I’d also note that I’ve spend >1 year helping a Tinderbox user trying to easily extract MarginNote note data on a repeatable basis. The problem, sad to say, is at the MarginNote. I understand the MarginNote developers are aware but I suspect usable export isn’t their priority. Currently, the best method seems to be to use some slightly flakey OmniOutliner-format export from MarginNote then script it into cleanly formatted data that OmniOutliner can export as OPML and can then ingest into Tinderbox.
In other words, I’d check you can get data out of MarginNote in a form usable by other apps before fully committing.
Assuming usable data, or inputting data directly with Tinderbox, I’d agree with @PaulWalters that on the limited data offered, Tinderbox has much to offer.
About aliases:
I’ve read somewhere in this forum that it is not possible to add links to an alias of my Paul. I guess this might be problematic for my friends’n’family case isn’t it? Or did I missed something?
(to be honest, this issue is the reason of my initial post)
About Marginnote:
I see what I’m going to lose in the switch (I’ve spent some times looking at tutos for beginners). I plan to keep Marginnote as a fast (temporary) note taker while reading (on ipad). I understand from what I’ve read so far (and @mwra just confirmed it) that the interraction between MN3 and TB8 might be hard job. I’m ready to copy all notes taken by hand if necessary.
An alias always shares the text links of its original note. An alias, by default, shares its original note’s basic links too, but importantly these may differ as you can create links to/from aliases.
Your annotation system seems to map best to attribute values, analysed via Attribute Browser view. I’m happy to be wrong, but I think multiple aliased based maps of the same notes with different links will become a difficult thing to as the document and content scale up in size. An approached based on attribute values will scale much better, though visualising the links (e.g. via a map) becomes a harder task.
Reading you, I understand that I will have, at some point, to choose between scaling and visualisation. Is that so?
And what about using containers? One ontainer named “family”, one named “friends”, etc… If I want to have Paul into all these, will I have to create Paul-aliases?
In my experience, ‘Yes’ but let’s not be too prescriptive in advance of more information. Also note, screen size/resolution may also form a practical limit. If your map is so big you can’t see enough of it at a readable zoom level, I’d treat that as a real-world constraint.
But we don’t even know how many notes there will be on the map. 50-100, OK: 5,000, probably not. Please don’t assume there are literal limits, i.e. exactly N items. The complexity of your maps, the degree of extra styling (shadows, patterns, fills, borders), and other ongoing activity like agents, rules, etc. all feed into the mix.
To ‘re-use’ a person note as an alias in numerous containers you will need to ensure the links all run to/from the original. Once an alias has its own basic links, it no longer shows those of its original. Nor does one alias plot the links of another alias. It only ever shows its own basic links (if any) OR those of its original note.
Many thanks for all these clarifications. I realise that my thread is not the simple question I thought. I guess I have to take a moment to figure out how I’m going to use Tinderbox. Thanks to @MartinBoycott-Brown: I speak french I’m sure this is going to be helpful.
May I take the liberty to develop a little bit (while suggested videos may answer the following)?
Let’s say Paul is any kind of information that can be read through the angle of theory-A or theory-B (namely family or friends in my example). Paul finds a place in the two theories at same time. One could just create Paul-A and Paul-B in two containers but how the likeness of Paul-A and Paul-B would be displayed then?
I guess lot of people, using TB8, have already faced that kind of situation. It looks pretty common to me.
How do most people deal with it in TB8? How do you guys do?
I don’t know what Theory-A / Theory-B represent, but it appears that instead of fixed containers they should be agents. If “Paul” has a an attribute that satisfies the query for the Theory-A agent and the Theory-B agent, then an alias of the “Paul” note will appear in both agents. Agents are, in essence, containers whose content is computed on the fly.
I understand you may not be able to discuss the details of your work in public, but the metaphor you’re using might be somewhat distracting here. Specifically, it’s always tempting at the start of a project to divide the information you plan to develop into a tree of categorical relationships. The information you actually discover, of course, is never quite what you expected — and the structure you originally built is likely to be wrong.
I talk about this problem of premature formalization in The Tinderbox Way, which you might enjoy.
Let’s look at a specific problem: a year from now, we’re going to have a big election, and you’re going to lead a group of volunteers to advance a particular cause. Let’s say, “awareness of climate change.”
What I’d do to begin is to make a Tinderbox document to hold some things I know I’m going to want to track.
People – volunteers, possible volunteers, experts. Candidates for various offices. Staff workers for those candidates. Lots more people.
News – articles, news clippings, research papers that are relevant to the campaign and our mission.
Events – stuff that’s going to happen. Not just dates and times — you have a calendar for those — but also planning meetings, conferences, canvasses, caucuses, conventions.
Now, a specific event might involve different people, and we can represent that relationship in a bunch of ways. We can:
have a link from Party Convention to people who are delegates.
have an attribute DelegatedTo that contains the name of the person responsible for the event.
have a set attribute that lists the members of the event’s steering committee
have a text description of the event that lists some possible speakers, each of which has a text link to contact information for that person.
assign a tag “JulyRockConcert” that’s added to the event, people involved in the event, and news about the event. Use an agent to gather all the notes that use this tag.
An important point is that you won’t know the best way to do any of this when you start out. You might expect to be organizing a dozen volunteers, and you turn around and you’ve got 500. Or you think you’re planning fifty weekly events for about 20 local volunteers, and your national coordinator calls you and says, “drop that — I want you to focus on getting 5,000 people to the July Concert fundraiser.” So, Tinderbox makes it easy to change representations.
I think the simple answer (crudely summarising what is written above) is: “don’t put items in containers – give them tags instead”. (OK, horribly simplistic, but for someone who is just starting, it might be best to cut to the chase.)
Amen. This approach allows you to make good use of Attribute Browser view and it gets you out of the problem of constantly adding agents. It also copes better with large numbers of notes (better than a map, for instance).
When considering how an imagined outcome will look on screen in a map (as many of us tend to), also ask yourself what data you need to make a map look like that? If it needs data you aren’t directly typing in/importing, you will need to figure out how to get or create that data before you can start to worry about making the maps look pretty.
I agree … but I would amend that to “…give your notes metadata with existing or custom attributes that makes sense for the context”.
There’s a tendency to rely on tags in a lot of apps simply because few bits of software have Tinderbox’s ability to work with all manner of metadata using built-in or user-defined attributes. Tags generally become a big bucket of glop. It’s hard to make sense of notes with Attribute Browser if the stuff in the tags can have different meaning depending on the note.
(If I was just going to use tags, I’d save my money and instead use files, Finder, Finder Tags, Preview, and Finder Saved Searches.)
I concur 100% - I should have been more specific. ‘Tags’ is a slippery term and can vary in meaning depending on how much exposure people have had to structured metadata. Breaking out strands of your information into discrete attributes makes your analysis much easier. Certainly, using only general ‘tags’ from other sources, such as import to Tinderbox’s $Tags, is very much driving the Tinderbox app in a very low gear.
Sure – I was just trying to keep the message as simple as possible for somebody trying to find their way in (and not working in their native language). Sometimes less is more. Sophistication can come later
No confidential material in what I’m trying to achieve. I took a simple situation involving Paul trying not to be boring with my personal stuff.
I’m a physics researcher and teacher. I’m trying to find new relations in a large set of theories and experiements. The situation I’m facing is somehow similar to the situation of my students: Learning different approaches for same phenomenons and figure out the interactions. I deal with it using the method I learned as a student (except that I now try to use numerical tools). The example below is a simplified situation that could be find in one of my student notebook.
I have the map of thermodynamics theory according to Lavoisier (image 1)
I have the map of thermodynamics theory according to Newton (image 2)
Some elements in Lavoisier theory can be translated in terms of Newton’s. (image 3)
Some elements are the same in the two theories (“Hot bodies warm cold ones” in the example, see image 4).
Just wanted to know if I can display these 4 maps in a clic or 2 using Tinderbox. Keeping other maps up to date after any change in one of them. Based on what all people here said, answer is yes but not the way I initially thought: I will have to get how agents and attributes work first.
Maps like those can definitely be composed in Tinderbox. Be aware, that maps portray all the notes, adornments, and aliases on the same outline level. Notes that are containers for other notes, or aliases, are shown in a particular way on the map. Like this:
There’s a little “viewport” in both cases showing little icons for the child notes inside the container or agent, and you can navigate into the container by double-clicking the background of the viewport. There are other navigation features available.
Isn’t that always the case, though? Large data sets always pose visualization challenges that smaller ones don’t. The sun that a five year-old draws is qualitatively different from the sun as seen through NASA’s tools, which is different yet again from the assembly of all stars in the observable universe.
I’m sure this is going to be helpful.
