I'd have bought several Epia boards by now if they had just put a useful number of Ethernet ports on board. Their 2-port boards look like one-offs and do not inspire. I want a board with 3 ports. If you're going to talk about "server appliance" you need 2+ network ports... Yet, from the photos I see here their latest stuff has
in the process of buying the pieces to start their own ERP suite
Oracle had a successful ERP platform years before they bought PeopleSoft. ERP is old hat for Oracle. The recent "fusion" work is their attempt to produce a new platform to replace the now rather mature Oracle ERP platform and provide a road for their recently acquired PeopleSoft and JDE customers.
As far as TCO costs go, I wouldn't be surprised if Oracle was cheaper. The stack is, while highly proprietary, fairly streamlined compared to SAP.
This is a prototype. It will be refined with lighter, stronger material and a viable power supply. The algorithms will be improved. It will acquire vision.
Don't be surprised when you learn the eyes have cross hairs.
No. "We" have the vision clearly in mind in all of the cases you mentioned. Gmail, for example, is available to you from almost any contemporary communication device, including your handheld, laptop, public terminal, cell phone, etc. The same functionality is also provided by nearly all other hosted email services, including those one might build for exclusive use. This is the "freedom" users value. Precisely which "ideals" and "important issues" do you suspect have been so "enthusiastically" abandoned?
.49 to.99 furlongs 3.24E-15 to 6.48E-15 parsecs 7407 to 14815 M&Ms (13.5mm nominal diameter, plain)
When a spec utilizes metric units there really is no need to convert from otherwise nice round, sensible figures like 100 and 200 to 328 and 656. While it is true that I, like most US citizens, default to thinking in terms of yards and feet, we're not (contrary to what is often asserted) incapable of coping with the occasional meter.
If I discover an obscure remotely exploitable security flaw in a Microsoft beta product (thus, unlikely to lead to a "critical" advisory,) why should I not sit on it until a few months after release and get paid?
Accuracate (sic) placement is not a high priority with fission bombs.
Although this is widely believed, it is not true. Placement is absolutely crucial to nuclear weapon effectiveness.
Nuclear detonations are large but still finite. Tactical weapons are deployed using highly accurate warheads because they must have a small CEP (on the order of a few hundred meters) to assure destruction of hardened targets. So called strategic weapons must be detonated at altitude to maximize damage.
Obstructions such as hills can greatly reduce the effect of overpressure. This was observed in Nagasaki, which was spared a large amount of devastation due to terrain. Thermal radiation can be stopped by terrain and even reduced by heavy clouds.
A small, or at least inefficient, nuclear detonation at sea level in a port will produce far less devastation than an optimally placed weapon. An interesting study of what might occur if New York suffered a hypothetical 150 kiloton detonation a ground level is found here. The result is 1.7 million casualties (800,000 dead) in a very densely populated city of 8 million. A detonation some arbitrary distance off shore of a less densely populated area, possibly mitigated by terrain and/or weather (Seattle, for instance,) would be far less effective. You might end up with total casually figures of a few hundred thousand. About average for a large scale carpet bombing operation during WW2.
My point is that placement is paramount to nuclear weapon effectiveness. Damage from haphazard detonations in ports will be relatively limited. More important is what happens after such an attack.
Interesting question. I've been lurking around blade platforms lately and I'm not happy. Why the proprietary chassis? Why the internal storage?
My ideal "blade" system mounts in standard 19" racks. 2 or 3 complete systems in 1U, 48VDC powered from another 1U transformer. Let me stack 1-8 of these in my rack and don't make me pay for the damn IBM/HP/etc chassis.
My system also has no storage inside the blades. Just give me 4 network interfaces per "blade", with at least 2 optionally capable of providing ISCSI TOE/HBA or 2/4 gigabit FC. No SATA/SCSI bus hardware to pay for, cool, power or otherwise. No CD/DVD/Floppy nonsense either. If I briefly need removable storage I'll use USB, thank you very much.
This doesn't exist as far as I can tell. If someone knows better, please chime in.
Yes. EVE has only one "server", which is a cluster of IBM hardware with a large Texas Memory Systems solid state disk. I'm not certain what operating system is running on the cluster nodes, but I know the database is MS SQL Server.
The game is implemented in so-called "stackless" Python. I believe they are using a now rather obsolete version of stackless. I continue to wonder when and how they will address that problem. Perhaps they have been maintaining an internal stackless Python fork...
Why not attack a communications company? Drag their asses into court and make them face judges. They can be made to respond. Get what you can into the public record. Time has a way of using such evidence. Find out if there is any low hanging fruit. If not then maybe this isn't worth pursuing.
I'm glad to see that the EFF has the confidence and resources to pursue such things. Props to the EFF. Enjoy.
Large data centers often have far more than 2400 operational disks. Under these conditions, at any given moment, some fraction of all storage has faulted and repair activity is continuous. This is one reason SCSI hardware is preferred: the disks are more uniform (capacity, electrical interface, etc.,) and replacements remain available over longer intervals.
This isn't the slightest bit unusual. At any moment some fraction of the power transmission and distribution system has faulted. Some percentage of all aircraft are grounded. Various segments of all wide area communications systems are down. Repairs never cease.
$350 equates to a few minutes of aggregate labor costs spent financing, provisioning, securing and monitoring a petabyte of storage. Other large ongoing costs include power and cooling. $350/day is lost in the noise.
EMC's new offering will reduce many of these costs for a given amount of storage. The thing to do then is build data centers to host these machines by the dozen.
Is there really a $140 billion dollar opportunity here? Does Itanium really offer something so superior to other available platforms that its creators are justified in believing they can acquire a large fraction of the market?
Itanium, a high-end processor, was once expected to sweep the server world. But because of delays, initial performance issues and software incompatibilities...
No mention of cost. Itanium is expensive. This fact obviates most of the server market that is well served by cheaper platforms. Why pay a premium for performance you don't need? Dollar for dollar, Itanium is slow. Until that problem is fixed they are throwing good money after bad. The market that was once willing to fund large margins for business hardware is gone.
In business computing performance is fungible. Today, I can adequately run most business software on Intel x86, Itanium, SPARC, PA-RISC, Opteron and POWER. It's understood my servers must be production grade equipment. Once that is assured the remaining question is simple; who provides the most performance per dollar? Itanium, as far as I can tell, has never even attempted to compete in that space.
There certainly are customers that need as much performance as possible from every core, damn the cost. This, however, is merely a niche. For typical business applications cost is an imperative with every purchase. Until Itanium is cost effective relative to the competition it will remain in a niche. A small, dwindling niche.
BTW, I think Sun may have a winner on its hands with the UltraSPARC T1. 8 cores with 8GiB of RAM in 2U for $13k. That could handle a lot of SIP streams with Asterisk. Anyone tried this?
...surely they couldn't be doing much more than...
There are no promises here.
Imagine something like a large rack mount UPS with legs. If it detects you it aims and fires. It's ambling along the street at night, along with the other several hundred deployed in the area.
No, it isn't likely to happen to you. Eventually, however, it will happen to someone, somewhere.
I tried. I really did. I just couldn't generate any concern particles reading this.
I have my own IP battles here in geek-land. If baseball fans are suddenly forced to become aware of the IP madness all around them, great. I suspect they won't be much use however; the MLB must only convince fans of a correlation between their IP rights and the construction of some new stadium or the acquisition of some uber pitcher.
This is news about IP and Baseball. The requisite parties are: lawyers and sports geeks. I'd rather the story appear on the non-MLB owned baseball fanboy sites (are there any?) and not here.
Read the CTCP document and you discover FreeBSD used as a router to simulate conditions of high speed, high latency/loss network links via DummeyNet. The results of this work lead directly to new algorithms that will probably appear in Vista.
James E. Carter and introvert? The guy who, for most of my lifetime, has been gallivanting around the planet advising and negotiating with anyone drawing breath, publicly criticizing all of his successors, participating in all manner of charitable causes, special interests and important events, and doing interviews for anyone capable of granting airtime? That "Jimmy" Carter? This guy visited TMI while the core was still molten.
Give me a ****ing break. The man has probably forgotten more friends than any ten of you will ever have.
from any of the various places sh/bash/etc source:
LESS='X'; export LESS
Now, Less and vim won'trestore the @#*$!%ing terminal on exit, permitting you to cut/paste/transcribe whatever you were just editing/viewing.
(whomever caused this behavior to be default; a pox on you)
p.s. Some bonehead in Usenet advises frobbing your terminal type to vt100 to get the same result. Do not do this. If you don't know why then especially don't do this!
Who says? Where has the EU or any other litigant specified sublicensing as imperative? I'd like to think they had this much insight. I really would. Until this is demonstrated I suspect any claims. Microsoft managed to bamboozle the EU once with the MediaPlayer settlement.
You presume a worm will "take a random address". Real IPv6 worms will do no such thing. IPv6 address allocation will form natural sequential groups. Any half-wit can produce an algorithm that searches the space immediately before and after the address of the compromised host. Never mind sniffing traffic (broadcasts, etc) to discover other hosts...
Slashdot Mirror
that has multiple Ethernet interfaces
I'd have bought several Epia boards by now if they had just put a useful number of Ethernet ports on board. Their 2-port boards look like one-offs and do not inspire. I want a board with 3 ports. If you're going to talk about "server appliance" you need 2+ network ports... Yet, from the photos I see here their latest stuff has
wait for it
1 port.
Sigh.
in the process of buying the pieces to start their own ERP suite
Oracle had a successful ERP platform years before they bought PeopleSoft. ERP is old hat for Oracle. The recent "fusion" work is their attempt to produce a new platform to replace the now rather mature Oracle ERP platform and provide a road for their recently acquired PeopleSoft and JDE customers.
As far as TCO costs go, I wouldn't be surprised if Oracle was cheaper. The stack is, while highly proprietary, fairly streamlined compared to SAP.
I squirm
Likewise, and I don't hesitate to admit it.
This is a prototype. It will be refined with lighter, stronger material and a viable power supply. The algorithms will be improved. It will acquire vision.
Don't be surprised when you learn the eyes have cross hairs.
No. "We" have the vision clearly in mind in all of the cases you mentioned. Gmail, for example, is available to you from almost any contemporary communication device, including your handheld, laptop, public terminal, cell phone, etc. The same functionality is also provided by nearly all other hosted email services, including those one might build for exclusive use. This is the "freedom" users value. Precisely which "ideals" and "important issues" do you suspect have been so "enthusiastically" abandoned?
The cynic in me envisions a trial lawyer engaging the survivors of people killed in a theater fire, cell phone in hand...
.49 to .99 furlongs
3.24E-15 to 6.48E-15 parsecs
7407 to 14815 M&Ms (13.5mm nominal diameter, plain)
When a spec utilizes metric units there really is no need to convert from otherwise nice round, sensible figures like 100 and 200 to 328 and 656. While it is true that I, like most US citizens, default to thinking in terms of yards and feet, we're not (contrary to what is often asserted) incapable of coping with the occasional meter.
If I discover an obscure remotely exploitable security flaw in a Microsoft beta product (thus, unlikely to lead to a "critical" advisory,) why should I not sit on it until a few months after release and get paid?
Accuracate (sic) placement is not a high priority with fission bombs.
Although this is widely believed, it is not true. Placement is absolutely crucial to nuclear weapon effectiveness.
Nuclear detonations are large but still finite. Tactical weapons are deployed using highly accurate warheads because they must have a small CEP (on the order of a few hundred meters) to assure destruction of hardened targets. So called strategic weapons must be detonated at altitude to maximize damage.
Obstructions such as hills can greatly reduce the effect of overpressure. This was observed in Nagasaki, which was spared a large amount of devastation due to terrain. Thermal radiation can be stopped by terrain and even reduced by heavy clouds.
A small, or at least inefficient, nuclear detonation at sea level in a port will produce far less devastation than an optimally placed weapon. An interesting study of what might occur if New York suffered a hypothetical 150 kiloton detonation a ground level is found here. The result is 1.7 million casualties (800,000 dead) in a very densely populated city of 8 million. A detonation some arbitrary distance off shore of a less densely populated area, possibly mitigated by terrain and/or weather (Seattle, for instance,) would be far less effective. You might end up with total casually figures of a few hundred thousand. About average for a large scale carpet bombing operation during WW2.
My point is that placement is paramount to nuclear weapon effectiveness. Damage from haphazard detonations in ports will be relatively limited. More important is what happens after such an attack.
Courts do not like people telling them that the law does not apply to them.
vee hauf vays oft deeling vit chu
Don't taunt the Court.
Interesting question. I've been lurking around blade platforms lately and I'm not happy. Why the proprietary chassis? Why the internal storage?
My ideal "blade" system mounts in standard 19" racks. 2 or 3 complete systems in 1U, 48VDC powered from another 1U transformer. Let me stack 1-8 of these in my rack and don't make me pay for the damn IBM/HP/etc chassis.
My system also has no storage inside the blades. Just give me 4 network interfaces per "blade", with at least 2 optionally capable of providing ISCSI TOE/HBA or 2/4 gigabit FC. No SATA/SCSI bus hardware to pay for, cool, power or otherwise. No CD/DVD/Floppy nonsense either. If I briefly need removable storage I'll use USB, thank you very much.
This doesn't exist as far as I can tell. If someone knows better, please chime in.
Simultaneously both a question and a statement!
The WHAT working draft Web Applications 1.0 certainly looks like Ajax with its XMLHttpRequest, drag and drop and other Ajaxish stuff.
Do they mean 23000+ people on one server?
Yes. EVE has only one "server", which is a cluster of IBM hardware with a large Texas Memory Systems solid state disk. I'm not certain what operating system is running on the cluster nodes, but I know the database is MS SQL Server.
The game is implemented in so-called "stackless" Python. I believe they are using a now rather obsolete version of stackless. I continue to wonder when and how they will address that problem. Perhaps they have been maintaining an internal stackless Python fork...
The Alternative is called picking your battles.
Why not attack a communications company? Drag their asses into court and make them face judges. They can be made to respond. Get what you can into the public record. Time has a way of using such evidence. Find out if there is any low hanging fruit. If not then maybe this isn't worth pursuing.
I'm glad to see that the EFF has the confidence and resources to pursue such things. Props to the EFF. Enjoy.
Large data centers often have far more than 2400 operational disks. Under these conditions, at any given moment, some fraction of all storage has faulted and repair activity is continuous. This is one reason SCSI hardware is preferred: the disks are more uniform (capacity, electrical interface, etc.,) and replacements remain available over longer intervals.
This isn't the slightest bit unusual. At any moment some fraction of the power transmission and distribution system has faulted. Some percentage of all aircraft are grounded. Various segments of all wide area communications systems are down. Repairs never cease.
$350 equates to a few minutes of aggregate labor costs spent financing, provisioning, securing and monitoring a petabyte of storage. Other large ongoing costs include power and cooling. $350/day is lost in the noise.
EMC's new offering will reduce many of these costs for a given amount of storage. The thing to do then is build data centers to host these machines by the dozen.
This is a $140 billion opportunity on hardware.
Is there really a $140 billion dollar opportunity here? Does Itanium really offer something so superior to other available platforms that its creators are justified in believing they can acquire a large fraction of the market?
Itanium, a high-end processor, was once expected to sweep the server world. But because of delays, initial performance issues and software incompatibilities...
No mention of cost. Itanium is expensive. This fact obviates most of the server market that is well served by cheaper platforms. Why pay a premium for performance you don't need? Dollar for dollar, Itanium is slow. Until that problem is fixed they are throwing good money after bad. The market that was once willing to fund large margins for business hardware is gone.
In business computing performance is fungible. Today, I can adequately run most business software on Intel x86, Itanium, SPARC, PA-RISC, Opteron and POWER. It's understood my servers must be production grade equipment. Once that is assured the remaining question is simple; who provides the most performance per dollar? Itanium, as far as I can tell, has never even attempted to compete in that space.
There certainly are customers that need as much performance as possible from every core, damn the cost. This, however, is merely a niche. For typical business applications cost is an imperative with every purchase. Until Itanium is cost effective relative to the competition it will remain in a niche. A small, dwindling niche.
BTW, I think Sun may have a winner on its hands with the UltraSPARC T1. 8 cores with 8GiB of RAM in 2U for $13k. That could handle a lot of SIP streams with Asterisk. Anyone tried this?
I still can't believe Bush hasn't even *apolagized* (sic) for breaking our fundamental American rights
It has been six years and people still hold out hope Bush can be harangued into apologies.
Yay for persistence.
And they both smoked tobacco!
...must work Sport Utility Vehicles into this somehow...
...surely they couldn't be doing much more than...
There are no promises here.
Imagine something like a large rack mount UPS with legs. If it detects you it aims and fires. It's ambling along the street at night, along with the other several hundred deployed in the area.
No, it isn't likely to happen to you. Eventually, however, it will happen to someone, somewhere.
I tried. I really did. I just couldn't generate any concern particles reading this.
I have my own IP battles here in geek-land. If baseball fans are suddenly forced to become aware of the IP madness all around them, great. I suspect they won't be much use however; the MLB must only convince fans of a correlation between their IP rights and the construction of some new stadium or the acquisition of some uber pitcher.
This is news about IP and Baseball. The requisite parties are: lawyers and sports geeks. I'd rather the story appear on the non-MLB owned baseball fanboy sites (are there any?) and not here.
Read the CTCP document and you discover FreeBSD used as a router to simulate conditions of high speed, high latency/loss network links via DummeyNet. The results of this work lead directly to new algorithms that will probably appear in Vista.
James E. Carter and introvert? The guy who, for most of my lifetime, has been gallivanting around the planet advising and negotiating with anyone drawing breath, publicly criticizing all of his successors, participating in all manner of charitable causes, special interests and important events, and doing interviews for anyone capable of granting airtime? That "Jimmy" Carter? This guy visited TMI while the core was still molten.
Give me a ****ing break. The man has probably forgotten more friends than any ten of you will ever have.
(whomever caused this behavior to be default; a pox on you)
p.s. Some bonehead in Usenet advises frobbing your terminal type to vt100 to get the same result. Do not do this. If you don't know why then especially don't do this!
Anything less won't cut it.
Who says? Where has the EU or any other litigant specified sublicensing as imperative? I'd like to think they had this much insight. I really would. Until this is demonstrated I suspect any claims. Microsoft managed to bamboozle the EU once with the MediaPlayer settlement.
Yo you know how large IPv6 space is?
You're advocating security through obscurity. Please stop now... mkaythnx, yo and stuff.
Bollocks.
You presume a worm will "take a random address". Real IPv6 worms will do no such thing. IPv6 address allocation will form natural sequential groups. Any half-wit can produce an algorithm that searches the space immediately before and after the address of the compromised host. Never mind sniffing traffic (broadcasts, etc) to discover other hosts...