Key-iP Compact (pronounced keep, the iP stands for iPhone) is the fix for your finger-aching, composition skill destroying virtual iPhone keyboard.

Slotting either side of the iPhone, Key-iP creates an ergonomic split keyboard, with the phone in portrait mode in the middle. Learning from the wonderful portable palm keyboards, a simple grasp on the sides of the board unhinged it into two folding flaps. Either keep the iPhone sandwiched in the middle, or slide it out. In folded mode, the keyboard is only slightly bigger that the phone.

Pulling power from the base of the handset, the keyboard requires no batteries, and communicates with the phone via a very low power Bluetooth signal. The updated Apple spec for external devices allows the Key-iP to work without the onscreen keyboard - more screen real estate for editing.

Arrow keys, shift, caps lock, copy, paste and numerical keys. No batteries required, and so light you won’t notice it in your bag. Ideal for writing on the go, finally killing off your persistent netbook temptation.

You want to write seriously with an iPhone? Time to Key-iP it real with a Key-iP Compact.


Open data scorecards are SeeClickFix for government data websites.

Let’s say you are interested in finding out how much money Christie spent getting elected Governor of NJ? After some Googling, you probably find the NJ Election Law Enforcement Commission website and the section with public filings and campaign reports.

But wait! No RSS feed? No feed opportunities of any kind? Great work with the full data download as csv though. Reach for an open data scorecard and rate what’s there and what is missing. The scorecard uses the Sebastapol Eight open data principles, phrased to make applying them to a website easy - access restrictions? recent data? feeds? machine readable files? Click click click submit.

Maybe you are more interested in the Nassau Executive race. The Board of Elections site doesn’t offer any data - “filings are open to public inspection” only means that printed copies can be viewed. Time for another scorecard!

The scorecard website aggregates all the results, showing the cumulative build-up of issues with any single site. And everything is republished through an API, so others can build tools from the scorecard ratings - making State of Data overviews easy.

Open Data Scorecards: aggregating the data smarts of the crowd, and letting frustrated data searchers express their displeasure, all for the common good.


Task lists always assume that “today” is before midnight, and “tomorrow” is after. In the small hours, scheduling new items becomes tricky - do I really mean today, or tomorrow? Is that task really overdue or am I getting to it later in this session? Wasn’t I going to do that tomorrow? Wait, is it tomorrow already? But I’m not done!

Introducing Adjustable Midnight: click Settings, then choose when the day ends. It’s your working day, don’t be hostage to the rigors of convention.

Applies to Remember the Milk, Todist, etc. Adjustable Midnight will be the One Click of task lists. But free.


design problem
hack needed
2009.10.30 // 1256950620

More on flickr.
Design problem.

Censuous - a beautifully designed, non-technical web interface to pull census info into kml or shapefiles.

Getting data out of the census is non-trivial, even when you know your SF1, block groups, GEOIDs, H04001, etc. The barrier to easy data access is still slightly too high - the place-based summaries are great, but mapped data requires just a bit too much knowledge to be really accessible.

This one is such a no brainer. No need to describe it further.


EmDeG (Exhibit my Delicious Graffle) combines three fantastic tools into one killer web app for making sense of data - here’s how it stacks up:

So why is EmDeG useful?

Let’s say you are collecting data on various map sources or perhaps open planning tools. Delicious is great for the first five or ten items, but then things start to get complicated - it’s hard to see all the items or have any kind of structure, and dynamically assessing what you have is difficult.

Enter Exhibit - a fantastic and under-used tool from the Haystack group at MIT. Filter your list by keywords, search dynamically and add grouping headers. Suddenly a long list becomes more manageable. It works well for lists of like and non-alike items - see this example of student groups - but doesn’t work so well for dynamically exploring data in the style of the all-conquering pivot table. Perhaps you want to view data by state, or source, or year.  Facette does a neat job with these first two steps, but wouldn’t it be great to view your data in a graphical layout?

And for flow charts and beautiful data layouts, nothing comes close to OmniGraffle. Wish it was cheaper and had a windows or web version, but these are side comments. For making charts of complex structured data, it’s really superb. Neat as graphviz iz, izznot OmniGraffle.

So why put these tools together? EmDeG takes the best of all three, giving fast exploration of structured data, even when the structure might be distributed across many tags (think state names, platforms, programming languages). And the delicious foundation means that everything is dynamically powered by a widely used data collection platform - I put a link into your inbox, and all your org charts update automatically. Putting information into flow charts and organization diagrams leads to meaningful visualization, beyond long list syndrome. Grab tags to make them headers, or to cluster items into blobs.  Of course, as a web tool, you get wiki-like collaborative editing and sharing, urls to view charts as an image, etc.


Meet Bethel: the first operating system with versioning and complete metadata, including intelligent linking between files.

For example, when you make a jpg from an illustrator file, Bethel knows that these files were once linked and maintains this link into the future, even if a file is moved.

When saving, Bethel invites you to enter useful information alongside the file name - a project number, perhaps a short description. You don’t need to do this in some parallel system of a wiki or a content tracking spreadsheet - it should be intrinsic to the file storage system. Lots of metadata fields are available, Bethel offers the ones you most commonly use - and gives the option to require documentation if you decide to be rigorous.

And then all the links and metadata are indexed and viewable by others. So you can search for a file made for a particular project, no matter who created it, and pull out a previous version from the integrated smart versioning.

Bethel also considers workflow essential to the creation and storage of files: if you do analysis or GIS, and create a final document, it probably contained data from other files or sources. And if you change a file early in the process, the subsequent files need to know about it. So in Bethel, just click on a file to see how it fitted into the process, and see its parents and children.

Returning to a project after a couple of months? Grab a file and switch to wiki-view, where comments are displayed seamlessly alongside file info - creating documentation from files on the fly, with details of who did what when, and enough pointers from them to describe why. See what else you did that morning, searching right inside the document.

Bethel. Because you’re sick of copying and pasting file paths into half-maintained documentation. Because you have a bad memory. Because files are not isolated but belong in a rich workflow. Because other operating systems fail on collecting all the metadata we leave behind.