Comment Re:NYC Subway (Score 0, Redundant) 276
boss hurries you to work
Work hurries your boss.
&
wife calls you home
Home calls your wife.
boss hurries you to work
Work hurries your boss.
&
wife calls you home
Home calls your wife.
The only things that belong in the title bar are the close button, the dock button, and the zoom to max content size button on the left, the window title in the middle, and the toolbar button on the right.
Are you some kind of an Apple HIG fanboy? Is this a sub-cult of the Apple cult of some sort? The way Chrome does tabs halfway in the title bar makes perfect sense. This approach leaves more screen real estate for the content, while retaining the ability to grab the top of the window to move it around. Besides, Apple breaks it's own HIG quite often. iTunes, Mac App Store - those are the main culprits in the current version of OS X. And God forbid you from using the Address Book in Lion.
Microsoft managed to squash Netscape, BeOS
While I agree with you to some extent, the arguments you provided are flawed.
BeOS was really never anything more than a tech demo. At that time Apple was interested in buying it and using BeOS as the base for their next OS when the Copland project failed miserably. They bought Next instead, and used NextStep. Be Inc. just couldn't compete. Too bad, because it was the most advanced operating system at that time.
Netscape was squashed by Internet Explorer, but it was not because of Microsoft's evil voodoo practices. It happened because IE was a far better product on both Windows and Mac, and to some extent even UNIX (yes, there was a UNIX port of IE 4 and 5).
It's almost impossible to resell, unfortunately. Who buys it on the street, or would trust you to cut it?
Hmmm. I agree that there are some poor design decisions but people that are in the physical sciences tend to only know Fortran because most of what they do is number crunching and, until recently, the Fortran compilers where the ones that got the best performance on any kind of numerical computation, and there's also 40+ years of libraries that they can draw upon. At least it's Fortran90.
However all the references in the linked article appear to be data preparation and not the actual modeling. Now certainly GIGO can be a problem. It does sound like CRU should hire somebody with a Data Warehouse background to do some decent conformal mapping and set up a data quality management framework. Actually it sounds like something the whole climate research community should get together and do. But while I haven't seen the code and only have the followed the link you point to, it's not clear to me that it applies to the actual computations doing the climate modelling simulations, which is where they would presumably primarily focus their attention. Again data quality could be a problem but, since it's for a large stretch of consecutive years, I would expect its main function is to compare with retrospective predictions. In that case the researchers may have considered this a necessary distraction from what was really important: the actual modelling. This may be prototype-quality code that they felt didn't need to be polished because it wasn't time-critical and provided sufficiently accurate results given some statistical analysis of the results. The real question is what does the production simulation code, the stuff that needs to be fast and right because it's iterated millions of times, look like?
8 Catfish = 1 Octo-puss