452 words on Software
While the weekend’s post on graphing focused on a lost opportunity in data presentation, today’s looks at the creation and presentation of data. An amusing – and thanks to Google and AppleScript quickly acquired – set of data would be Apple’s build numbers of their upcoming MacOS X.5 operating system. Whenever a new build is released, ‘rumour’ sites are quick to learn about it and put out ‘reports’ and screenshot galleries. In fact, frequently they are so good at their job that even people who have legitimate access to the builds learn about them through the magic tubes of the internet before the news make it through the expected channels.
Be that as it may, it also means that the somewhat obscure build numbers Apple uses (think 9A343) are widely mentioned on websites. They are also relatively unique, hence easily googlable. So if you want to know about which builds knowledge made it to the outside, you can just google all terms from 9A1 through 9A600, say, and look at the result count. You can then feel useless but amused after drawing a graph from those values and try to guess which builds Apple sent around the world. Might save the nausea that visiting ‘rumour’ sites causes.
Of course I’ll be the first to acknowledge that this is far from scientific as it is completely unclear how reliable Google’s result count is and how the number of search results for such a term changed over time. But the results correspond quite closely with what aforementioned rumour sites report.
And I’ll leave it at the discretion of the reader to draw conclusions on how the web works and how worthwhile it is when tens of thousands of pages are created within days – and duly indexed by Google –, all of which contain sorry ‘news’ about unreleased software. Even worse, all of them contain exactly the same news about said software. Going all the way to straight copy and paste jobs or some automated equivalent of them in many cases. Now tell me in which way a quarter million web pages on the same terms are better than two…
The graph I managed to kick out of Excel isn’t particularly good. But I challenge any Numbers fanboy to show me that Numbers isn’t complete crap on this rather trivial example: I give you a few hundred numbers, you create me some sort of reasonable diagram representing them. Simple as that. You’ll find that Numbers’ performance is abysmal even with such a range of values. It will stall for long periods of time while using around 300MB of RAM. For a few hundred numbers and a single graph, that is. The future is now.
Received data seems to be invalid. The wanted file does probably not exist or the guys at last.fm changed something.