating links and attributes to use in the hyperbolic & attribute views because it’s already to slow. Create some actions and the project will load from pane to pane yester-year; the “beach-ball” effect will become your best friend. I stopped using Tbx for this very reason. I simply cannot split a document for performance benefits when that document has relationships which are not mutually exclusive…that would be non-sense.
mind you I have an extremely capable Mac to handle a full Logic Pro music ensemble loaded with hundreds of plugin instruments. Loading up a virtual machine to test Tbx will diminish performance as well (I use VM for MindManager Windows version).
That’s one book that I have a hard time managing in Tbx. It’s a tremendous piece of software otherwise…
8700 notes is, for Tinderbox, a big document. 1000 notes in a containers is a big container, too; it might make sense to split those containers into smaller sub-containers
Please send a copy of your document; if Tinderbox 8.1 doesn’t perform well with your document, I’ll be (a) a little bit surprised, and (b) will make it perform better.
I do not worry about performances. All notes I have are very un-styled notes (I should say “thoughts”) with a few words. Most of them are in fact just references: book name, relevant page (where I can find highlighted lines in my IRL books).
My concern is about links. I need a tool that enable visual exploration of notes to find new relations. I’m particularly interested in Tinderbox precisely because:
I’m ready to learn a bit of code synthax if this can open the way I look at my notes.
So… I manage to have the VM working, I intalled the Tinderbox trial and tried to design my example:
I have the notes about Lavoisier thermodynamics that have “Author2=Lavoisier” attribute
I have the notes about Newton thermodynamics that have “Author3=Newton” attribute.
I created an agent called “Lavoisier+Newton” that query “$Author2==Lavoisier | Author3==Newton” and collect the notes I want to see. I use the Agent as a note filter.
I realized that this method do not diplay the links in the Agent (see image below). I also see that linking two notes in the agent (Aliases) do not impact the originals. That’s the two things I’d like to be able to do.
Is there something I did wrong? Or do I simply not use the right global approach of the problem? I think I understood aliases as “replicants” but they are not. I’ve found some threads about replicants in Tinderbox and it feels like a touchy question…
I’ve looked over the references given above by many of you but I have to confess that there is a lot to read and learn, too much for a first try (for me at least).
May I ask you some help to buid this simple case?
This is expected behaviour as aliases export but don’t display their original note’s basic links.
Aliases can have their own basic links. If you want to use an agent to create links between the original notes represented by the aliases, after dragging a link, use the pop-up selector on the create link pop-up to choose ‘original’ for both source and destination. The link will be created between the two original notes rather than the aliases but will not show on the agent’s map for the reason already explained.
Tinderbox has no concept of a replicant (such as found in some programs like DEVONthink).
Ok so if I link the two originals (via their aliases) and I also link the two aliases, I get what I want. It just requires 2 linking operation, fine. The link between the aliases are kept while updated so perfect.
However, notes selected by the agent are not initially smartly placed on screen. So I move it by hand but it is all reverted at update. Is there anything I can do to avoid auto-placement at update?
I also have an issue with “thermodynamics” note that is taken by my agent while it has no attribute. (It had the attribute “Author2 == Lavoisier” but I removed it, see image below
I may be wrong in my thinking, but I feel that one of the key features of Tinderbox is its dynamism – the fact that as you add items (notes) the results you get from Agents will change, or the Agents will change things (appearance, etc) as you add material or metadata to exising notes. Perhaps the key is in the name – Agents act, they do things. Other programs are static. You arrange stuff and it stays like that. But Tinderbox, I feel, is about exploration, which means that constant change is (or may be) an advantage. But that is just the way I see it, and others may disagree.
It appears you removed the $Author2 attributed from Key Attributes. But that action does not remove the attribute or its value from the note – the action merely hides the attribute from view in the KA table. You will need to remove the “Lavoisier” value from $Author2 in order for the note not to appear in the agent results.
This issue above is a narrower noe - and not arguing against dynamism & discovery. Originally, each update cycle refreshed the agent’s map and in doing so re-sorted it. If the user wants to use the dynamism of the agent query finding the right things but also further working that map, the user ends up fighting the agent or having to turn it off. The re-sorting of the map isn’t a necessity, but just how things started out.
The addition of the $CleanupAction (c. v5 if I recall) allows some flex. IOW, the status quo remains and in most cases the agent just refreshes the map as it sees fit. For those who want to work that map the option exists to turn off the clean-up (i.e. auto-re-sorting of the map). I assume that there is some extra work for the app or likely the new feature would have been a change as opposed to a new option.
Sure – I wasn’t really replying to the narrow issue – merely allowing myself to ponder about the general philosophy of using the program. But I acknowledge that Tinderbox offers a heck of a lot of options, and the way I think about it may be opposite to someone else’s view. And I suppose I found myself wondering if producing a static view using Agents was the best way of doing things. But now I know that it is possible, it adds yet another tool to the box. Cheers! Martin.
A few questions about a few things I did not manage to do yet:
It seems that clicks on bottom link stub of my notes do not work. Is it due to trial version? or maybe VM? Or did I missed something?
I want to draw a link between nodes stored in separate containers. I guess I have to use the “parking space” but I did not manage to figure out how to use (reading this).
Is it possible to add an action to the agent that could automatically apply a link created between aliases to their originals?
#1 is for creating links. Click and drag to start a link, drop onto the target note or a link park.
#2 is an inbound stub. Click the numbered circle to see a pop-up list of inbound links from links from other maps. Click on an item in the pop-up list to select that note in the text pane. Note: _the map does not shift to the newly select note’s map.
or
#3, as #2 but for outbound links.
An up to date reference for that link is here. The process is in 2 parts:
Drag a link from the selected note (see #1 in the first screen grab above) and drop it onto the link park on the tab bar:
Now change the view so that the target note for the link is visible, i.e. scrool the map or move to a different map. Click on the link park and drag out a link dropping it on the target note.
If successful you will see the link creation pop-up, like below, which lets you complete creating the link:
An action? Yes, see linkToOriginal() but note this action can’t be triggered by a manually linked drag.
If you have more questions, can you please start a new thread for each discrete question. It makes it simpler to treat the questions separately and helps leader readers find answers to the same question. Thanks
