That egg was hatched a long time ago. Atom is going nowhere in smart phones, and cellphone handsets are outselling PCs by a huge factor. Nobody cares that their phone can't run Windows. x86 compatibility has absolutely zero value in this market, and this is the fastest growing market, bar none.
By the way, in case you haven't noticed, apps haven't been written exclusively in x86 assembly language in a long long time. Source code written in C can be compiled for ARM just as easily as it can be for any other CPU family. Programs written in Java, Python, whatever, don't even need to be recompiled. x86 ISA compatibility's importance as a selection criteria in hardware is dead and gone in the only markets that are growing.
But we can't stay at that 100x level, and in reality we don't need to be there all the time. Intel Atom proves that - you can get *enough* useful work done with a simpler design, and fewer transistors. Unfortunately, when you get down to the number of transistors that Atom uses, suddenly the frontend decoder *is* a significant proportion of your die real estate again. Inefficiency *always* costs you, and it's stupid to pretend that it doesn't. Atom may try to challenge ARM but it will fail, as long as it keeps the x86 ISA baggage. Efficiency *matters*.
I wonder if this stuff will dissolve dental biofilms. Would be cool to finally have a simple, 100% effective treatment that totally prevents plaque, gum disease, cavities... Tho I suppose if it's that good, the ADA will bury it.
I remember when someone posted their implementation of John Conway's "Life" to usenet. That was a fun way to tie up a printer and waste a ream of paper. (Basically you could prepend it to any document and it would iterate off the document's first page. Each generation printed on a separate page.)
I think they're just acknowledging that they can't write firmware to save their lives. I had a WG602 that would always lock up after a few days of use; the lockups would happen sooner after big ftp/scp sessions. Basically the damn thing had a memory leak. Updating to the latest firmware didn't help; I finally replaced it with a Linksys.
(Oh yeah, and they also promised upgradability to 802.1x WPA when I bought it, and never released a firmware update with WPA support.) AFA I'm concerned, this is the smartest decision they could possibly make. Now they don't have to bother with fake promises of future firmware upgrades, they can just leave it to their customers to upgrade at will. And people buying these routers won't have to put up with buggy firmware without any recourse.
Of course I still think it's too late; I've completely sworn off ever buying Netgear again and have stuck to Linksys...
Yes, that's always been one of my problems with the GPL. Way back before it existed, I used to release all my freeware projects with a license clause "You are free to use this code as you see fit, but any bug fixes you make must be sent back to me for incorporation into the master source." When the GPL came around I got a lot of pressure to relicense my code, because my license wasn't "free" enough, or didn't fit some johnny-come-lately organization's definition of "free software license."
Freedom doesn't mean lack of responsibility. If you're getting a benefit from the code I write, and you find a flaw in it, I think you owe me a bug report at least.
It also implies they'll be completely out of the software business by then. It only makes sense if they're planning to totally reinvent themselves along the way. Personally, if I were at Sun and thought SaaS was going to be the model of the future, I'd be making moves to ensure that other companies would be getting their services from me, not dismantling anything I owned that could possibly be used to offer such a service.
Still, the whole model is predicated on networking technology becoming so efficient that there's no significant cost to running your apps and accessing your data at an arbitrarily distant data center. To believe this will ever be true is to deny reality. E.g., disk usage always expands to fill available disk capacity. Network usage always expands to fill available bandwidth. Service levels in this brave new SaaS world will always be prone to outages and traffic overloads, and they will invariably cause failures at the least convenient possible moment.
The summary for OpenLDAP on the Coverity web page is out of date; there are no uninspected or unresolved bugs in the code base as of their last actual run. There have been no uninspected or unresolved bugs for several months.
The most obvious thing: USB was invented by Intel and designed to keep Intel microprocessors as an indispensible part of the system. That's why it uses that stupid "speak when spoken to" protocol. Intel did everything they could to proliferate USB in preference to Firewire, because they wanted to make sure the x86 remained king of the hill in system architecture. That's why USB is so cheap today. Nobody with equal marketing muscle was really pushing Firewire the same way, even though Firewire is better for 3rd party hardware manufacturers.
It may well be just business, but still it's Bad business.
If you develop a cutting edge system that has value to your employer, then they should make it worth your while to stay there. A non-compete is the same as saying "we don't give a damn about you, we just want to pay as little as possible for your output." It's the mark of an extremely unenlightened management style, the kind that believes that any employee is interchangeable with any other. Despite the obvious cognitive dissonance, if you can be so damaging at a competitor, then you're obviously not just an interchangeable cog in their big machine.
When you develop something of value to your company, you deserve both compensation and recognition. We are not interchangeable droids, we are individuals, and deserve to be recognized as such. Any company where the management doesn't understand this doesn't deserve to be in business, and they certainly don't deserve to profit from my labor.
Been there, done that. Their idiotic agreement would have claimed ownership of *ANY* intellectual property I created, even on my own time and my own gear. And it was so general it would have applied to all the music I've composed for my band, as well as software I wrote in my hobby activities. (Both totally ludicrous; I was still developing on Atari STs back then. Stuff that would have absolutely zero relevance to my day-job enterprise software coding. But it's all "IP" and that contract would have laid claim to it.) I refused to sign, left, and started my own company. Today we market commercial support for the fastest LDAP server in the world, and my code totally obliterates the performance of anything that other company makes. And of course, all my code now is open source. No more proprietary bullshit.
Well, monitors are a perfectly good place to start. I have a couple monitors always in standby mode in my house, and several at the office. But sure, there's plenty of other devices that need to change, like all of my stereo gear with IR remote receivers.
Personally I think the idea of using a photovoltaic cell with a capacitor is pretty good. It doesn't take much energy to operate a photocell. (Heck, they are after all the same technology as photovoltaic cells, just tuned for a different optimal wavelength and lower output.) Get rid of all the Standby LEDs, that'd help.
The obvious trick with the wireless remotes is to use the photocell output to turn on the output from the cap, to power the FM detector and decide if the incoming energy is a remote command or not. If not, ignore it. In the meantime, continue to absorb all that light and keep charging the cap.
Close. Only Federally registered marks are allowed to use the (R) mark. You can register at the state level but that's essentially meaningless; there are no laws giving you any particular protection for those.
I happen to know all of this because I went thru the legwork of registering my band name "Highland Sun" with the USPTO. A horse ranch somewhere in Kentucky (IIRC) wrote me a letter claiming I was infringing on their name "Highland Sun Farms", registered in their state and I wrote back essentially (1) we're not in similar businesses, no one will ever get us confused and (2) mine is Federal, yours is state; if you want to push this, you'll lose. They went away.
I remember sending the Life game to an Apple LaserWriter back in college. It was almost as fun as watching it run on a DECwriter back in high school, but used a lot more paper... I don't remember getting to send it to a NEWS or DisplayPostscript server, that might have been amusing.
He didn't specify what kind of cellphone access he has to choose from now. CDMA is pretty good with Verizon, 1xRTT at 144kbps worked pretty well for me and now there's EVDO. I used 1xRTT all the time when traveling around the US, and it's good enough for ssh...
It seems odd that the number would be completely meaningless. You take a huge complex chain, you decide that the sequence of base pairs in the chain is significant, but you decide that breaking this chain into two pieces is not significant? If it's so insignificant then why is it (usually) replicated so consistently? If a chromosome just randomly breaks somewhere, will a centromere just automagically form in the center of the two pieces so that they can replicate themselves the next time the cell decides to divide?
We know that trivial little things like deleting the wrong portion of a chromosome can have pretty big effects (e.g. mental retardation, other physical changes too) so it seems that there's more than random breakage to account for.
re: bold moves - people already think they're entitled to create artificial life, that seems pretty bold. I think it's smarter to understand more about the life that already exists, before going off and creating something totally new...
I think it would be interesting to take a person's stem cell and try to remove all the "junk DNA" from the nucleus, then grow the cell thru a few generations (perhaps even to a full clone) and see how different it is from the original person. Very likely a lot of what we think is junk DNA isn't useless after all. Probably the reason we have 46 chromosomes in the first place is that we've been accumulating genetic material from other microbes over the span of millions of years...
And what exactly is the lowest denominator that you see here? Heck even back in 1993 we could do 8 channel double-buffered digital audio glitch-free on an Atari Falcon. Why should the fact that a network interrupt handler needs a few extra CPU cycles be an excuse for audio glitches to occur?
Slashdot Mirror
That egg was hatched a long time ago. Atom is going nowhere in smart phones, and cellphone handsets are outselling PCs by a huge factor. Nobody cares that their phone can't run Windows. x86 compatibility has absolutely zero value in this market, and this is the fastest growing market, bar none.
By the way, in case you haven't noticed, apps haven't been written exclusively in x86 assembly language in a long long time. Source code written in C can be compiled for ARM just as easily as it can be for any other CPU family. Programs written in Java, Python, whatever, don't even need to be recompiled. x86 ISA compatibility's importance as a selection criteria in hardware is dead and gone in the only markets that are growing.
But we can't stay at that 100x level, and in reality we don't need to be there all the time. Intel Atom proves that - you can get *enough* useful work done with a simpler design, and fewer transistors. Unfortunately, when you get down to the number of transistors that Atom uses, suddenly the frontend decoder *is* a significant proportion of your die real estate again. Inefficiency *always* costs you, and it's stupid to pretend that it doesn't. Atom may try to challenge ARM but it will fail, as long as it keeps the x86 ISA baggage. Efficiency *matters*.
I wonder if this stuff will dissolve dental biofilms. Would be cool to finally have a simple, 100% effective treatment that totally prevents plaque, gum disease, cavities... Tho I suppose if it's that good, the ADA will bury it.
No, it's the plain truth. The AD database design is inherently flawed, and AD is the least scalable directory server in the world.
http://www.mail-archive.com/ldap@umich.edu/msg01464.html
Bricking a laserwriter wasn't hard at all.
I remember when someone posted their implementation of John Conway's "Life" to usenet. That was a fun way to tie up a printer and waste a ream of paper. (Basically you could prepend it to any document and it would iterate off the document's first page. Each generation printed on a separate page.)
Some folks prefer a little more control. Personally, I run my own caching DNS on my laptop, as well as on my main server at home.
These stanzas in my /etc/named.conf are the only thing that make web browsing bearable for me:
zone "intellitxt.com" in {
type master;
file "junk.zone";
};
zone "doubleclick.net" in { ...
type master;
file "junk.zone";
};
viola:/var/lib/named> cat junk.zone
$TTL 1W
@ IN SOA @ root (
42 ; serial (d. adams)
2D ; refresh
4H ; retry
6W ; expiry
1W ) ; minimum
IN NS @
IN A 0.0.0.0
I think they're just acknowledging that they can't write firmware to save their lives. I had a WG602 that would always lock up after a few days of use; the lockups would happen sooner after big ftp/scp sessions. Basically the damn thing had a memory leak. Updating to the latest firmware didn't help; I finally replaced it with a Linksys.
(Oh yeah, and they also promised upgradability to 802.1x WPA when I bought it, and never released a firmware update with WPA support.) AFA I'm concerned, this is the smartest decision they could possibly make. Now they don't have to bother with fake promises of future firmware upgrades, they can just leave it to their customers to upgrade at will. And people buying these routers won't have to put up with buggy firmware without any recourse.
Of course I still think it's too late; I've completely sworn off ever buying Netgear again and have stuck to Linksys...
Ah, so they've finally developed the technology for the Star Trek handheld medical scanner. Cool.
Funny, I was benchmarking an SGI Altix with 1TB of RAM back in 2005. What's this "not available yet" nonsense?
Most likely they should have written their own archive parsing libraries, rather than relying on the existing binaries for those formats.
As a footnote, there are no such buffer overrun vulnerabilities in my ARC program, which is now more than 22 years old.
http://sourceforge.net/projects/arc
They should just weave this stuff in with silk or wool and channel off the electricity from static buildup. Simple.
Yes, that's always been one of my problems with the GPL. Way back before it existed, I used to release all my freeware projects with a license clause "You are free to use this code as you see fit, but any bug fixes you make must be sent back to me for incorporation into the master source." When the GPL came around I got a lot of pressure to relicense my code, because my license wasn't "free" enough, or didn't fit some johnny-come-lately organization's definition of "free software license."
Freedom doesn't mean lack of responsibility. If you're getting a benefit from the code I write, and you find a flaw in it, I think you owe me a bug report at least.
It also implies they'll be completely out of the software business by then. It only makes sense if they're planning to totally reinvent themselves along the way. Personally, if I were at Sun and thought SaaS was going to be the model of the future, I'd be making moves to ensure that other companies would be getting their services from me, not dismantling anything I owned that could possibly be used to offer such a service.
Still, the whole model is predicated on networking technology becoming so efficient that there's no significant cost to running your apps and accessing your data at an arbitrarily distant data center. To believe this will ever be true is to deny reality. E.g., disk usage always expands to fill available disk capacity. Network usage always expands to fill available bandwidth. Service levels in this brave new SaaS world will always be prone to outages and traffic overloads, and they will invariably cause failures at the least convenient possible moment.
The most obvious thing: USB was invented by Intel and designed to keep Intel microprocessors as an indispensible part of the system. That's why it uses that stupid "speak when spoken to" protocol. Intel did everything they could to proliferate USB in preference to Firewire, because they wanted to make sure the x86 remained king of the hill in system architecture. That's why USB is so cheap today. Nobody with equal marketing muscle was really pushing Firewire the same way, even though Firewire is better for 3rd party hardware manufacturers.
I'm surprised you need to ask. http://www.openldap.org/
It may well be just business, but still it's Bad business.
If you develop a cutting edge system that has value to your employer, then they should make it worth your while to stay there. A non-compete is the same as saying "we don't give a damn about you, we just want to pay as little as possible for your output." It's the mark of an extremely unenlightened management style, the kind that believes that any employee is interchangeable with any other. Despite the obvious cognitive dissonance, if you can be so damaging at a competitor, then you're obviously not just an interchangeable cog in their big machine.
When you develop something of value to your company, you deserve both compensation and recognition. We are not interchangeable droids, we are individuals, and deserve to be recognized as such. Any company where the management doesn't understand this doesn't deserve to be in business, and they certainly don't deserve to profit from my labor.
Been there, done that. Their idiotic agreement would have claimed ownership of *ANY* intellectual property I created, even on my own time and my own gear. And it was so general it would have applied to all the music I've composed for my band, as well as software I wrote in my hobby activities. (Both totally ludicrous; I was still developing on Atari STs back then. Stuff that would have absolutely zero relevance to my day-job enterprise software coding. But it's all "IP" and that contract would have laid claim to it.) I refused to sign, left, and started my own company. Today we market commercial support for the fastest LDAP server in the world, and my code totally obliterates the performance of anything that other company makes. And of course, all my code now is open source. No more proprietary bullshit.
Some things must never be compromised.
Well, monitors are a perfectly good place to start. I have a couple monitors always in standby mode in my house, and several at the office. But sure, there's plenty of other devices that need to change, like all of my stereo gear with IR remote receivers.
Personally I think the idea of using a photovoltaic cell with a capacitor is pretty good. It doesn't take much energy to operate a photocell. (Heck, they are after all the same technology as photovoltaic cells, just tuned for a different optimal wavelength and lower output.) Get rid of all the Standby LEDs, that'd help.
The obvious trick with the wireless remotes is to use the photocell output to turn on the output from the cap, to power the FM detector and decide if the incoming energy is a remote command or not. If not, ignore it. In the meantime, continue to absorb all that light and keep charging the cap.
Close. Only Federally registered marks are allowed to use the (R) mark. You can register at the state level but that's essentially meaningless; there are no laws giving you any particular protection for those.
I happen to know all of this because I went thru the legwork of registering my band name "Highland Sun" with the USPTO. A horse ranch somewhere in Kentucky (IIRC) wrote me a letter claiming I was infringing on their name "Highland Sun Farms", registered in their state and I wrote back essentially (1) we're not in similar businesses, no one will ever get us confused and (2) mine is Federal, yours is state; if you want to push this, you'll lose. They went away.
I remember sending the Life game to an Apple LaserWriter back in college. It was almost as fun as watching it run on a DECwriter back in high school, but used a lot more paper... I don't remember getting to send it to a NEWS or DisplayPostscript server, that might have been amusing.
He didn't specify what kind of cellphone access he has to choose from now. CDMA is pretty good with Verizon, 1xRTT at 144kbps worked pretty well for me and now there's EVDO. I used 1xRTT all the time when traveling around the US, and it's good enough for ssh...
It seems odd that the number would be completely meaningless. You take a huge complex chain, you decide that the sequence of base pairs in the chain is significant, but you decide that breaking this chain into two pieces is not significant? If it's so insignificant then why is it (usually) replicated so consistently? If a chromosome just randomly breaks somewhere, will a centromere just automagically form in the center of the two pieces so that they can replicate themselves the next time the cell decides to divide?
We know that trivial little things like deleting the wrong portion of a chromosome can have pretty big effects (e.g. mental retardation, other physical changes too) so it seems that there's more than random breakage to account for.
re: bold moves - people already think they're entitled to create artificial life, that seems pretty bold. I think it's smarter to understand more about the life that already exists, before going off and creating something totally new...
I think it would be interesting to take a person's stem cell and try to remove all the "junk DNA" from the nucleus, then grow the cell thru a few generations (perhaps even to a full clone) and see how different it is from the original person. Very likely a lot of what we think is junk DNA isn't useless after all. Probably the reason we have 46 chromosomes in the first place is that we've been accumulating genetic material from other microbes over the span of millions of years...
And what exactly is the lowest denominator that you see here? Heck even back in 1993 we could do 8 channel double-buffered digital audio glitch-free on an Atari Falcon. Why should the fact that a network interrupt handler needs a few extra CPU cycles be an excuse for audio glitches to occur?