Saturday, November 10, 2012

XMind: Impressions and comments on the mind mapping market

It's been two years since I first looked at XMind. During that time I used MindManager at work and experimented with MindNode Pro at home. I mostly use the tools to explore new terrain, and as a visual aide to some teleconferences (share the mind map while discussing).

MindManager wasn't ideal, but it was a decent tool when we could buy it for $100 or so. Their current pricing is too high for team use, and I really did want the option of sharing maps. So when I switched projects I also switched to XMind. I don't have time for a proper review, but I can share some bullet points on why I chose it, what it's like, and what I would love to see.

Why I chose XMind

  • It runs on Windows 7 and it's nice I can also use it on my Air.
  • Price: Free for a very solid version, upgrade to pro was $80 for me. I don't like free software, but we can't afford MindManager - so this freemium model is a good balance.
  • Longevity: It's been on the market for several years and just went through a significant update.
  • Quality: it's got bugs, but it's tolerable so far.
  • It's a simplified clone of MindManager so it has a good feature set.
  • The base version is "open source". A weak form of insurance, but could be worse.
  • Freemind lacks the corporate look and seemed a steeper learning curve for non-geeks.
Impressions, including problems
  • Data lock: The inevitable for all but Freemind
  • Java: The UI is native, but the back-end requires Java. That's bad enough on Windows, but for a Mac user Java installation feels like installing a malware-welcome sign.
  • There's no built-in Help, only web help.
  • It is slow to load what I consider a mid-sized map.
  • It is pretty reliable, but I have run into a significant bug with string search. Search sometimes fails unless the map is fully expanded.
  • It's made in China, and the language localization is imperfect. "Extend" is used in place of "Expand" for example, and the mouse-over tooltip text is quaint.
Thoughts on the mind map / concept visualization marketplace
I've seen cognitive-support apps come and go for twenty years, and I don't think we're making much progress. We're shuffling in place. This definitely isn't a technology problem -- we had similar apps running on the computing-equivalent of medieval tech. I don't think it's due to lack of imagination, though that has occurred to me. I think it's a business problem -- the market for high-end cognitive-extension concept modeling software is tiny; probably not more than 1 in 10,000 adults, perhaps 300,000 worldwide on all computer platforms. If we then ask how many can/will pay $30 a year for a product … we're talking a modest income stream for 1-2 developers owning a world market.
Yeah, this is a business problem. So we're not going to get what I want through traditional market-driven mechanisms. We're going to have to figure a way to grow something from modest means, and it's going to have to be built atop something else.
So here's how I think it could work. Start with the standard data formats used in other apps like Notational Velocity for the nodes. That means UTF-8 including "plain text", RTF, and markdown with a simple title, tag, date/time and text metadata model. That way the "nodes" can live in a simple Spotlight/Windows Search indexed folder and can be used by SimpleNote or Dropbox.
Now put the graph structure as XML or XMLized RDF in just another note in the same folder with a special name.
Optionally, allow the folder to contain other files, images, and so on (future).
That's the data. Now the app reads in the RDF and the nodes and renders the relationships. Ideally many different apps work with the same data structure. There's very little income here, so we're taking labor-of-love with a bit of cash to pay for a new computer. From this base, over time, with full data portability, we can slowly build a concept-visualization ecosystem with full data freedom.
Anyone have other ideas?

See also:


Matthew Blount said...
This comment has been removed by the author.
Matthew Blount said...

What about Evernote? The approach there is generally to take pictures of pencil & paper/whiteboard diagrams, and have Evernote OCR the text for you. Personally, I'd rather draw in a notebook than use Freemind.