One of my favorite gadgets is Gullible Info, its always on my Google Home Page. Its displays things like:
«Japanese scientists have created a mechanical version of a human hand with similar motor skills; it requires a computer 1,500 times more powerful than a standard desktop computer to function.»
The problem with this is that this kind of information takes a month to be outdated and becomes misleading in some cases. In this case, I don’t know when the “Japanese scientists” developed the mechanical human hand, so I can’t really be sure how many nowadays standard desktops would be needed. Sure, this isn’t a real deal to me because I’m not into robotics R&D but, shouldn’t this kind of information be updated? If so, how?
Another example of this problem is when you read some blog post or news article, it usually says something like “Today…”. Well, we do have the date on the article, but as shown by the first example, that isn’t always enough to place the entire data into context. It is also obvious that we can’t expect people to keep updating each and every article but still, somethings become more like history facts than news and need to be placed in context so future kids can really understand the data.
Taking the dates as a simple example, what we need is a tool to update articles daily so that it will show dates accordingly: today, yesterday, this [weekday], last week, etc… and eventually “on [month], [year]”. This is pretty simple, and in fact I would be surprised if there aren’t tools and content management software already doing this for dates. Yet, I never noticed any evidence of its existence.
Another case, and the reason for the example in this article, is much more complex. A computer knows dates very well, but how can it know whats the power of “standard computers” today? And how can it compare them to the article’s computers? To be fair, it is near impossible, we can’t flatly compare 2 nowadays processors because each technology advancement provides improvement over some tasks but not another. Still, a pretty close estimate can be done, for example, using the number of instructions per second.
In this case, this kind of data can be easily collected, a bot like google’s indexes many computer stores, so it can calculate the “standard” computer for today and also has access to the processor manufacturers homepages, enabling a very approximate estimate of nowadays standard computers power. In the end, all that is needed is a webservice to provide this kind of data in a way that we can easily access and update articles with it, but how do we write them?
Writing this kind of article seems complex, it would require programming at first, but not some tools are extended to support this features. In my idea, something like Microsoft Word or TinyMce, which I’m using to write this article, can be extended to detect dates (or date references like “yesterday”) and change them so they can be updated by the article engine in use. Somethings need to be updated more often that others, for example, dates need to be updated once a day, or, since they are actually quite easily calculated they can be calculated on the fly.
On complex cases like “how many nowadays computers are needed to handle that mechanical hand?” we would need some more work. To start with, the editor wouldn’t be able to detect what is time-context sensitive data, so we would need to select it and click a button to edit the preferences to that piece of information. In this particular example, I believe the best way to setup the data would be to create an “Excel”-like formula, using the webservices that would provided updated date. If at the time of the article the writer would check the webservice to the actual power of one standard PC, he could easily produce a formula to change it whenever it was outdated.
At last, a standard computer power isn’t updated daily, its something like once a month, to either the writer could setup the update time or even the webservice itself would provide it. At a best case scenario, something like RSS feeds could be used to let the article engine know there were some updates.
Well, I believe that I made my idea clear, a mega database of some time sensitive data could help improve our understanding of history when we read something on the internet. Widespread of this kind of tools would enable even more historical data to be stored and used by future generations, or even ourselves, without the risk of misunderstandings related to different technological or social contexts to name a few.
[…] while ago, a little over than a year to be precise, I wrote a post called Time Sensitive Data. It outlines some basic ideas on how each time anyone publishes something online its highly […]