Honestly, if you're lucky, several of the lead developers have kept copies of the source of their own (legally or otherwise), and have a decent idea about what is and isn't worth reusing. I've seen this in a few places, both very big and small.
The brightest people I've seen (coders, managers, producers, etc) have "personal scrapbooks" of this sort.
If they're still at the same company, or otherwise have rights to the code, they use or encourage use of the code directly; if they're not, they generally use the practices and ideas to implement it again. New implementations on the same ideas can be even better than old ones.
If you work someplace with some examples of great code or ideas, it's worth prodding the authors until you understand the hows and whys. Often, if you get the reason behind _why_ something was done the way it was done, (re)building it becomes much easier. People generally enjoy explaining if you take an interest at an opportune time. This kind of information can be even more valuable -- and yet less legally/morally questionable -- than taking the code with you.
Better yet, if *you* have coded or helped with a great system, share it. If you can't share the code, share the ideas: lecture at an industry conference, talk to students, write a blog, submit magazine articles, publish a book, or whatever you can get away with; even company-internal workshops. You'll get far more information than you give, almost guaranteed. It's rewarding on several levels.
All that having been said, some of these things really do work out for the best; some of the code that was lost deserves to undergo a death/rebirth (it would stagnate and never improve otherwise), and some rare pieces should be utterly forgotten (too complex or useless to even present an effective antipattern -- the stuff of nightmares).
If you're willing to get into "reasonable future technology", the Minority Report user interface had a lot of thought put into it. John Underkoffler gave a presentation on it at the 2005 Game Developers Conference, discussing what thought went into the design of the interface. (Here's an article covering his lecture.)
Googling around, you can see that more than a few people have pointed out flaws with such an interface, but also that many places (including slashdot) have reported on multi-touch or gesture-based interfaces in the last few years. These start to scratch the surface of what was seen in the movie. I'd highly recommend viewing some of these movies; I was especially impressed with some of the work being with multi-touch displays.
I don't think this is the first transparent speaker ever to come to market. I say this because I spent a fair portion of my childhood in front of a pair of Stax speakers -- the 6' high variety, two panels apiece. One of our audiophile friends had a pair of the 3' high speakers with only one panel each.
You could indeed see through these; there was only a layer of something like a coarse cheesecloth in front of them, and a metal grill of sorts behind, protecting what I understood was a pair of sandwiched plastic layers that looked like celophane. Our cats would eye the whole arrangement from time to time and flex their claws. They learned to stay away from them eventually.
This particular pair was a factory-rejected "showroom model", coming at a significant discount: the engineers had the bright idea of putting two LEDs on the bottom of the speakers to indicate whether you were overdriving the speakers. Green was loud but acceptable, and Red meant that you'd probably already committed one or more of the precious panels to the garbage. Unfortunately, their reviewers, who like to listen in dark rooms at high volume, found the presense of a bright green LED "distracting". The company purportedly removed it from subsequent models.
The panels were apparently very low-yield -- something like only 11 pairs a year were manufactured, and it's seems pretty obvious from Stax's headphone-centric website that they're no longer making them. We did manage to get a replacement panel from them once, about 2 years after we purchased the speakers, but I'm pretty sure that won't happen again. The speakers had another problem recently with capacitors in their power supply leaking -- just recently we found a second batch when the second speaker's started going bad, but I don't believe they've been installed yet. A bit of a shame, really.
As a side note, these speakers completely spoiled me -- nothing which I've had since sounds nearly as clear. With a good recording, you could close your eyes and completely lose yourself. These taught me to appreciate vinyl far more than I would have otherwise; with some listening and comparison, I could even understand why my father had gone with a tube amp rather than a solid state one...
Slashdot Mirror
Honestly, if you're lucky, several of the lead developers have kept copies of the source of their own (legally or otherwise), and have a decent idea about what is and isn't worth reusing. I've seen this in a few places, both very big and small.
The brightest people I've seen (coders, managers, producers, etc) have "personal scrapbooks" of this sort.
If they're still at the same company, or otherwise have rights to the code, they use or encourage use of the code directly; if they're not, they generally use the practices and ideas to implement it again. New implementations on the same ideas can be even better than old ones.
If you work someplace with some examples of great code or ideas, it's worth prodding the authors until you understand the hows and whys. Often, if you get the reason behind _why_ something was done the way it was done, (re)building it becomes much easier. People generally enjoy explaining if you take an interest at an opportune time. This kind of information can be even more valuable -- and yet less legally/morally questionable -- than taking the code with you.
Better yet, if *you* have coded or helped with a great system, share it. If you can't share the code, share the ideas: lecture at an industry conference, talk to students, write a blog, submit magazine articles, publish a book, or whatever you can get away with; even company-internal workshops. You'll get far more information than you give, almost guaranteed. It's rewarding on several levels.
All that having been said, some of these things really do work out for the best; some of the code that was lost deserves to undergo a death/rebirth (it would stagnate and never improve otherwise), and some rare pieces should be utterly forgotten (too complex or useless to even present an effective antipattern -- the stuff of nightmares).
If you're willing to get into "reasonable future technology", the Minority Report user interface had a lot of thought put into it. John Underkoffler gave a presentation on it at the 2005 Game Developers Conference, discussing what thought went into the design of the interface. (Here's an article covering his lecture.)
Googling around, you can see that more than a few people have pointed out flaws with such an interface, but also that many places (including slashdot) have reported on multi-touch or gesture-based interfaces in the last few years. These start to scratch the surface of what was seen in the movie. I'd highly recommend viewing some of these movies; I was especially impressed with some of the work being with multi-touch displays.
Incidentally, their headphones were very nice too.
I don't think this is the first transparent speaker ever to come to market. I say this because I spent a fair portion of my childhood in front of a pair of Stax speakers -- the 6' high variety, two panels apiece. One of our audiophile friends had a pair of the 3' high speakers with only one panel each.
You could indeed see through these; there was only a layer of something like a coarse cheesecloth in front of them, and a metal grill of sorts behind, protecting what I understood was a pair of sandwiched plastic layers that looked like celophane. Our cats would eye the whole arrangement from time to time and flex their claws. They learned to stay away from them eventually.
This particular pair was a factory-rejected "showroom model", coming at a significant discount: the engineers had the bright idea of putting two LEDs on the bottom of the speakers to indicate whether you were overdriving the speakers. Green was loud but acceptable, and Red meant that you'd probably already committed one or more of the precious panels to the garbage. Unfortunately, their reviewers, who like to listen in dark rooms at high volume, found the presense of a bright green LED "distracting". The company purportedly removed it from subsequent models.
The panels were apparently very low-yield -- something like only 11 pairs a year were manufactured, and it's seems pretty obvious from Stax's headphone-centric website that they're no longer making them. We did manage to get a replacement panel from them once, about 2 years after we purchased the speakers, but I'm pretty sure that won't happen again. The speakers had another problem recently with capacitors in their power supply leaking -- just recently we found a second batch when the second speaker's started going bad, but I don't believe they've been installed yet. A bit of a shame, really.
As a side note, these speakers completely spoiled me -- nothing which I've had since sounds nearly as clear. With a good recording, you could close your eyes and completely lose yourself. These taught me to appreciate vinyl far more than I would have otherwise; with some listening and comparison, I could even understand why my father had gone with a tube amp rather than a solid state one...