Where is Tinderbox heading?

It’s cat amongst the pigeons time.

Doug Miller and Jeffrey Radcliffe have both commented recently on their (admittedly rather mild) concerns about Tinderbox. These were prompted, I think, by some of the new features in Tiger, but they struck a chord with me for more fundamental reasons.

Tinderbox has always been a “quirky” programme and, if one may be forgiven the comment, it bears much of the stamp of one man, Mark Bernstein[2]. That is inevitable, perhaps, but it can tend to make the programme look a lot like a pet project. Compare that with, say, DEVONThink which looks more like a committee production.

The trouble is that the very success of Tinderbox tends to militate against its retaining this “one-man’s product” feel. I am conscious that this may sound harsh on Mark, but one does detect in the way the product is developing that there are two connected strands:

a) First, a rather techie, Puritanical approach to the look of the thing - “it does what it does very well (and nothing else comes close), so it doesn’t matter what it looks like”; and

b) Secondly, a slight resistance to implementing features that are (i) common enough in modern applications and (ii) actually quite useful (I am thinking about the implementation of Services, the failure to implement which severely hampers Tinderbox, IMO).

But the fundamental problem is that Tinderbox is a weird product - while it can do very simple things like outlining (as Ted Goranson’s exhaustive analysis shows), at heart it is a very ‘techie’ environment that calls for a lot of time and effort to get the best out of. In fact, I’d go further than that - it demands a fair degree of effort to get much out of it at all beyond the very simple.

This is fair enough if you are going to “live in Tinderbox” but I find that I like to use other things as well and to interchange data between them. One of Tinderbox’s much-vaunted features is the ability to export data to text and to HTML - it is good, but it demands effort to set it up and is likely in my view to put off the casual user.

Which brings me to the question of Windows.

I am very far from convinced that porting Tinderbox to Windows is going to be a winner. The sort of mindset you need to get the best out of Tinderbox is the sort of mindset that tends to use UNIX-based systems [Sweeping Generalisation #1]. I cannot see most - or even very many - Windows users getting to grips with Tinderbox. First, the non-techie types are likely to find it much too daunting at the advanced end, and at the simple outlining level there are already enough things that do the job and are (much) cheaper and/or free. Secondly, the WinTel tech types tend to be into .Net and Java [Sweeping Generalisation #2]. There tends not to be that sort of open engagement that is called for to put your trust in Open Source and/or things from small software houses like Eastgate [Sweeping Generalisation #3].

So where would that put Tinderbox in a Windows world? Hmmm. I think that it is likely to be a largely wasted effort. Look at something like ECCO, which some of you may remember from years back - it was a very different sort of Outliner-on-steroids that was killed by Microsoft bundling outlook with everything, yet it still retains a loyal user base and indeed people are still writing new scripts for it. The point is that ECCO was quirky, but less quirky than Tinderbox, and useful in a way that I have never been able to replicate in anything else (including Tinderbox, incidentally - though no doubt someone will prove me wrong, but then that brings me back to my point on useability: I simply cannot face jumping through all the hoops to try). Yet it still died, because it was too quirky for a mass audience.

To be honest, I cannot see Tinderbox making it in this world - after all, if it is not exactly widespread in the Mac world (and I have no figures as to registered users), it is hardly going to take off in Windows. And yes, I know that having 0.1% of 95% of all computer users is better than, say, 0.5% of 5% (those are indicative figures only) would be better in cash terms, but I somehow don’t see it happening. Tinderbox is essentially a personal information manager; it’s networking capabilities are not great and that’s where Mark would make a killing - imagine GE suddenly buying licences for all its employees… And I just cannot see hordes of Windows users rushing out to buy Tinderbox[1].

So where do I think Tinderbox should be going? Well, here are my modest proposals for some useful things for the end-user:

• Update the whole look of the thing - even if Mark (and others) think it is unnecessary and perhaps superficial to have the program looking modern, it does need a radical overhaul to bring it into the 21st century. I do not suggest that this should be at the expense of functionality, but can we not have beauty as well?

• Write a decent manual - I am afraid that I have never been persuaded by the line that “it does so much that it is pointless to tell you all about it so take these indications and discover it for yourself”. It does need a proper manual and it is possible to write such a document.

• Bring in support for Services. Please.

• Improve/organize the wiki pages - they are horribly tangled at present and it is impossible to do anything except stumble across pearls of wisdom.

• Share the ideas from Tinderbox weekends more. I cannot imagine that these are enormous money-spinners for Eastgate, and it would help people buy into the Tinderbox ethos to be a little more open about what it can do. And no, the File Exchange and wiki do not cut it (not least because they are infrequently updated and rather random).

• Have a supported forum. There have been one or two efforts, I recall, but every time I checked in, I found little activity.

• Cut the price. Notwithstanding those who say that Tinderbox is to words what Photoshop is to pictures (sorry, I forget precisely who said that), the cost is too high. And I think that the whole idea of paying an annual fee of $70 for updates is off-putting and potentially unfair (suppose you buy just after an upgrade). I know there is nothing directly comparable, but most of the “kind-of” competitors are significantly cheaper (the DEVONstable, NoteBook, NoteTaker) or even free (SubEthaEdit, VoodooPad, Instiki etc). Oh, and make the price consistent - while it may be nice for those in luck, there is little as teeth-gnashing as buying something full price and then finding that, 10 days later, you could have bought it for 60%.

These are just a few thoughts. No doubt others will have their own suggestions and a fair proportion of users will probably tell me that they disagree with everything here. Fair enough. These are just one user’s experience over about 2 years of using the software.

Ultimately, Tinderbox is too good to be a hobbyist’s tool; it deserves wider recognition and I submit these thoughts and suggestions respectfully to the developers in the hope that it will at least engender some debate.

Update: [1] A few hours after I write this, Mark mentions that anxious Windows users are pressing him for news on the progress of ‘Tinderwin’, so obviously there is some pent-up demand in the Wintel world.

[2] (the following day) Richard writes appositely that this “is almost certainly true, however I have to say if you’re going to have that kind of influence then Mark is a sure fire choice to produce something worthwhile“. I entirely agree - one does not want an anodyne ‘product’ and it is only thanks to visionary (and highly talented) types like Mark that one gets true quality in all walks of life. Look at how photo.net slid downhill after Philip Greenspun disengaged. Horribly bland and uninteresting.

Related posts:

1. Tinderbox (reprise)I made some comments about Tinderbox the other day. Judging...
2. TinderboxingI have re-written the whole of the static part of...
3. Virtual dimensionJust found this little program which neatly (for a Windows...
4. “Operating system”Lovely. I was looking at the mini-itx pages and saw...
5. Stationery in Mac OS XI found this useful, on the Tinderbox wiki, though as...