AmberV, (or any Tinderbox user)

I remember reading somewhere that you use Tinderbox for your notes. I plunged in to the fairly deep end with TB about 3 years ago, enjoyed it for a while but ended up concluding that it’s huge tweakability factor – especially when exporting to HTML – meant that I was spending way to much time fiddling and not enough writing. I also wasn’t wild about the upgrade policy and balked at sending the developer 70 (now 90) dollars each and every year for pretty incremental updates. Mostly I was heavily using DT at the time and couldn’t quite figure out how to blend TB and DT in my workflow.

Fast forward to this week’s Voodoopad discussion and while fooling around with VP I discovered that one of it’s web export modes meshed pretty well with DT. This was similar to an experiment i once tried with Tinderbox. Like many people here, I’m frustrated by DTpro’s limitations when it comes to actually writing and organizing/linking notes, but I love it’s strengths as a central depot for all kinds of material. it occurs to me that notes generated VP or TB might export well to HTML and be linkable and searchable from within DT. If I remember rightly the last time I tried this I had naming issues for the exported HTML files and stability issues in general with TB.

At least 2 years later both ny system and TB itself have been upgraded and since I have less time to waste I’m less worried about fiddling too much. So I’m considering giving some VP/TB+DT combination a try alongside Scrivener.

I’m not particularly cheap, but it bugs me in principle to renew this bloody Tinderbox license fee if it’s not going to pay off. The trial versions are way too crippled to provide an adequate test. And though Mark Bernstein seems like a nice guy and all, why should I subsidize his windows development while he pretty much ignores what’s happening with OS X?

I was wondering if you had an opinion. TB’s learning curve doesn’t really frighten me - I’m at least half way along it anyway - but the fact that it’s neither cocoa nor universal is a worry. Right now I can imagine fairly smooth workflow between TB, DTpro and Scrivener - but i do have an overactive imagination.

If you have a second, could I have your two cents on this?



Okay, I’ll blather for a while about Tinderbox. It is, after all, one of my favourite applications on the Mac. :slight_smile:

Actually, I used Tinderbox for more than just a place to keep my notes. I used it to define the actual structure of my books, and allow me to view various components of it in different collections. For example, I could have one view that shows the narrative order, and another view that isolated various plot threads, so that they could be read as a linear sequence. I also used its mapping abilities to place each of the narrative sections onto a huge time chart.

This was when I was also using Ulysses for all of my creative writing. Like you, I was never able to find Tinderbox as all that useful for writing creatively. There is just too much interface in the way for that. So I would write in Ulysses, utilising its full screen, and then dump the resulting text into my book structure. This was back when only a handful of applications had full screen. Nowadays, with WriteRoom around, we have access to full screen anywhere, and if Scrivener had not have come along, I would probably just used WriteRoom in conjunction with Tinderbox.

One of the things that Tinderbox really does right is allow a strong way of viewing data in multiple metaphors. Outlines are useful for some things; visual maps with arrows between them can be useful for other things; being able to follow a trail of links and branch off based on link type is subtle and useful; and collected text views such as what Edit Scrivenings provides, are useful for still other tasks. For real mapping, Tb is not as strong as something like Inspiration or CMAP, but it gets the job done, and is much much stronger in other areas than either of those applications.

So, Scrivener has changed things a lot. I no longer use Ulysses that much, and I have less need for filling the organisational gap, so my usage of Tinderbox as a way of keeping track of my stories has waned. I still have a lot of use for it, but mostly with web publishing and managing large projects from a task level basis.

I still find VP to be the best and most natural way to brainstorm and take notes on things. Tinderbox has a WikiLink mode, but it is rather limited for a number of reasons. Tb’s strength is in organising and displaying what is already in it. It isn’t so hot for making new things in it. At least that has been my experience. Fiddling with its millions of options are an asset when you are finding relationships between things – but not when you are trying to write.

The pricing scheme for Tb has never bothered me. I think it is unfortunate in that it is rather unorthodox and commonly misunderstood. I suppose, if you paid your dues every year and only got your incremental updates, it would feel a little expensive. There isn’t anything requiring one to do this though. The application will work forever with or without a yearly upgrade fee. The option is only there if you really do wish to keep up to date. You could treat it like any other major application that has a two or three year release cycle and only pay for the one year upgrade to get the benefit of all those years worth of upgrades. The difference is that, unlike those major applications, you get whatever else comes along for an entire year, as well. That is actually more generous than most offers which only give you the features up until that version point you paid for. Also unlike other applications, waiting doesn’t make things more expensive. If you wait five years to update, the upgrade fee doesn’t change. You get all five years of development, plus whatever happens in the sixth, for the same cost as if you had upgraded one year later and then stopped. So, patience pays off.

All of that said, some years are much better for Tb than others. This year so far has seen Tb grow in many substantial ways. A lot has been added to the AgentAction, OnAdd, and Rule fields that have massive implications on how data can be dynamically evolved. A lot has been added visually, too. Integration with OS X is much better. Other years have been less impressive. There have been years that I’ve skipped going for an upgrade; plenty happy with the girth of things that have been added in the two year interval.

While the trial is useless for new users, it can be pretty useful as a way of gauging whether or not it is time to spring for an upgrade. You can still load your 8,000 note masterpiece of information theory, you just cannot add new notes to it. Everything else is available including full access to export.

The concern you have with OS X integration is definitely one of the bigger ones. The whole Cocoa vs. Carbon thing doesn’t bother me as much. There is a valid place for Carbon applications, especially those that have a long standing history into pre-X days. If they already have a strong tool kit either hand-coded, or from professional libraries, there is little reason to use Apple’s Cocoa kit. You can still accomplish a lot of integration, including Services. Tb already has incorporated some of them – Services included. The fact remains, it is less integrated than a lot of applications are.

At least, it is less automatically integrated, and that is an important distinction to make. To an advanced user, its level of open integration is much higher because it is malleable to the user. Someone who has the ability to manipulate XML files can do quite a bit with Tinderbox. Its export engine is also extremely powerful. As long as your output format is text based, you can pretty much do anything. One of the first things I did with Tb was make an RTF exporter by using the HTML export engine. I ran into a few snags (which I posted on the Wiki) and since then those snags have been addressed. For those of you wondering why I wrote an RTF exporter, being such a plain text freak – well blame writing on a Palm Pilot and using WordSmith, for that. :slight_smile: It would have been possible, for an example of integration, to make a synchronisation script between Ulysses and Tinderbox, so that I could write in Ulysses and organise in Tinderbox. They would both use a common data source and never even realise it. I never created it though, because in the time when I started researching it is also when Scrivener came out. While in theory such a system could work with Scrivener, it would be a lot more complicated because of the whole rich text thing. A lot.

For the long term direction of development. Yeah, that concerns me. I know a Windows port is in the making, and I also know it has been in the making for many years now. I am not sure how much it impacts the Mac development, actually. Tb has grown a lot – more than the average Mac project does, as a matter of fact. Given that, and how long the Windows port is taking, I do have to wonder just how committed they are to the port. As a long time Mac user, you always get a little worried when one of your favourite developers says “Windows Port”. Too many times we’ve seen ourselves side-lined once the “real” money starts rolling in from the countless hoards of Windows users. There are two things that lower my concern: A) Tinderbox is always going to be a niche program, no matter what platforms it supports. The learning curve is too substantial for most casual users to get in to it. If they keep their 15 note trial policy, it will continue to keep people at bay (and that remains, in my opinion, one of the most absurd trial limits, ever! It would be like a Photoshop demo where you could only make fifteen brush strokes or menu commands). B) The fact that they continue to support the OS 9 crowd to this day, causes me to consider that they are willing to put money beside legacy support. There seriously cannot be a lot of money is an OS 9 anymore. So, even if they do end up going Windows, and they do make a lot of money there, I am not too worried about the Mac side getting closed down. Things are just not as bleak as they used to be, too. It isn’t 95% to 5% anymore. These days, Macs are everywhere.

But, I remain prepared. I only buy a yearly upgrade when the price is fair for the number of new features I’ll be getting – rather than banking on the chance that the coming year will have significant upgrades, too.

Oh, and while Universal is still in development, I’ve tested Tb under Rosetta, and it works just fine. Tb is one of those rare complex applications that actually operates on a system with very minimal resources. That makes it suitable for emulated execution. I’d imagine you would want to lower agent priority a bit for complex projects, though.

Okay, I’ve blathered enough.


Thanks for the detailed and considered reply – definitely not blather and worth way more than 2 cents. I really appreciate it.

This is precisely the sort of thing I’m looking – and failing – to do. Tinderbox came closest for me but I had stability issues - damn thing kept crashing me at the worst of times. I gather mine was a very isolated case; I see few complaints of TB’s stability around the tubes. But when you’re the victim, rarity is cold comfort. I’m hoping the new setup (24" Imac & Tiger) is more stable.

I’m not clear on how Scrivener would replace or supersede Tinderbox. Given that it has nothing like agents, or even aliases or clones, how could it help with the sort of things you list in the previous quote? Has Scrivener changed things or have you adjusted your process to Scrivener?

It’s funny: one of my bugaboos is marketing that tries to fool or manipulate rather than appeal to reason, yet some thick part of my brain refuses to accept the reason behind TB’s pricing scheme: it just feels wrong to me. Anyways, if this was just about saving 90 bucks I wouldn’t bother you with it. What I’m really trying to save is X wasted hours failing to set up a system that works between TB, DT and Scrivener. That said, your defense of Mark’s scheme is better than his!

Given your ability to address these issues so clearly, I’m tempted to bug you for more detail here, but I’ll search the wiki myself.

I’d find it a lot more useful if I could add 15 new notes to my 8000. How can I fully test new entry features if I can’t add new notes? Pisses me off; and I bet I’m not alone.

We come to the crux of the matter. I’m wondering what it would be like to hump my notes back into TB, massage them a bit (this is sounding a bit pornographic) using agents and so on, then export into a nice well presented and integrated HTML format that could be linked in Devonthink. That way I’d have a couple of powerful tools/means for searching the same data. And I’d have access to all my material, including notes and sources within DTpro.
Hopping from there to Scrivener is a bit more problematic, especially since Scrivener desperately lacks something I’ve come to take for granted in DT: the ability to create a new note from dragged text and automagically title it for the first line of said text. For me, this is Scrivener’s biggest failing. Nonetheless, I’m thinking of only transferring essential, up-to-date, strictly draft-related material into Scrivener and thus using it purely as a “focus and write” tool; having to do this more manually might actually help in the winnowing process.

The new Imac is in the shop for drive replacement, so I have a few days to consider this before I start fiddling. My powerbook and TB just don’t get along so I’m not even going to bother for now. I guess I’ll look more closely at the TB wiki to see if my earlier concerns have been addressed. It sounds silly, but just being able to export HTML pages that are clearly named – spaces and all – and thus easily searchable in DT would make big difference to me. I never managed that in earlier versions.

Anyways, now I’m blathering.

Thanks again for the input.


What version are you using, out of curiosity? I used to have stability problems back when 3.0 was brand new, but lately I cannot even remember the last time it crashed on me.

That is a good question. For structuring the novel, my usage of Tb was pretty basic, actually. While Scrivener does not automatically populate folders based on an expression, it does have saved searches which seems to be working fine for me so far. Between that and keywords, I’ve yet to run into a situation with Scrivener where I really felt as if it was lacking in some way of presenting the book to me. Yeah, I am missing some of the things I used in Tb, but not needing to keep two different (wildly different) applications synchronised with each other is quite nice.

But who knows. I might find that once the drafting is done, I prefer something a little more dynamic for editing. That was definitely the case with Ulysses, but it lacks a lot of things that keep a Scrivener draft fairly dynamic.

There may not be a whole lot of that, yet. Some of the potentials that I am seeing with the newest capabilities are rather advanced, and I have not fully explored them yet. For instance, Rules can now be self-adjusting, as they can modify their own expressions based on logical syntax. In the beta version, external shell scripts can be piped into attributes, and even modify the contents of the notes themselves. So Tb is no longer limited to its own small pool internal commands – it can literally operate as a processing hub for hundreds of scripts and UNIX applications.

Right, the defaults are set to be as browser friendly as possible. Short names, no spaces, et cetera. Depending on your version: Something you could try is to change HTMLFileNameMaxLength to 200 (256 is the Mac’s limit, I think), and then set up a Rule with this code:


Note, that will only work in version 3.4 and up, I believe. Perhaps 3.3. What it does is check the HTMLExportFileName attribute, and if it is empty it will assign the value of the Name attribute to it. If you want it to always keep them synchronised, at the expense of processor time, then remove the pipe symbol in front of the equal sign.

Dump this Rule into a common Prototype, and you’ll never have to worry about it again.

Thanks for the detailed reply AmberV, I’ve been in rehearsal and only just noticed you posted it. I ended up renewing my license but haven’t had much time to play with TB since then. I’m using the latest beta of 3.6 and I’ll have a detailed reply in a couple of days.

Thanks Again,


So, Eiron, have you gotten around to integrating TB into your workflow? I’ve been looking at it lately, but, as AmberV said, trying to keep the two apps synchronised is too much for me.

I’m still tantalised with some of TB’s abilities though – free form brainstorm, timeline abilities, creating a linked map of all my characters, even outlining the contours of my locales, etc.

I’m using the same setup as you – iMac 24", Scrivener, Tinderbox and DT. The world is too standard these days.

After a long period of hesitation, and after having read a lot of articles and reviews, yesterday I finally took the jump and bought Tinderbox. Why? Well, because I never could get rid of the idea that I missed something fundamental by not owing, not knowing, not using it. Maybe within a couple of months I’ll have to conclude that it would have been better to spend those $ 229 in another way. But in that case I’ll say to myself: it’s always better to regret that you tried than to regret that you did not try!

And yes, it would be great to hear about new (positive or negative) experiences of those members of the Scrivener community who also own and use Tinderbox!

I’ll jump in. I’m a reluctant (and maybe temporary) ex-Tinderbox user. It’s a brilliant program for putting context and order into notes, but for me it has a couple of critical failings.

One: it doesn’t harvest stuff as well as the competition. There’s no Services or hot key to clip material from the web or email, no automatic registering of a URL against a web or mail clipping. It’s still drag-and-drop or cut-and-paste. Once you get used to DevonThink’s way of clipping material, everything else seems cumbersome. (I wished it retained links in web clips, too, but can live without that.)

Two: It collects text (and images), but not whole documents. These have to be linked to notes. So if your references are scattered through a PDF, you either clip the items that interest you into separate Tinderbox notes, or link to the PDF somewhere in the Finder but lose the ability to search for those references in your Tinderbox document. Again, programs like DevonThink swallow and search both text and documents, which is immensely convenient.

Three: the geek factor. I’m decidedly non-geek, and while I can work out how to write agents, I’ve forgotten the language by the next time I need to write one and have to re-learn all over again. A plain language graphic interface would help here.

All of the above could possibly be solved by making Tinderbox a Cocoa program hooked into a database. I’m sure there are good reasons, apart from the pain of re-writing the whole thing, for not doing this. But not adopting all the Cocoa tricks leaves Tinderbox vulnerable to looking increasingly dated, like a film camera in a world of digital imaging. (Which is not to say that film cameras don’t have their virtues. Only my beloved Olympus OM-3 and OM-4 are gathering dust, while my Olympus E-1 is collecting wear and tear.)

Four: My needs are just not that demanding. I read The Tinderbox Way, loved the logic behind it all, but have found that I don’t leverage the power that Tinderbox offers. Every few weeks I fire the program up and look for an excuse to use it again, but inevitably return to the plain digital filing cabinet capabilities of DevonThink (or increasingly, Scrivener).

In a nutshell, I want to use Tinderbox, but I can’t find a compelling reason to continue using it. Same with my old Olympus 35 mm kit. Dammit.


I have been a temporary ex-Tinderbox user a number of times in the past as well; but I always come back to it and every time I do I am glad for it.

Harvesting difficulties: I really don’t think that is the core goal of the application. It isn’t meant to be something like DEVONthink or EagleFiler. It’s a tool for turning ideas into concrete conceptualisations or implementations. I think it can be used as a research tool, and be quite effective at it, but to me it seems more in line with the philosophy of acting as a concept layer above the actual material layer. Each note can be linked to a file on the drive—and once it is linked it extremely easy to retrieve the file and work with it alongside Tinderbox. The information that is necessary to have expressed in Tb can be copy and pasted into it and become a part of the network. But even in a really beneficial document, there is rarely a need for 100% of it to be available from within the concept network.

In my own experience, encouraging bulk document import tends to lead to data bloat. You get huge databases full of 60% or less useful data. I’d rather just select precisely what needs to be known and link to the file for all the rest. Keep the kernel clean.

As for the actual mechanics of getting data in: I really don’t see what is wrong with copy and paste or drag and drop? These methods are very efficient? How is a clipping service any more efficient? It’s just a matter of what you are used to, I think. Services are non-targetable, due to their nature, and thus require a sorting phase after they have been clipped. Paste is direct to context no sorting required. And if you don’t want to think about it you can always just paste your clips into a holding area. Since Tb creates notes dynamically when you paste into the outline—this is extremely simple and easy to do.

I’ve certainly tried clip-centric programs in the past, but honestly I never got it. Most of the time you have to access the service in some menu with the mouse (or contrive some incredibly arcane universal keyboard shortcut that will not conflict with anything), and unless you spend time keeping that menu clean, it can get really messy loaded up with a lot of stuff that has nothing to do with what you need at the moment. Cmd-V is always there, always doing exactly what you expect.

Difference of habit, I guess.

Well, I’m a geek, so the bit about agents doesn’t bother me. There is a graphical interface for that, by the way, and what it doesn’t address, it does show you how to expand by demonstrating the syntax as you build it with the standard drop-downs. But yes, it is pretty geeky once you get past the basics. I will definitely concede that. But I think it is a valid question whether or not that advanced stuff is necessary all of the time? I have plenty of Tb documents that are extremely simple—mostly just leveraging the map+outline+linking features that hardly anything else does. It’s nice to know that if I need to do something I can, though.

I have to completely disagree with the go-Cocoa+Database analysis, though. I wouldn’t mind the interface getting optimised for say, Intel. That’s my biggest grief with it right now. It runs a bit slow through Rosetta on my MacBook; the fan comes on a lot. But it already does most of the things that Cocoa can provide. What were you thinking of that it is missing? Carbon can be a valid alternative in some situations. As for going database? Ugh. No. :slight_smile: I love that it is an XML file. I take advantage of that frequently. That was one of the big selling points, for me. I have saved many weeks of manual labour simply by having access to the raw data in a logical XML manner.

That’s true, and so what I’m probably describing is the fact that I need a research repository more than a concept-generation tool.

Nothing wrong with them per se. But compared to, say, Eaglefiler 1.3’s F1+pop-up window to name, tag and direct a clip to a specific destination (or OmniFocus’s neat system), dodging around windows with d&d or c&p feels cumbersome. Because so much of my information comes via the web or email, I want collection to be as painless as possible. If all I’d known was drag and drop, I’d be fine with it. But I’ve been spoiled.

My Tinderbox documents tend to be basic, too. It’s just when I try and get fancy, I have to do a crash refresher course in agent syntax. It shouldn’t be that hard.

To be frank, I have no idea what a Cocoa’ed & databased Tinderbox would look like. I’m just imagining some happy mating of Tinderbox and DevonThink. One thing XML is not so good at is accumulating data mass. My bigger Tinderbox files are taking a while to save (“a while” in impatient computer language - actually, 5-10 seconds) and I wonder about the implications of accumulating further mass. DevonThink eats gigabytes as a light snack. But you’re right: most of the gigs I have in Devonthink are padding - a whole PDF for three lines of quotes, a Word doc for a figure.

Oh no. I feel the urge to take another look at Tinderbox…

Two quick responses:

  1. On data-mass and XML: Definitely, yes! I have definitely run into that problem with Tb in the past where I’ll have a document that is 100k+ words and it isn’t instant like we expect things to be in this day and age. I have since adjusted my approach to Tb, as a result. That file was a journal. I was storing all of my thoughts in it every single day. I’m not sure that is what Tb is best at. I use Journler for that now, and have been largely happy with it. I’ve also dabbled with using Scrivener as a daily journal. But Tb definitely does seem to be happier with lots of little notes, rather than lots of huge notes. A database back-end would help out there—but is that in its core interest? Not sure. I’m inclined to say no, and I think Mark B. would too.

  2. Regarding mail collection, have you played around with the new POP mail synch? It requires a dedicated POP account, but these days free POP accounts are easily obtainable. I just set up a little mailbox specifically for each Tinderbox document (one at the moment), and then using Mail Act-On assigned a keystroke to forward anything destined to that document to this special account. The next time I boot up Tb, that project automatically downloads all of the mail and turns them into notes. This is extremely handy once it is set up. When I’m out in about I can write little notes to myself on my iPod Touch, file them in the Mail outbox, and the next time I’m around wi-fi they all get sent to this holding area where Tb can access them.

just a thought on ‘harvesting’ data into DevonThink, Tinderbox, or wherever. I got a copy of iClipboard bundled with the latest SOHO Organiser upgrade. It’s available as a standalone app. too. it sits off the bottom left edge of my screen, and offers an unlimited number clipboards - . You can also have named clipboard sets. IT’s instantly accessible, and means that you don’t have to do the drag and drop thing all over the computer. IT’s pretty much similar too, but much smoother than, what i used to use a lot, which is CopyPaste, for the same tasks. So if you prefer not to use the services menu, this i think is a pretty practical alternative.

When is firefox going to support services?

Storage and cruft: yes yes yes to the entire 7Mb pdf for one quote… What I need is the actual quote, but then later on I need to know where it came from. Together will let you drag-and-drop a text section from any web page, and even if it’s just a few lines, the URL is imported along with the text. This is really very useful. I wish Devon THink did something like that. Hell, I wish Tbx did something like that, though Keith has already explained why that’s tricky.

Michael, you lost me. DevonThink has always been able to keep the associated URL of some text clipped from some webpage. That is actually one of its strong points, I use it all the time. Like you, I wished that Tinerbox had this feature as well, though.
Or maybe I am mising something here.

Lordy lordy, so it does. So it is. I am a fool. I suppose there must have been some point in its distant past (I’ve been using it since before v1) that it didn’t keep URLs, and I just never looked again.

Mind you – I just did look again – DevonThink will only do that if you clip to it using the supplied AppleScript. If you drag-and-drop, no source URL is saved. Together manages to grab the URL from any bit of text you drag from the browser. Neat.

At the risk of thread drift, I’d just add that I’m currently working on a G5 20" iMac and Together is unspeakably slow. With a modest collection of about 1,100 items, it uses up to 95% CPU, every byte of RAM it can find, and is still like molasses. I’m a bit disappointed. Tinderbox, with the same data put in as individual notes, is several orders of magnitude quicker even though it is just one vast XML file. I wonder why?


you most certainly are anything but a fool. Considering how much I have learned from you, Amber and the generally enlightened ones, I am glad to have returned a favour. Actually, after writing the post above, it occured to me how much we have accepted that in our digital age the pace at which features change has changed the perceived duration of time. Forever, ha. Since 2005 which is when I switched to the Mac and explored the exciting world of information apps.
Last year I went to a small, unsuspecting museum in Uppsala, Sweden, and to my own astonishment found the most excuisite collection of eqyptian mummies and hieroglyphs, Inca knots and what not. The structure of these information tradition systems have changed so slowly that societies evolved away from it. Rather amusing (and disturbing) to think that it is exactly the other way around today. People even get bored if a program does not evolve quickly enough over time.
I am not one to develop instant hysteria over digital amnesia, but I suspect I am growing old enough to get drawn to plain text systems such as Mulitmarkdown and (why not) plain text. Writing languages other than english, even the question of encoding this plain text (UTF-8, ANSI…) can make this seemingly trivial task interesting.

Back on topic: I usually just highlight the text of interest and hit the shortcut (Cmd closing parenthesis) and it gets tucked away along with the URL.


In Devonthink, using the hot keys listed in the DT Services menu makes clipping completely painless. I have DT setup to clip straight to an Inbox, and anything of note gets sent there with the Cmd-Shift-) hot key combination (or, in Mail, the “send to Devonthink” command under the Messages menu.
Fact is, I haven’t found anything like Devonthink for absorbing information. I’d like file system integration, like everyone else, but I wait and trust for the wizards of Devon Tech to produce v2 and resolve all the issues that v1 users have been whinging about.

and wait … and wait …

I’ve read somehwere on the DevonThink site that the resources that had been devoted to developing a mystery DT spin-off have now been returned to the core task of DT 2.0.


Amen for that. I wonder what Tinderbox would be like now if Mark Bernstein wasn’t slogging away on a Windows version. It must be a tough call on whether to diversify the product range and widen your sales opportunities, or focus on making the core product outstanding - and accept that it will have a limited number of buyers, many of whom will gripe furiously over upgrade pricing …

As a user, I prefer the latter model, as exemplified by Scrivener.