The extra diagnostic LEDs are certainly very easy to design in these days and don't even cost any extra plastic. There is a part available that integrates one or two LEDs directly into the plastic RJ45 housing. It's got to cost a bit more, but certainly gives you those LEDs at no extra design and/or space cost.
One could argue that this is a security measure. Remember all this talk about how you can reconstruct the bit stream from a LED's blinking remotely?:)
What good will it do without the application support? I expect this to be similar to Windows, which also still has useful mappings for 2 buttons only, not the middle one.
I suppose I could map the extra buttons to common keyboard modifiers via software myself, but wouldn't that be an un-Applish procedure? I mean, an Apple user is assumed to be a total idiot:)
I disagree. Internet-wide email took off the way it did exactly because it was so interoperable with all the existing email systems and it became the glue between them. It will be difficult to make FWD as useful as it is in US because the local calls are not free to end users elsewhere. There is some hope though due to all the IP-enabled cellphones...
What is badly needed is a way to integrate disparate directory services into one. This will probably end up being done on the client rather through interoperation at gatekeeper level (don't you wish...). For example, I have an ATA 186 configured for FWD. So, I can now call my friends with a similar setup for free, but I can no longer call my grandma in Russia through a commercial VoIP provider's GK...
That clunky single-button of theirs is probably the reason #1 I don't want an Apple machine. (most of their other hardware and the OS are quite all right meanwhile, IMHO) I so don't buy their argument that more than one button is confusing. Most Apple programms end up using keyboard modifiers to achieve the same thing they could've done with a couple of extra buttons and that's a lot more confusing to me. I think a mouse needs as many buttons as users can tell apart by touch without looking, which is probably about 3.
Apple just don't want to admit Microsoft beat them there on a UI-related issue, so they come up with idiotic pseudo-inventions like this new mouse just so that they could sneak in extra buttons. I figure, M$ also added a wheel mostly because they finally figured out 3 buttons were better than 2 after all, but didn't want it to look like they followed Logitech's lead. In the interest of full disclosure, my perfect mouse would be an IBM ScrollPoint, but it would have those Honeywell double legs for sensors. (allows for a fully sealed case and works on absolutely any surface) Extemely unlikely anyone will ever build such a combo, as both the ScrollPoint stick and Honeywell sensor are covered by expensive patents.:(
These IBM mice are called ScrollPoint and they are still being made. I am using a recent optical version right now. These are my favorite mice and I had to order mine from US (those are only available here in Europe with a complete system from IBM)
What force to counteract? When I am using a mouse, the back of my hand still touches the tabletop, so pulling this hand away by moving the same is equivalent to pulling yourself up by the ears.
I find those mice a much better idea than the Microsoft-style wheeled mice, because:
1. You can also scroll horizontally.
2. If you want to keep scrolling, just push the stick into desired direction and hold. With a wheel, you have to keep scrolling. The harder you push the stick, the faster you scroll. Wheeled mice are only good for contant-speed scrolling.
3. I still have a full-width middle button. Your typical mouse wheel is not really a very convinient button for prolonged use (say, paste in X)
The mice are very well made and look cool, especially that blue LED backlight inside the little stick. (which is quite a bit larger and more comfortable too than the little clit they put onto laptop's keyboards)
Overall, I'd rate the ScrollPoint as one of the most significant advancements in HCI, to which effect it also won some awards. As usual, IBM can neither hype it properly themselves, nor make the related patents properly accessible to other manufacturers.
Dasher is all right, I suppose. One way or another, there is going to be a need for a one-button interfa ce. I think we agree on that. There was also some German company that caused a bit of a stir at the last Cebit with their PDA/cellphone platform designed for one-handed input. (Linux based, too) OK, found it: http://www.invair.com/
As far as a voice interface as good as Star Trek one is concerned, I don't buy it. Human level understanding of human speech may very well require human-level AI, and we are no closer to that today than we were 30 years ago. I don't think Moore's law is going to help you any.
P.S. Don't go to Austria, Andy. This place is a dump and nobody here knows what Sound of Music is anyway:)
They were supposed to have even higher energy density than Li-Ion and family, but haven't seen them recently outside of niche hearing aid market. Is there a problem with them breathing atmoshperic air or something?
I think the point is moot, because there aren't really any battery factories left in the US. The big name US/Western European/Japanese brands simply rebadge batteries coming from the same OEMs. For something as commoditized as standard-sized rechargaable battery, badge engineering makes the most sense. I'd say most of the actual battery production goes to Taiwan, continental China and South Korea.
I kid you not. Consider this. The only reasons PDAs/cellphones/portables stopped getting smaller is that they still need to accommodate some sort of a display and input device(s). But, it will soon be perfectly feasible to have a very high-res display integrated into some sort of eyewear like regular looking sunglasses (actually, the only way to make a very small yet high-res screen readable is to put it right in front of your eye) and the rest of the device could be just a single button dangling off your keychain. Morse input is a natural for that. Palm graphitti (sp?) clearly demonstrated that a large number people can be moved to learn a new input mechanism if there is a clear benefit to it. In this case the benefit is being able to input without looking, eq while driving. You'd be surpised how much else you could do with a single button. (see www.xenote.com, now defunct:( , for an example)
The best natural pesticide is cocaine. Insects O.D. on it much easier than warm-blooded animals, cause coke is much more effective reuptake inhibitor for octamine (sp?) a neurotransmitter specific to insects, than dopamine. (the way over which it affects humans) THC is no slouch either. Most bugs can't eat weed:)
I think Mazda should be on the list for their Wankel rotaries. Just two moving parts in the whole engine! I hope this Merc lunacy with by-wire controls is going to be a fad that goes away. The computer industry is actually way ahead of car industry in understanding that every complex system needs a manual failover. Why isn't there a single car with an additional manual control for electric windows? As the cars get old, gizmos are the first to go, so you have to be able to still use the functions manually. I know somebody, who had the battery die on him in an armored 7-series beemer. He couldn't open the door or even knock out a window. Good thing he had his cellphone, so the towing company came and got him out of the can.
Would you consider a language like K if your code got 100 times smaller? A 1000 times? I mean, at some point it's got to be worth it. K code is so extremely dense so that you could grok the whole thing at once. (no K program is ever supposed to go over a single page:) I've personally replaced 2.5K lines of Java (132 character wide, mind you) with just 4 lines of K.
You really want to be able to "splice" your database into separate files by row, column or whole table, so that you can put those files into different partitions in order to minimize disk arm contention.
TPC results are posted by database vendors themselves. It is a really expensive procedure, because the compliance to the official rulebook has to be meticulously documented. This is why you don't see many posts from smaller vendors, even though their product is often faster. (The fastest db I am aware of is kdb by kx systems (www.kx.com), by far)
In Europe, a fancier version called Dual Play is available from some British online stores. It goes for 299 pounds, but has more features. (like 256MB Flash and realtime MP3 encoding, to make it a recorder)
'non-addictive' cannabis alternative. Merck was a leading cocain manufacturer. Our current choice of 'good' and 'bad' chemicals is in fact completely arbitrary and illogical. The list demonized substances in different cultures and different times have been quite different from ours and just as arbitrary. One would expect at least/. readers, who consider themselves digirati to have some clue. Coke speficially powered some good work. To name a few avid users - Thomas Edison, Freud, Conan Doyle.
The 'neurological damage' caused by E is non-conclusive at best. I have been an occasional user for several years and I don't notice any ill effects myself. The research about damage caused to seratonin receptors is cooked, IMHO. If the receptors were damaged, don't you think that E wouldn't work anymor? There is no evidence meth causes any brain damage directly. The health risks of meth are indirect and are related to malnutrition and sleep deprivation. What can I say, don't be an idiot and remember to eat and sleep when you are on meth. BTW, I love meth I do have a job. If anything, it increases your capability for work manyfold. Think of it as industrial-strength caffeine.
The C++ kitchen sink approach to OO just plain sucks, but I find the newfangled generative stuff intriguing. Back in school we tended to settle on a more manageable subset of C++ functionality. You really don't have to use every obscure feature and than it becomes better. You can also write perfectly fine object code in plain C, using structs and function pointers. In fact, the most beautiful OO design I've seen in a real program is that of the Freetype library, written in C. For embedded folks, C++ is usually a non-option. On many small platforms, you'd be happy if your C compiler is working as advertised.
I don't follow the problem you have with compiler flags. Even if you compile parts of your code with different flags, you can surely still link them together?
You are indeed correct in that every two player zero-sum game has a 'solution' and a 'perfect strategy'. Theoretically. Solving chess as a matrix game is, however, not possible in practice due to its extreme size. If you had infinite processing power and memory, it'd be trivial though:)
From a database/storage standpoint, it makes a lot more sense to stream an object collection spliced vertically (column-wise?) rather than object-/row-wise.
I have a whole list of things that Oracle should have fixed many many years ago. My guess that the intern that last touched the code is long gone:)
1. No readline support in shell SQL monitor. There is no good way to browse previous commands with one click.
2. Varchar is limited to some small and completely arbitrary number (I believe it was exactly 40000 bytes) I don't see the point in varchar anymore, anyway. Compare that to Informix/Postgres type 'text' - arbitrary large text of non-specific length. (40000 may be too small for some webpages for example. I know they sell a plug-in to store large text, but that means bye bye standard SQL)
3. I am sorry, but 56-digit long strings are a screwed-up way to represent numbers.
4. Non-support for ANSI SQL timestamp type. Oracle's best timestamp definition is 1 second, which for example renders it near useless for trading applications (timeseries data is often keyed on the timestamp) This is probably why so many financial firms use Sybase.
Positive things about Oracle:
Oracle writers do not lock readers, but you pay with 'handle expired' problem for that.
I find it really relatively quite stable, amazingly so for something this large and convoluted. Must be all that burn-in time from the many users.
Everything else you can get elsewhere, only with much less admin overhead. Oracle's got to be some sort of secret government public works project for Oracle admins.
Slashdot Mirror
Have a pity on a poor foreigner with a limited vocabulary.
The extra diagnostic LEDs are certainly very easy to design in these days and don't even cost any extra plastic. There is a part available that integrates one or two LEDs directly into the plastic RJ45 housing. It's got to cost a bit more, but certainly gives you those LEDs at no extra design and/or space cost.
:)
One could argue that this is a security measure. Remember all this talk about how you can reconstruct the bit stream from a LED's blinking remotely?
What good will it do without the application support? I expect this to be similar to Windows, which also still has useful mappings for 2 buttons only, not the middle one.
:)
I suppose I could map the extra buttons to common keyboard modifiers via software myself, but wouldn't that be an un-Applish procedure? I mean, an Apple user is assumed to be a total idiot
I disagree. Internet-wide email took off the way it did exactly because it was so interoperable with all the existing email systems and it became the glue between them. It will be difficult to make FWD as useful as it is in US because the local calls are not free to end users elsewhere. There is some hope though due to all the IP-enabled cellphones...
...). For example, I have an ATA 186 configured for FWD. So, I can now call my friends with a similar setup for free, but I can no longer call my grandma in Russia through a commercial VoIP provider's GK...
What is badly needed is a way to integrate disparate directory services into one. This will probably end up being done on the client rather through interoperation at gatekeeper level (don't you wish
That clunky single-button of theirs is probably the reason #1 I don't want an Apple machine. (most of their other hardware and the OS are quite all right meanwhile, IMHO) I so don't buy their argument that more than one button is confusing. Most Apple programms end up using keyboard modifiers to achieve the same thing they could've done with a couple of extra buttons and that's a lot more confusing to me. I think a mouse needs as many buttons as users can tell apart by touch without looking, which is probably about 3.
:(
Apple just don't want to admit Microsoft beat them there on a UI-related issue, so they come up with idiotic pseudo-inventions like this new mouse just so that they could sneak in extra buttons. I figure, M$ also added a wheel mostly because they finally figured out 3 buttons were better than 2 after all, but didn't want it to look like they followed Logitech's lead. In the interest of full disclosure, my perfect mouse would be an IBM ScrollPoint, but it would have those Honeywell double legs for sensors. (allows for a fully sealed case and works on absolutely any surface) Extemely unlikely anyone will ever build such a combo, as both the ScrollPoint stick and Honeywell sensor are covered by expensive patents.
These IBM mice are called ScrollPoint and they are still being made. I am using a recent optical version right now. These are my favorite mice and I had to order mine from US (those are only available here in Europe with a complete system from IBM)
What force to counteract? When I am using a mouse, the back of my hand still touches the tabletop, so pulling this hand away by moving the same is equivalent to pulling yourself up by the ears.
I find those mice a much better idea than the Microsoft-style wheeled mice, because:
1. You can also scroll horizontally.
2. If you want to keep scrolling, just push the stick into desired direction and hold. With a wheel, you have to keep scrolling. The harder you push the stick, the faster you scroll. Wheeled mice are only good for contant-speed scrolling.
3. I still have a full-width middle button. Your typical mouse wheel is not really a very convinient button for prolonged use (say, paste in X)
The mice are very well made and look cool, especially that blue LED backlight inside the little stick. (which is quite a bit larger and more comfortable too than the little clit they put onto laptop's keyboards)
Overall, I'd rate the ScrollPoint as one of the most significant advancements in HCI, to which effect it also won some awards. As usual, IBM can neither hype it properly themselves, nor make the related patents properly accessible to other manufacturers.
Dasher is all right, I suppose. One way or another, there is going to be a need for a one-button interfa ce. I think we agree on that. There was also some German company that caused a bit of a stir at the last Cebit with their PDA/cellphone platform designed for one-handed input. (Linux based, too)
:)
OK, found it: http://www.invair.com/
As far as a voice interface as good as Star Trek one is concerned, I don't buy it. Human level understanding of human speech may very well require human-level AI, and we are no closer to that today than we were 30 years ago. I don't think Moore's law is going to help you any.
P.S. Don't go to Austria, Andy. This place is a dump and nobody here knows what Sound of Music is anyway
And?
Whatever happened to Zinc-Air rechargeables?
They were supposed to have even higher energy density than Li-Ion and family, but haven't seen them recently outside of niche hearing aid market. Is there a problem with them breathing atmoshperic air or something?
There is a well sorted listing of battery manufacturers, by type of product, by country, etc.
a tP /batt/batt.shtml
Knock yourself out:
http://energy.sourceguides.com/businesses/byP/b
I think the point is moot, because there aren't really any battery factories left in the US. The big name US/Western European/Japanese brands simply rebadge batteries coming from the same OEMs. For something as commoditized as standard-sized rechargaable battery, badge engineering makes the most sense. I'd say most of the actual battery production goes to Taiwan, continental China and South Korea.
I kid you not. Consider this. The only reasons PDAs/cellphones/portables stopped getting smaller is that they still need to accommodate some sort of a display and input device(s). But, it will soon be perfectly feasible to have a very high-res display integrated into some sort of eyewear like regular looking sunglasses (actually, the only way to make a very small yet high-res screen readable is to put it right in front of your eye) and the rest of the device could be just a single button dangling off your keychain. Morse input is a natural for that. Palm graphitti (sp?) clearly demonstrated that a large number people can be moved to learn a new input mechanism if there is a clear benefit to it. In this case the benefit is being able to input without looking, eq while driving. You'd be surpised how much else you could do with a single button. (see www.xenote.com, now defunct :( , for an example)
The best natural pesticide is cocaine. Insects O.D. on it much easier than warm-blooded animals, cause coke is much more effective reuptake inhibitor for octamine (sp?) a neurotransmitter specific to insects, than dopamine. (the way over which it affects humans) THC is no slouch either. Most bugs can't eat weed :)
I think Mazda should be on the list for their Wankel rotaries. Just two moving parts in the whole engine! I hope this Merc lunacy with by-wire controls is going to be a fad that goes away. The computer industry is actually way ahead of car industry in understanding that every complex system needs a manual failover. Why isn't there a single car with an additional manual control for electric windows? As the cars get old, gizmos are the first to go, so you have to be able to still use the functions manually. I know somebody, who had the battery die on him in an armored 7-series beemer. He couldn't open the door or even knock out a window. Good thing he had his cellphone, so the towing company came and got him out of the can.
Would you consider a language like K if your code got 100 times smaller? A 1000 times? I mean, at some point it's got to be worth it. K code is so extremely dense so that you could grok the whole thing at once. (no K program is ever supposed to go over a single page :) I've personally replaced 2.5K lines of Java (132 character wide, mind you) with just 4 lines of K.
You really want to be able to "splice" your database into separate files by row, column or whole table, so that you can put those files into different partitions in order to minimize disk arm contention.
TPC results are posted by database vendors themselves. It is a really expensive procedure, because the compliance to the official rulebook has
to be meticulously documented. This is why you don't
see many posts from smaller vendors, even though their product is often faster. (The fastest db I am
aware of is kdb by kx systems (www.kx.com), by far)
In Europe, a fancier version called Dual Play is
available from some British online stores. It goes for 299 pounds, but has more features. (like 256MB Flash and realtime MP3 encoding, to make it a recorder)
'non-addictive' cannabis alternative. Merck was a leading cocain manufacturer. Our current choice of 'good' and 'bad' chemicals is in fact completely arbitrary and illogical. The list demonized substances in different cultures and different times have been quite different from ours and just as arbitrary. One would expect at least /. readers, who consider themselves digirati to have some clue. Coke speficially powered some good work. To name a few avid users -
Thomas Edison, Freud, Conan Doyle.
The 'neurological damage' caused by E is non-conclusive at best. I have been an occasional user for several years and I don't notice any ill effects myself. The research about damage caused to seratonin receptors is cooked, IMHO. If the receptors were damaged, don't you think that E wouldn't work anymor? There is no evidence meth causes any brain damage directly. The health risks of meth are indirect and are related to malnutrition and sleep deprivation. What can I say, don't be an idiot and remember to eat and sleep when you are on meth. BTW, I love meth I do have a job. If anything, it increases your capability for work manyfold. Think of it as industrial-strength caffeine.
The C++ kitchen sink approach to OO just plain sucks, but I find the newfangled generative stuff intriguing. Back in school we tended to settle on a more manageable subset of C++ functionality. You really don't have to use every obscure feature and than it becomes better. You can also write perfectly fine object code in plain C, using structs and function pointers. In fact, the most beautiful OO design I've seen in a real program is that of the Freetype library, written in C. For embedded folks, C++ is usually a non-option. On many small platforms, you'd be happy if your C compiler is working as advertised.
I don't follow the problem you have with compiler flags. Even if you compile parts of your code with different flags, you can surely still link them together?
You are indeed correct in that every two player zero-sum game has a 'solution' and a 'perfect strategy'. Theoretically. Solving chess as a matrix game is, however, not possible in practice due to its extreme size. If you had infinite processing power and memory, it'd be trivial though :)
From a database/storage standpoint, it makes a
lot more sense to stream an object collection spliced vertically (column-wise?) rather than object-/row-wise.
I have a whole list of things that Oracle should have fixed many many years ago. My guess that the intern that last touched the code is long gone :)
1. No readline support in shell SQL monitor. There is no good way to browse previous commands with one click.
2. Varchar is limited to some small and completely arbitrary number (I believe it was exactly 40000 bytes) I don't see the point in varchar anymore, anyway. Compare that to Informix/Postgres type 'text' - arbitrary large text of non-specific length. (40000 may be too small for some webpages for example. I know they sell a plug-in to store large text, but that means bye bye standard SQL)
3. I am sorry, but 56-digit long strings are a screwed-up way to represent numbers.
4. Non-support for ANSI SQL timestamp type. Oracle's best timestamp definition is 1 second, which for example renders it near useless for trading applications (timeseries data is often keyed on the timestamp) This is probably why so many financial firms use Sybase.
Positive things about Oracle:
Oracle writers do not lock readers, but you pay with 'handle expired' problem for that.
I find it really relatively quite stable, amazingly so for something this large and convoluted. Must be all that burn-in time from the many users.
Everything else you can get elsewhere, only with much less admin overhead. Oracle's got to be some sort of secret government public works project for Oracle admins.
1. Transaction support is a joke. Correct me if I am wrong, but doesn't compiling in Berkeley DB means each writer locks a whole table?
2. No subselects. My class project db did subselects, for Pete's sake!
3. Query planning/optimizing close to non-existent.