Carrying on in the portable apps vein, R is rapidly becoming the de facto standard in statistical analysis software for research. Its command line driven so not for the point-and-click brigade but incredibly powerful and, importantly, all the algorithms are peer reviewed. If they get in to R then they have to be good. It therefore surprised me when I realised that I hadn’t looked for a portable version. And it’s not quite as obvious to find as I thought it might be, but Andrew Radd has put together a portable version in the portableapps.com vein. Well worth a download.
For those developing with Python it’s nice to have a (USB) portable environment to work in, particularly when you have multiple versions of Python you need to test in. I’ve written before about scripting ArcGIS with Python, but this equally applies to any Python programming (or scripting) you might be doing. Portable Python delivers the goods and is stuffed full of lots of software (including PyWin) in addition to Python itself. And there are 3 versions: 2.5.4, 2.6.1 and 3.0.1 which can be run entirely separately.
The ASPRS have been hard at work converting some of their old video footage of things remote sensing and turning them in to short “info-clips”. Its a worthwhile exercise and they have a YouTube channel where they can be viewed. Worth a look.
Mike Baker, the BBCs Education Correspondent, again has an interesting this week about changes to university admissions. In essence the government wants to move towards universities offering places based upon a student’s actual grades, rather than predicted grades (which aren’t that accurate). In a half-hearted move we now have a 5-day “adjustment period” in which students who performed better than expected have the option of “upgrading” and applying elsewhere. This sounds all very good in theory, but as Mike outlines the likely net effect (because “good” universities are already full) is that there will be no change other than getting the hopes up of many students.
I came across this Windows search engine recently called, not surprisingly, Everything Search Engine! Its a 300kb stand alone application that ties in to NTFS file indexing, allowing you to search for any filename on an entire filesystem. What’s amazing is that, besides being so small, it indexes your entire disc in ~1s. Results are instantaneous. A worthy program to have at hand.
The Shapefile 2.0 manifesto has been doing the rounds on a few blogs lately. It’s a nice article (and worth the read) simply because it identifies the the need for open file formats that allow us to easily work with data (and it makes the point that GML is a distribution format which is a different kettle of fish). This harks back to an earlier post on file formats to use for archiving data (at the Journal of Maps). One of the points Alex misses is that shapefiles do not contain topological data. So whilst SHP has its limitations, as he says, “in my opinion this makes Shapefile the best thing ever to happen to GIS, without it the GIS market would be a fraction of it’s current size.” Time is ripe for a replacement and I would agree that the file geodatabase would be a good option if it was made open.
Landsat 5 has reached the remarkable age of 25, having been launched 1 March 1984, and remains operational. In satellite remote sensing this is an unheard of feat, particularly when you consider that the design life was just 3 years. A nice article detailing the success of the mission has been written by NASA. This long lived success is even more important in the context of the launch failure of Landsat 6 in 1993. And whilst Landsat 7 launched successfully in 1999, the failure of the scan line collector in 2003 severely restricted its usefulness. That said the USGS reckons they have enough fuel for both satellites to operate until 2012, which is just as well as Landsat 8 is not due for launch until 2011.