Today's Scrivener newsletter & Tinderbox

While I’m grateful to both Scrivener & Tinderbox developers to offer us a huge discount on Tinderbox, the mentioned newsletter raises the following questions, TMHO.

  1. Tinderbox is not an esay app to use, mainly for those who enjoy Scrivener’s simplicity. Why don’t “they”, both Scrivener developers & Tinderbox’s developer Mark Bernstein, make a detailed tutorial on how to make the best use of both applications?

  2. Mark Bernstein is always looking for Tinderbox users willing to show online how they use Tinderbox so that everybody can learn from them.
    May I suggest Ioa make public part of his workflow? (relating to what we get in today’s newsletter, I quote: "Tinderbox helps me organise and develop ideas in a fashion that is easy to pair with Scrivener. I’ll use it to aid in world-building and detail-tracking, manage my timelines, find patterns in research data, and even just jot down everything that still needs doing in my current book. The program scales with me depending on what I need, addressing everything from simple lists to intricate visual maps depicting the relationships between all of the characters in the book, depending on where they are, and when. If I need to move these details into Scrivener, Tinderbox’s export capabilities produce files which can be bulk imported into the project file as needed, either as rich text files or OPML outlines, or by using both Scrivener and Tinderbox’s respective SimpleNote synchronisation features to share information. If there is something I want to do with raw information, chances are I can make Tinderbox do what I want.")

  3. The problem with Tinderbox is that we are told ALL the great things we CAN accomplish using Tinderbox but we are very seldom told HOW.
    Why not enjoying the opportunity of the big discount on Tinderbox to let us know a little bit more HOW great things are accomplished?

Note: I personally took advantage of Christmas 2010 Tinderbox discount. I’ve now owned Tinderbox for 6 months. In spite of a lot of time and patience (Tinderbox Wiki and so forth…), I must say I’m still not able to accomplish much with Tinderbox. So, my personal opinion, at least for the time being, would be: due to both my ignorance and the lack of well targeted Tinderbox tutorials, I shouldn’t have purchased Tinderbox.

I fully agree. I too purchased Tinderbox at Christmas and have found it incredibly frustrating to do anything with it. The ‘manual’ or ‘help’ is very poor and simplistic. I sincerely regret purchasing the product. It may be wonderful in some people’s hands, but not so far in mine. Right now, it is a hidden application on my Mac and is one that once in a blue moon I start up and fiddle with and ultimately write it off for another day when I have more patience.


I concur: I downloaded a trial of Tinderbox and fiddled with it for a while, but ultimately could not figure out a way to make it useful (or maybe just how to use it). I also am trying DevonThinkPro on my computer and am studying the tutorial and trying to make use of it as well. Is there any one who has or uses both of these? Or can make suggestions as to the usefulness of one over the other? My writing ranges from technical essays to a biographical project to literary criticism. (Yeah, I know, jack-of-all-trades, master of none.) Lots of data bits of varying kinds to organize and integrate.

Thanks in advance–


I’ve used Scrivener off and on for over 2 years and love it. I’ve had DTP for about 18 months and use it almost every day. I bought Tinderbox during the same offer last Christmas, and so far, I can see ways in which it will be useful, but I’m really still coming to terms with it.

I wouldn’t say I was an expert in any of the programs (certainly not Tinderbox), so please don’t treat any of these thoughts with anything but polite suspicion: your mileage will certainly vary, and those who know the programs better will probably able to point out errors and omissions, and there certainly isn’t space to list everything the programs do: they are all very fully-featured.

At the moment I’m using DTP far more than the other two, because I’m in the process of taking notes on a subject that may one day (possibly, hopefully, fingers crossed) turn into a historical novel. I collect everything relevant into it (websites, pdfs, images, notes from books etc) and DTP is really good at this ‘accumulation and classification’ phase, where the AI will help you classify your research and highlights links between items that you may not otherwise have found. I also use it to create my own notes on the things I’ve accumulated. It’s OK at that — there are some useful scripts included which allow you, for example, to highlight sections of a document and automatically create a new document where you can add your own notes, cross-referenced to the original. Text entry is bog-standard TextEdit standard, but serviceable enough (if you don’t mind swearing a lot at the programmer from Planet Zog who wrote the styles and lists implementations). So, a long way from Scrivener 2 in ease / pleasure of use for text editing, but OK.

Tinderbox is based upon notes, although its text-editing features are non-standard and more basic than DTP’s - bold, italic, colours and limited bullet points and indenting. Perfectly fine for the intended purpose of taking notes: I personally wouldn’t want to spend all day editing in it, but lots of people do. It’s a lot better than DTP though at outlining and visualising the structure of your notes. When you factor in its ability to have prototypes for different types of notes, to define attributes for each note, to define different types of links between notes, and to run regular programs (‘agents’) to act on the attributes and links, it is incredibly powerful — so powerful that it’s overwhelming at first.

I’m feeling my way with it, but the basic workflow I’m developing is:

  • Collect all the documents etc I need in DTP and annotate them there

  • Import relevant annotations into Tinderbox, automatically assigning each note under its source, character, date, location as relevant

  • Using agents to pick up the links between notes: for example, finding out what different authors have said about Ostorius’ spell as governor in Britain, sorted by date and displayed on a visual map.

Each of the notes has more detailed text within it, taken from the DTP note:

The lines at the top include the user-defined attributes, including a link back to the original Devonthink file.

The notes are all pulled together as you see them in the first screenshot by the following ‘agent’:

, which also automatically adds the date of the event to the note title.

  • The next stage will be for me to extend this sort of thing so I get an idea of what the general consensus is, where there are gaps or conflicts in dates / interpretation etc. For example, I can build up similar maps of all references to Verulamium and display them in a timeline, add my own characters to link to events and locations etc.

  • Eventually, I will try to pull all this together into something that makes sense as a coherent narrative into which I can write the actual story: if you like a research spine on which I can develop characters, and develop the outline of what I think should happen.

  • The final stage will be to transfer the outline to Scrivener and refine and actually write the Opus futurum sed incertum

Bearing in mind that I’m right at the beginning of this process, my initial thoughts are that each of the three programs adds something useful - DTP for the massive data collection; TBX for visualisation, assimilation and large-scale organising; Scrivener for final outlining and writing.

Well, that’s the theory anyway: there are a few obstacles stopping this being as seamless as I’d like. e.g. DTP and Tinderbox don’t talk to each other particularly well, so there are a few hoops to go through to get data from one to another, which in my case involved amending some applescripts to avoid endless copying and pasting between the two, and as I’m not a programmer there has been some fun in trying to set up agents in Tinderbox the way I want them. But both programs do have very helpful user communities and I think I’ll be able to do everything I want to do in one way or another in the end.

So that’s how I plan to use them: apologies for the length of this post from someone who isn’t an expert, but I hope it’s given you a bit of a flavour of what’s possible and how I see the three programs interacting. (And if anyone has any easier ways of doing any of this, please let me know…)

Use DevonThink Pro constantly, maybe even more than Scrivener. It’s absolutely indispensable for anyone who collects large quantities of reference material.

Try Tinderbox every six months or so, struggle for an afternoon to bring in enough information about my current project to be useful, but typically abandon the effort within a week. For that reason, have utterly failed to climb the learning curve. (And by climbing the learning curve, I mean a Himalayan expedition with climbing gear and Sherpas, not one of your afternoon hill walks.)


brookter –
Thank you, a very useful summary.

I also bought Tinderbox in the Christmas offer, but I haven’t had time to approach it properly yet (I’ve been finishing off a massive project). I think one of the problems the program has is that it is so flexible and powerful it is difficult to see what you could use it for until you have learned to use it. There is a kind of vicious circle which I can’t quite define. You have to use it a lot and become quite an expert before you can use it. The experts at Eastgate seem to think that you can start small and work your way into the program, but I think the experience of many people with the program belies this. A lot of people seem to start small and give up. If I can use an analogy from my experience as a gliding instructor, we used to teach a lot of people who got to go solo, and then couldn’t progress to the stage of flying cross-country, and eventually gave up. The reason was simple – we didn’t train them while they were trying to make the transition. Everybody thought that getting to solo standard was the difficult bit, and that then they could teach themselves the rest. Some could, but in fact most of them couldn’t, so we lost them. So I would suggest that there may be a large and difficult gap in using Tinderbox that comes between starting small and progressing to the next stage. Perhaps I am wrong about this, and my analogy is leading me astray. It would probably be worth it for Eastgate to collect impressions from people on this forum who have struggled with the program and given up to find out what the problem was.

There is a thread about the difficulty of learning Tinderbox on their forums: … 1304060068

Which points to an introductory article that is quite useful: … ent-151779

My two penn’orth.

Cheers, Martin.

I’ve been using Tinderbox for a little bit over a year now. Like many users, I would cuss Tinderbox and run away, but for some reason (seriously, I have no idea) I kept coming back to it. Now, I can tell you that Tinderbox is such an amazing piece of software that is very much a part of my workflow.

My advice is to start simple and let the program assist you when you are ready for it to assist you. At the beginning (and probably still guilty of now) I would over complicate my documents. Before I even had a single note written in a document, I’d have half-dozen or so user attributes and then confine my notes to whether the attributes would work with them. A huge mistake. Just write notes and go from there. If you need attributes later on, create them and there you go. Just start writing notes…

Also there is no limit to the number of attributes you can create. Let me say that again…you can create as many as you need, so if you think that you need an attribute for a particular note, but decide not to create the attribute because it only pertains to ONE note, CREATE the attribute and assign it to the note. There is no limit.

Go to the forums often and read. Download some of the Tinderbox cookbook files and the reference file that Mark A. has written and reverse engineer them. Open them and read how they wrote their codes for specific notes.

I’ve used Tinderbox for everything to keeping a home inventory of my items for insurance purposes to writing to doing my taxes. I’m serious when I can really think of no limit to what you can do with this program once you’ve decided to give it time.

I commend you on your patience, but most people, when they buy software, expect something a bit more functional and user friendly out of the box. I don’t ask to have my hand held every step of the way, but I do expect some manner of useful help file that goes beyond the absolute minimum. Scrivener is a relatively complex program, but its manual is exemplary in creation and usability, hence I could get up to a functional speed in Scrivener relatively quickly. The same for Devonthink, which I use daily, but am quite sure do not really employ its full functionality. With these programs I feel I am getting my money’s worth. With Tinderbox, I am pretty sure I threw good money after bad. I hope someday to prove myself wrong, but I have a life to life, and jobs to do, and Tinderbox is not easing the tasks.


It might be useful to talk about a specific task you’re trying to do, or a specific obstacle you’re encountering. Tinderbox needs of, say, a fiction writer who is plotting a thriller would be quite different from those of a graduate student collecting notes for a historical dissertation.

Of course, if there are mechanical obstacles (“where IS that button again?”), just ask and we’ll do our best. But, in my experience, the issues are not chiefly mechanical, but conceptual.

I suspect that is precisely why better documentation is needed. Some simple step-by-step worked examples (say, one for laying out a novel, one for laying out a dissertation, and so forth) would, I believe, be enormously helpful for people trying to get to grips with what is, after all, very unusual, and very complex software. I know it is very difficult to put oneself in the shoes of the beginner (I’ve done a lot of teaching in my time, and it’s a problem that I have often had) but I believe some better documentation would make an enormous difference. I can tell from visiting the forums that the two Marks both want to help, but I get the impression that total beginners are not really getting the right kind of help. Of course, I may be wrong! And, as I said above, perhaps the program does suffer from this closed circle in which knowing how to use the program (having already used it) is the only way of knowing what you can use it for. I realise that that is a rather obscure statement, but I don’t quite know how to express the idea. The only analogy that I can think of right know is that sometimes in order for stroke victims to relearn how to move their limbs, people have to move their limbs for them, so that the messages flow in reverse, so to speak, from the limbs to the brain, before they can then be sent in the other direction. Sorry – this is getting even more complex than Tinderbox! I’d better stop rambling.

Best, Martin.

Moved to “Feedback”.

I share some of the frustrations with Tinderbox, but, having engaged in discussions similar to this one on this forum and elsewhere in the past, I also have an inkling of why super-users like Ioa value the application so much.

It seems to me that Tinderbox effectively has two modes: there’s what could be called “OmniOutliner Plus”, which is essentially an outliner with added mind-mapping (and now, time-lining) and is excellently summarised by Steve Zeoli’s article (N.B. a review, not a tutorial) referenced above.

The help and instructional documentation for this mode could be better, but it’s not too difficult to understand, and in that mode, what Steve if I remember correctly calls the “shallow end of the pool”, Tinderbox can be a valuable idea-structuring tool for use in a writing workflow pre-Scrivener.

And then there’s Tinderbox’s more advanced, deep-end mode, as “a tool for creating tools”. That’s more complicated, more difficult to understand and requiring much harder and persistently applied trial and error, it seems to me, but, it is said, very rewarding for those users who need to use it in this way and are prepared to put in the effort. In this mode, Tinderbox is far from the user-friendly, relatively straightforward tool that, say, OmniOutliner is. But to provide the breadth of functionality it does, it probably cannot be anything else. I think it’s often this mode to which Eastgate are implicitly referring when they say “Tell us what you want to do and we’ll advise you on how to do it.”

I think you can splash about in the shallow end of the pool without ever approaching the deep end; I do. The trouble is, the deep end is ever-present and distracting and in some ways very enticing, and, of course, you’ve paid for entry to the whole darn pool.

As a postscript: Eastgate have implied for some time that they are planning to enhance their instructional and help resources. If I remember correctly, Mark Bernstein recently talked about an announcement in the very near future.

I think that mbbntu hit it close to the mark with his/her comments about learning a new task/program. The eyes of the teacher have to mimic the eyes of the learner and not view the problem only within their own, vastly greater, knowledge of the program. They have to go back to how they initially viewed it and work from there. Some step-by-steps of how to approach a technical problem (rather than, say, a novel) would be very useful. Some more extensive videos would also be appreciated. I learned quite a bit about Scrivener’s possibilities from those videos; not necessarily how to do it, but what could be done. Something a bit more complex than in the manual, and less complex than in some of the ‘examples’ on the TB website, which to me are incomprehensible.



There is no lack of documentation for Tinderbox, though some of it needs updating. I think the real problem is that the applications one can use Tinderbox for are so customizable as to make it very difficult to create a “how to” type manual. As an exaggerated analogy, it is somewhat like expecting the manufacturer of your car to provide detailed directions for your drive to the office. You will notice that in my Mac Appstorm review I rely heavily on examples of uses to illustrate features. By necessity, then, every other type of use is short-changed.

So, yes, it can be frustrating to use Tinderbox, especially at first, and I wouldn’t try to tell anyone what level of tolerance they should have for frustration. However, if you’ve already purchased it, why not try to make use of it, even if for only small chores?

As an example, here’s a simple way to put Tinderbox to use. Set up a map to chart your day:

Tinderbox Today.jpg

This is very easy to do. Just add adornments and notes. Move the notes around to where they make sense. Keep the map open and refer to it during the day as needed. Next day, just delete items you no longer need. Ones you want to keep you can drag into the “Archive” note. When tomorrow rolls around, move items from the Tomorrow adornment onto their appropriate places on Today’s grid. This application doesn’t require using agents or knowing any regular expressions. It can all be handled manually very easily. (Set up your adornments first, then use the lock feature to make sure they stay in place and you don’t grab and move them by accident.)

Here’s an even more basic usage: Just open a blank map and leave it open. When you have a thought or idea, switch to the map, make a note. Why is this better than any other note-taking application? Maybe it isn’t, but maybe one note generates thoughts for others. In the Tinderbox map you can cluster these together, then at the end of the day you can cut and paste the text to whichever program you choose.

Obviously, you wouldn’t purchase Tinderbox just to do these simple chores, but since you’ve already purchased it, why not try to put it to use? You may start to enjoy Tinderbox and can then begin to use some of the other features.

Vermonter’s “day management grid”, incidentally, is also a very handy framework for coordinating several characters’ movements in a complex day – something that might easily get out of hand in a thriller, say, or indeed in a Dickensian braided novel.

The same approach can be handy for film continuity, especially in documentaries.

Finally, a very similar use of the map is an ideal vehicle for planning events and conferences, where different things will happen in different places. You don’t want the Chancellor to be scheduled to introduce the Dean in the Albatross Room at the same time the Chancellor is to deliver her paper on the lifecycle of the Brant in Cockatoo Hall.

I’ve been using Tinderbox for years and it’s the second stage in my workflow (or the first, if I’m not using reference materials).

1: Everything goes into DevonThink (sometimes using Notational Velocity as a lean front end on my 2Gb Air).

2: Stuff winnowed for a particular project then goes into Tinderbox. I use Tinderbox to organise my thinking visually (by Map views), sequentially (by Timeline views, which are wonderful) or in straightforward (but super-powerful) Outline views. Then I use Agents and the more powerful live-search tools in Tinderbox to pick up emerging properties from my stuff.

3: Once a decent structure has emerged, it’s off to Scrivener for writing.

4: Word processor or Final Draft for gussying-up.

I also use Tinderbox for taking notes which then, once ordered and cleaned up, go into DevonThink via HTML Export.

I agree Tinderbox can be tricky at times. But Mark Bernstein’s technical support is unmatched (a similar level to Keith’s, which is saying something) and I have yet to find anything I want to do in Tinderbox which I discover is impossible. It’s a remarkable tool but like many remarkable tools, it’s complex beyond the top level. But unlike a lot of apps, there IS something beyond the top level. WYSIBNMWYG*.

It does need some focus at first. That’s not the same as money down the drain. You could go and drop £2,500 on a Hiren Roy sitar then find you’re unable to play Raag Manomanjari and consider you wasted the dough. But the truth is you need lessons, practice and patience.

Perhaps the truth is we’ve come to believe software should be immediately transparent. But even Scrivener is seemingly complex to beginners. ALl I’d say is that if you just need a simple outliner, Tinderbox is massive overkill. If you want an “idea processor” or “tool for notes” of unparalleled power then there’s nothing like it, though that comes at the cost of some complexity and a longer learning curve.

EDIT: * (vid. sup.) was going to expand that initialism into something really really witty but I’ve forgotten what it was going to be now. But it was good. Damn good. Trust me.

Not only have you paid for entry to the whole pool, the deep end is what makes this particular pool so enticing.

Regarding the “tell us what you want to do and we’ll help you do it” idea… It’s a circular problem, because if I had a better understanding of Tinderbox, I’d be more able to explain what I want to do with it.

For example:
I write a lot of technical articles centered around closely related topics. I use DevonThink Pro as my main data repository, so if I have a new project I can pull in related materials from past projects. However, as part of the writing process, I generate notes: generally no more than a sentence or two, capturing a key point from my research. It would be nice to be able to re-use these notes for future projects. DTP isn’t well-suited for that, because the overhead of creating a new file for just a sentence or two is too high. (For me. I realize that others do use DTP for this purpose.)

It seems like an ideal application for Tinderbox. But I’ve been completely flummoxed by Tinderbox’s note retrieval tools. Search doesn’t appear to be smart enough: I can’t find what I’m looking for unless I already know what it is. The map view is too unwieldy: I find myself squinting at lots of tiny boxes, but can’t see enough of them at once. The outline view requires that I know what the future use for the notes will be at the time I create them, which I don’t, as does tagging (which I don’t quite understand in Tinderbox anyway).

Essential information for a note includes:
Title, source, creation date, note contents, and ideally a link to the original source in my DTP database.

In the best of all possible worlds, I could import small PDF files. I use the Livescribe pen to digitally capture notes on paper, which can include drawings and other non-text elements.

A single project might create anywhere between a few dozen notes, and hundreds, depending on the complexity of the project. For Tinderbox to be better than disposable notes on paper – my current solution – I need to be able to retrieve notes that are old enough for their original context to be forgotten.

Once I’ve collected notes relevant to a project, I’d like to organize them in a mindmap/cluster map which I’ll then use to structure the draft of the article.

Suggestions? I understand how to create notes in Tinderbox, and I think I understand the mindmap creation piece. It’s the part in the middle that I’m having trouble with.


Yes, but how? This is where I get stuck.

(And this is the “deep end” of the pool that is the reason why I bought the program in the first place. It’s very frustrating.)



Agents are not as difficult as they may at first seem. The complexity (in my view) comes when you start needing to make sophisticated regular expressions to winnow out the data.

Here’s a screen shot describing how to create a simple agent that searches a key attribute called “project” for a specific project name:

That’s really all there is to it. You can also have the agent perform some action on the notes it finds, and that starts making it a little more complicated.


Katherine, if you want to find these notes that are so old that you’ve forgotten the context, how are you planning on identifying them? You can do it by tags, by custom field, by searching for strings of text, anything you like. If you let me know how you’ll be finding them, I’ll tell you how to do it. (For example, it’s actually dead easy to identify all notes referring to Fred OR George OR Mary OR cephalophods which were created between Jun 23 2008 and yesterday AND which are classed as referring to criminals AND have at least one note descended from them OR are linked to notes which contain the word “plesiosaur” but have no more than 20 words of text which mustn’t include the words “clavichord” or “marzipan”…

… and having found them, make them bigger than all the other notes, turn them bright red, and put aliases into a nice Map window so that you can organize them visually.

No, really. It is.

The trouble is that with power comes responsibility. You have to be reasonably logical about what you want to do. Like I once turned a huge book file into over 7,000 notes all suddenly renamed “Test Of New Agent”. Thank God for Undo…

As I said, post a reasonably concrete example of what you’re trying to do, and we’ll fix you up to do it.