Re: Desktop monopoly
I think Tom 7 made a valuable point. As for mugs who have to operate these corporate computing dung heaps "not caring", I disagree with the sentiment that they shouldn't care. The plebs bashing away at their work PCs should care that because vast sums of money are making folks like Ellison and Ballmer richer when that money could be making them richer instead.
Tom hit the nail on the head with his observation that often the software used is a poor fit for the problem at hand. One particular system from recent memory comes to mind... It was a .NET WCF service that used Coherence as a it's primary storage to stream a bunch of updates to a bunch of unreliable remote clients, this thing had burnt $zillions of licensing, required a bunch of very expensive servers to cope with it's memory consumption, and had a bunch of expensive SSDs slotted in to work around Windows' poor file I/O performance.
While bored and frustrated waiting for this large expensive and complex system to work, I read the requirements + interface specs and hacked up my own version within a couple of hours with some Python scripts and an Apache webserver running on my cruddy 5 year old Linux ultra-portable laptop.
The resulting hack did the same job faster, burnt less network bandwidth, consumed less memory, was more reliable, more robust (ie: a client going offline didn't kill the whole system) and cost about 2 hours of my time to implement (vs. 6 people taking 8 months). My code was a lot easier to maintain too, < 1K lines vs >100K lines. ;)
The guys implementing that system were not stupid or bad at their job, quite the reverse in fact. The root cause behind this expensive nightmare was that the *Requirements* specified that the system comprise of WCF services backed by Coherence running on Windows boxes, so I guess technically my hack didn't qualify as a valid solution either.
It's a shame, an expensive one for that company, but that's what you get if all you have are hammers and every problem looks like a nail.