Comments on Doing Today's Job with Yesterday's Tools

Margaret Rasor — Thu, 22 Mar 2007 19:06:41 GMT

I agree that there should be one way to access all types of data, and also that it should not just be about the platform but can work across various platforms. Someone else mentions the Raskin work, it’s definitely worth looking at – sounds like people are doing what should be done – responding to the need by pushing through the current limitations and barriers and taking it to the next level. Very interesting, and long overdue concepts!

Avi Soudack — Tue, 12 Feb 2008 16:44:46 GMT

Patrick,
Thanks for raising this. I was taken by your point about how different applications are mini-environments that we have to negotiate between. What a waste.
If you haven’t already, have a look at Jef Raskin’s The Humane Interface. (Summary here: http://jef.raskincenter.org/humane_interface/summary_of_t… He argues for abolishing: file names, folders, the distinction between files and applications, and anything that makes you switch modes (your uploading photos example comes to mind). He’s for: ways to use humans’ innate spatial wayfindng abilities and focusing attention on the task at hand. His zoomng video space metaphor is really fascinating.

p.s. His ideas are being actively pursued (http://rchi.raskincenter.org/index.php?title=Home) though I’ve only just discovered this.

Rhett Brown — Tue, 12 Feb 2008 16:45:19 GMT

Some branches of inquiry that your article inspired when I read it – that’s a compliment, by the way ;)

1) When discussing information accessibility and portability, one invariably strays into “universal object format” territory, and some interesting thought has been done on that subject. We have different formats – because we have different applications – because we have different users – who have different ways of thinking about the data they care about. Is there a useful meta-meta data format that will allow ANY information to be uniformly described, stored and delivered in a way that is useful to ANYONE? To realize your vision, you need that.

2) Personal data are weird – they don’t organize exactly like objects and tasks in the physical world. With data, we use applications and tools in part to achieve useful groupings (“Recent documents” lists are a good example) but we don’t find the hole we were digging last week by grabbing the shovel we used to dig it. It is somewhat intuitive to group information by what we do with it and the tools we use to manipulate it…

3) Are tags and labels universally useful? Maybe, but there are other ways to approach organization. In a perfect implementation of what I call the “Google Mail Vision”, you would have only 2 classifications for your data – “visible” and “not visible”. When you want to see data that is in the “not visible” state, you search for it to make it “visible” (in a search results list). What if the data itself does not contain search-able bits that are intuitively useful for searching on directly (e.g. compressed media, or music and picture files)? Currently, you bandage on a fix in the form of a tag or grouping that will immediately come to mind when you are looking for those data. But one day, your search tool might be smart enough to fingerprint and relate your data across many dimensions (e.g. English language and audio, emotion and imagery, etc.) and show you useful results without you having to pre-organize it. I see the benefit in combining that with tags, but there is a point of diminishing returns. You end up searching for tags and within tags anyway…

4) In my opinion, this brings into focus a somewhat perpetual issue in the development of humanity – we have and always will advance our ability to collect more and more information. We then develop methods for storing and processing that information, and as a species, one of our primary methods is specialization within societies. But the kicker is that until an individual can physically exist forever, he or she will always be forced to decide what to store and process PERSONALLY. Thus, we must each devise a unique methodology that implements certain tools and information taxonomies, spending time to organize either in front of the process (i.e. creating folders and tags for browsing) or on the back end (i.e. searching). In my HUMBLE opinion, this means the affliction of disorganization is most effectively treated by pruning the set of data you are trying to organize. That strategy even works great for life away from your computer!

Thank you for the article…

Patrick Dubroy — Thu, 01 Mar 2007 22:13:00 GMT

BumpTop is definitely cool. I would say that it is approaching things from the opposite direction though. I am interested in creating in the semantics of the objects, in defining a sort of vocabulary in which you can describe and interact with them. If this were in place, then BumpTop (or any other interface) could be used to manipulate and organize the objects.

Dante Murphy — Fri, 02 Mar 2007 18:49:18 GMT

Nice article, Patrick. A very good conversation starter on a ubiquitous problem.

Having just seen the BumpTop demo I can’t help but think that there’s another dimension at play in achieving peak productivity, in finally getting “organizized”. And that is to take al the rich information associated with single objects and begin to apply them to groups of objects, either permanently or as temporary instances.

If I envision the BumpTop dektop metaphor as a portable workspace (which I do, because I am a dreamy sort), then I can use their “pile” interaction to create homogeneous collections of objects and begin to apply enriched labelling and metadata to this temporary collection. Because BumpTop is a platform already (albeit for tablet PCs only), the only remaining hurdle is to extend the object model to include all the whiz-bang features of Flickr, del-icio.us, Gmail, etc.

Who knows when, or if, this will happen, but after years of wandering in the same dark forest you portray I can at least say I have a vision of one possible solution.

Look forward to reading more about everyone else’s.

Jordan Frank — Sun, 11 Mar 2007 18:55:30 GMT

In reference to a Burton Group report on the subject, I blogged about the importance and utility of hypertext as the core for interactive and compound “documents” where, in essence, the idea of a document dissolves, replaced with a set of mechanisms for joining pages and data in the manner that is required given a certain context.

In the context of project requirements, you could have hundreds of separate pages that you need to allocate and re-allocate across milestones, then produce “document” views which slice the requirements by milestone or even by status.

read the full entry here

By the way: 1 month ago, my inbox dropped to 2 messages. A business trip caused me to let it slip into the 20s but I will get back down to 1 or 2 by end of week. My tools:

1) Traction for all the text content that I need to remember, share or track/manage
2) ACT to record contact actions and next actions.
3) A spreadsheet for a few simple needs like managing my trade show calendar and tracking sales

Patrick Dubroy — Wed, 28 Feb 2007 17:23:31 GMT

Kevin, I think you’re right—I do see Microformats as being part of the solution. As I mentioned in my earlier comment, there are several semantic web technologies that could be helpful in this area.

Vincent Clark — Sun, 11 Mar 2007 18:54:49 GMT

What you are proposing exists at the application level as General Knowledge Base including multiple classification, search across object types, etc. I enjoyed your article and will follow this thread more closely.

Vincent Clark — Wed, 28 Feb 2007 13:26:15 GMT

What you are proposing exists at a the application level as ">General Knowledge Base including multiple classification, search across object types, etc. I enjoyed your article and will follow this thread more closely.

Kevin C — Fri, 02 Mar 2007 18:49:56 GMT

I wonder if microformats will offer any solution to what you’re describing. It’s more web-based and still emerging, but does have some potential.

I just found a decent presentation on them today: http://www.xfront.com/microformats/Microformat-Demonstrat…

Clifton Evans — Sun, 11 Mar 2007 18:54:18 GMT

Finally, someone managed to sum up at least a good half of the usability problems with the desktop metaphor! And, without going into the somewhat abstract Sematic Information Modelling Paradigm Logic (SIMPLE), again.

I particularly liked the quote: ”...is like a dialect private to an isolated town: as soon as we leave, we are forced to use the computer equivalent of grunts and hand signals.”

Great example! What isolated towns usually have is a language that is common with their neighbors, so they can trade photos, money and email! I did quite a bit of time on languages and communities in my masters so I’m very happy with this metaphor. My only worry is that some languages will die out if there is too much ‘extensibility’ (like OpenDoc) if there isn’t any support in the surrounding ‘villages’, the languages will disappear, perhaps even faster than languages die in the real world.

So, what’s the practical option? Want to make one??? Should we?? A village-to-village platform? VTV? I’d be happy as a dog in a field running from village to village to be a part of it.

A good place to look for starters, and perhaps what not to do, would be Haystack, via MIT:

http://haystack.lcs.mit.edu/

Now, they’ve been barking up this tree for a while, but something about it doesn’t quite sit right with me. I don’t think they are heading in the right direction. If you want to chat about it, email me, I’ve got tons of ideas on the back burner waiting to become soup de jour!

CD.

Matt Johnson — Sun, 11 Mar 2007 18:53:50 GMT

Very good article. I have been having trouble keeping it all straight. I need a person mashup server. Anyone got any good ones? Tried PageFlakes.com, but then too, it can’t view anything on my person computers.

One of my co-workers did suggest Google Desktop. It took it about an hour to index one of my machines, but it has made a lot easier to find that one document that I created 5 years ago. Then I use targeted web searches to find my information on other sites (ex. Hex site:del.icio.us).

Anna Rouben — Tue, 27 Feb 2007 01:49:45 GMT

Thank you for an interesting article. I agree that organization of data is far from optimal. I think integrating web with desktop would be quite useful but very challenging at the same time. Having desktop be linked with web resources might raise some security issues. Combining several applications into one can create some bulky products with some features that not every user wants.
But I definitely, agree that there is a place for improvement. I use IPhoto to organize pictures on my computer and use Flickr uploader to upload the pictures. Finding an album in the IPhoto folder can be tricky because the names of the folders are different from albums’ names; the folders are numbered. Kodak offers a desktop application that has an easy synchronization with the online Kodak albums. As Patrick mentioned there are still problems with compatibility; I can use Kodak to synchronize with my Kodak album but not with Flickr. It would be nice if there were more standards and online applications could talk to desktop application independently from the brand. I can also see Web Services as an option to achieve this goal. I am not sure that the companies are interested in such friendly compatibilities though, i.e. Kodak might loose customers of Kodak online galleries because now these customers can easily work with Flickr.

Patrick Dubroy — Sun, 11 Mar 2007 18:49:20 GMT

You’re right, the desktop OSes are finally giving us some better ways to manage our information. The problem is, it’s too late. For a lot of people (myself included), our personal information isn’t sitting on one computer. We’re using web-based email, Flickr, Del.icio.us, etc.

I deliberately didn’t go into that much technical detail in the article, but I think the real solution is going to need to deal with data that is distributed across many machines and web services. It’s really almost just an extension of semantic web ideas.

Kevin C — Sun, 11 Mar 2007 18:49:10 GMT

Thanks for the good discussion of the issue. I think at least some of the issues you’ve mentioned have begun to be solved by next generations OSes. OSX, and Windows vista support, things like tagging and virtual folders which are akin to what you’ve mentioned. They’ve also allowed search to be more powerful so it doesn’t just deal with files. Search can return browsing history, emails, contacts etc…

Here’s an interesting demo of Vista’s search and tagging facility: http://www.istartedsomething.com/uploads/search.mov

There’s a good bit more work to be done before it comes to the level you’re talking about, but some of the ideas have begun to be incorporated ta least.