Software’s Opportunity Cost

By | April 8, 2007

I’ve never seen this properly studied, and only rarely taken into account by software developers: the opportunity cost of committing to one service or program over another. In a word: Why is it software that’s in charge, not the data itself?

An obvious one is Twitter vs Jaiku. Which one to embrace? Jaiku actually has more features in a way than Twitter, but more people are on Twitter. And, perversely, because one of Jaiku’s features is being able to easily include your Twitter stream into Jaiku, it makes more sense to stick with Twitter as your main presence/communication service, since those updates will automatically feed into Jaiku. Jaiku loses out because it’s better.

But usually it’s a starker choice: choose one program or service over another, and you’ll find it harder and harder to reverse engines and try another. I’ve had two versions of this blog going, one on TypePad and one on WordPress, because I can’t decide which is the better service. It’s a lousy solution and often ends up confusing people and diluting the conversation. I haven’t committed to either yet, a makeshift solution made easier by tools like BlogJet, which allow me to post to both blogs, and the import/export tools that both blogging services provide. But it’s still a dumb compromise.

Worse is the commitment one makes to software. I love PersonalBrain, but I also love mindmapping tools like Freemind. And outliners like MyInfo. I also want to explore stuff like Topicscape. ConnectedText has potential too. But because I want to use them in the real world, with a real project, I don’t want to find that by committing to one I’m foregoing using the others. But that’s inevitable. There are import and export tools available to make it easier for these kinds of about turns (or occasionally starting out in one simple program and then moving the data to something heftier when the data gets too big).

But surely there’s a better way of doing this — by making data so open that we can easily move it between programs without these hurdles? Instead of the programs being the dominant tool, they become servant to the data? A case in point: I want to look through all the blog postings I’ve written in the past five years. I want them somewhere I can see them, but also some way I can index them, and view them in different ways. I want to be able throw them at a Bayesian filter to look at the language I use, the topics I choose, the arguments I present. I want to be able to view all the data as a big mind map, or a treemap, with the categories and tags as branches. I want to be able throw them at a Wiki builder so it becomes one big Wiki without me having to do anything fiddly. I want to throw all the posts into a PersonalBrain, where the links between articles turn into links between thoughts. Then I want to throw all my emails into the mix and see what pattern they make. I want to move between all these ways of looking and manipulating my stuff without me having to worry I can’t ever go back.

In short, I don’t want to commit to one program. I want my data to be in charge, and the programs themselves conform to the data, not the other way around. Perhaps this is impossible. But why should it be?

4 thoughts on “Software’s Opportunity Cost

  1. John Orford

    Open formats are part of the solution.

    Committing a lot of time and energy to closed formats will more likely end up in tears.

    Reply
  2. Argey

    Jeremy,

    Topicscape (thanks for the mention) is firmly set on the interoperability path. It can import from Personal Brain and, via text, from MindManager but that was the start. The latest Beta can import from FreeMind and re-export to FreeMind, it can export to OPML, or read OML (from ShadowPlan, for example). It also has an XSLT translation function to allow the tech savvy to import and export to other XML formats. But now we’ve built the capability to handle XSLT, Topicscape will be adding new interfaces, with the intention of supporting two-way connections (and a round trip where possible) between many other types of mindmapping and information management software.

    Different circumstances mean different needs, and the ability to switch software as needs change will distinguish the useful tools from those that try to lock users into a proprietary standard. The most obvious example is where a small 2D mind-map is friendly, easy to read and often just what people want for the start of a project, but as the project grows and the mind map with it, it gets unweildy. And when there’s a need to organize information according to the mind map without making the mind map large with every added file or note, then switching to 3D is the approach that Topicscape takes.

    if small sub-projects emerge later, then a topicscape can be copied, pruned, converted back to a 2D mind map and passed to someone who doesn’t need to see all the detail that is not relevant to them.

    Interoperability rules!

    Reply
  3. andrew

    [While I’m the sysadmin type at jaiku, this is me with my opinion, not official jaiku-speak. I look after the servers, not the business decisions. Just for the record.]

    You say ‘Jaiku loses out because it’s better.’. Sure, we lose your presence, but we gain your content (via RSS of your twitter feed) for our users.

    I think it’s more important for services like Jaiku and Twitter to grow the overall market, rather than fight over who gets a larger share of it.

    I firmly believe that more inter-op is the a vital part in how we can do that.

    It would be very easy to end up with umpteen different nano-blogging/precence/other-buzzword services, none of which really talk to each other, just like the IM world. No-one wants that.

    Reply
  4. economics

    We have to more emphasis about the free resources of software………like………
    Capital cost reduced (exemption from payment or low cost of the software) ….
    Independence with respect to an editor …
    Evolutions controlled by the community of the users…
    Moreover, the free software is often characterized by the reliability and the quality of their code, like their conformity with the standards and a good interworking.
    Lastly, the availability of the source code facilitates the development of new functionalities, the adaptation to the technical context, or the correction of the anomalies.

    In spite of these assets, the adoption of the free software is sometimes slowed down by several factors:

    Difficulties of choice (in certain fields, there are multiple open implementations source)
    Complexity, in certain cases, to evaluate in an exhaustive way and objectifies the advisability of resorting to open solutions source instead of solutions owners
    Difficulty in evaluating perenniality, the robustness, the performances of an open solution source because of the difficulty sometimes of obtaining sufficiently reliable and precise information of the community or others utilisateurs*
    Lastly, since one plans to deploy on a large scale of the solutions based on the free software, the question of obtaining contractual engagements about the quality of this software arises and of profiting from support in the event of incident – services usually obtained from the editor when a solution owner is chosen.
    The open offer source of Atos Origin has vocation to make it possible our customers to fully benefit from the benefit of the free software, while raising these obstacles

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.