Hey... my (primary) video card as well (I also have two Riva TNTs for monitors 2 and 3). It's plenty fast if your main workload consists of a few xterms and/or Mozilla... even does Unreal Tournament on WinXP decently. Nearly 3 years old and still going strong! (Actually, my oldest hardware still in semi-regular use is a monitor, CD-ROM drive and video card from a Dell P166, circa 1996 - not in my main workstation though).
I've got 3 CRTs (1 19in in the middle plus 2 17in) on 3 video cards in my Linux box and I must say it kicks ass for development, among other things.
I have XMMS and GAIM on my left screen, two xterms on my 19in in the middle, and online docs or two more xterms on the right.
When doing homework I have VMWare running Word (sad I know, but required to exchange docs with school account) on one screen, and then I can pop up Mozilla for some research, or check my mail quickly, on another screen.
It is amazingly useful to be able to have three maximized windows at once, and even more smaller windows (XMMS, GAIM, small xterm, etc.). Two everyone who says they don't "get" multihead: you will get it once you have enough space that you never need to have one window on top of another - everything visible at once. Even better with virtual desktops.
smtp.central.cox.net isn't authorized for my domain
The proposed change is to require (presumably with assume-authorized for backwards compatibility, at least for a while) the _sending_ domain's DNS to have record(s) for IPs that are allowed to send mail _from_ those domains. So, all you have to do if you have your own domain is to add your IP, or your ISP's SMTP server's IP, to your domain's DNS.
In this case, you'd just have a record in your DNS with smtp.central.cox.net's IP, allowing it to legitimately send with "your.domain" as the source.
Sheesh, calm down - it's only an apostrophe, and I doubt it takes you 10 min - 20 s = 9:40 to hit that one extra key. It's always nice to avoid typos when possible, but Slashdot doesn't necessitate perfect composition. I certainly don't care whether or not posts sound like formal essays.
In some states, it is no longer a requirement to know how to parallel park
True in Oregon - it's been replaced by parking and then backing along a curb (at least that's what my driving instructor said - I'm not 16 yet and so haven't taken the test).
Re:Sounds like a NYC black out waiting to happen
on
MIT Roofnet
·
· Score: 3, Insightful
Consequently overloading that node and causing the latter to fall off the mesh as well.
I think you're associating a wireless network a little too much with a power grid. Routing everything through one node won't cause it to "fall off the mesh" - it will just start dropping the excess packets. What do you think happens when you send a 100mbps stream of Ethernet packets to a 256k upload cable modem? Same thing. The connection speed of all nodes funneled through a single bottleneck would merely suffer somewhat.
Starting up the network (after a power outage, say) wouldn't necessarily need a certain order either. It just wouldn't reach full speed until all the critical nodes (ones with lots of links to other nodes) came up.
Yes - theoretically (as mentioned elsewhere) it can be solved in 21 (I think) moves. I simply offered the reversing moves strategy as proof that a method always exists.
Actually, the proof (informal) of that is quite easy - it starts solved (when you buy it), and to scramble you rotate the faces, so to solve you simply reverse the moves you did to scramble. If you start flipping corners or edges (an odd number of flips) then it becomes unsolvable, however.
Could you couple WebDAV with SSH to achieve the requisite security?
You read the part about the preferability of no extra client software to install, right? Last I checked, PuTTY on Windows was a separate download. SSL is built into Windows.
Sorry if I'm going against the conventional Slashbot wisdom here, but what exactly about building a standard PC (note the original question - comparable specs, not an exact replica) is illegal under the DMCA?
Last I checked, it was quite easy to legally pick up a few PC parts and slap them in a case...
Hardware floating point is only so accurate - if you need more floating point (or integer) precision, use GNU MP - a library for C with bindings for many other languages too. It came in quite handy when I wrote some cryptography code with very large numbers.
Note "_an_ open source OS" - to keep as even as possible they should have either used Linux on both i386 and ppc, or Darwin on i386 (I'm pretty sure it runs on i386) and ppc.
How about taking that a step further, and benchmarking with an open source OS as well? That way the only variable is the processor/system itself. The Linux kernel runs on both x86 and ppc64...
I have to say that this is one of the more fun areas of development (having been involved for the last 3 yrs or so). To anyone wanting to get started I would highly recommend the newsgroup alt.os.development - many of the members there also have very informative pages.
(sorry for posting kinda late, I was busy working on my OS!)
Re:No more car tinkering...
on
42-Volt Autos
·
· Score: 1
constant of proportionality is resistance
Sorry, should be 1/resistance - higher resistance == less current.
Re:No more car tinkering...
on
42-Volt Autos
·
· Score: 1
Thus, a high amperage at a low voltage is just as deadly as a low amperage at a high voltage
Right - current will cause fibrillation, but even low current at high power will cause the dissipated heat to burn your flesh.
But - remember Ohm's law? Current is directly proportional to voltage - the constant of proportionality is resistance. Human skin has a fairly high resistance, thus 12v or even 48v won't cause very much current to flow (for dry skin, less than a milliamp).
I can't speak for the UK, but at least here in the US it's actually closer to 240V (for the total voltage coming into the house). I just put a voltmeter across an outlet here, and I got 118V (and a single outlet is one-half of the total voltage - so about 236V).
Symmetric multiprocessor - two or more processors sharing a common memory space and peripherals (sometimes a common FSB too). Basically, all multiprocessor PCs.
If you think about it, how is this any different than, say, an Apache log, or an intrusion detection system log? Well, the data is recorded on the server and not on the client, but other than that it's basically the same. What happens when someone attempts to crack a system? The logs are evidence. With black boxes it would be exactly the same in a car accident.
It is equally true that "If we can access something, we can copy it". No one ever seems to make that logic leap, either. Especially as applies to DVD's - If we have to decrypt them to watch them, we can also copy them. Or those copy protected audio CD's - if we can hear them, we can copy them.
You're assuming, by saying that anything heard can be copied, that the destination medium doesn't require things like digital signatures. Although your two examples, CDs and DVDs, don't require signing, things like console games do (thus the need for a modchip). The only way to get around a restriction like this would be to copy a disc bit-for-bit, which is many times not an option with proprietary formats.
OK, let me expand on that: it's specifically bad for open source software, because most open source licenses (the GPL and the BSD license at least) contain a "no guarantee" clause.
If this goes through, it could set a precedent of liability for software bugs... that's bad, of course.
Here's an interesting thought: maybe closed source software could be hit harder by this because keeping the source closed could be considered hiding the vulnerability? IANAL, of course.
Another thing - aren't there liability issues for engineers in other fields as well - like holding a bridge engineer accountable if the whole thing falls down? Of course, a software bug isn't quite that serious, but still...
Slashdot Mirror
Hey... my (primary) video card as well (I also have two Riva TNTs for monitors 2 and 3). It's plenty fast if your main workload consists of a few xterms and/or Mozilla... even does Unreal Tournament on WinXP decently. Nearly 3 years old and still going strong! (Actually, my oldest hardware still in semi-regular use is a monitor, CD-ROM drive and video card from a Dell P166, circa 1996 - not in my main workstation though).
I've got 3 CRTs (1 19in in the middle plus 2 17in) on 3 video cards in my Linux box and I must say it kicks ass for development, among other things.
I have XMMS and GAIM on my left screen, two xterms on my 19in in the middle, and online docs or two more xterms on the right.
When doing homework I have VMWare running Word (sad I know, but required to exchange docs with school account) on one screen, and then I can pop up Mozilla for some research, or check my mail quickly, on another screen.
It is amazingly useful to be able to have three maximized windows at once, and even more smaller windows (XMMS, GAIM, small xterm, etc.). Two everyone who says they don't "get" multihead: you will get it once you have enough space that you never need to have one window on top of another - everything visible at once. Even better with virtual desktops.
smtp.central.cox.net isn't authorized for my domain
The proposed change is to require (presumably with assume-authorized for backwards compatibility, at least for a while) the _sending_ domain's DNS to have record(s) for IPs that are allowed to send mail _from_ those domains. So, all you have to do if you have your own domain is to add your IP, or your ISP's SMTP server's IP, to your domain's DNS.
In this case, you'd just have a record in your DNS with smtp.central.cox.net's IP, allowing it to legitimately send with "your.domain" as the source.
Sheesh, calm down - it's only an apostrophe, and I doubt it takes you 10 min - 20 s = 9:40 to hit that one extra key. It's always nice to avoid typos when possible, but Slashdot doesn't necessitate perfect composition. I certainly don't care whether or not posts sound like formal essays.
In some states, it is no longer a requirement to know how to parallel park
True in Oregon - it's been replaced by parking and then backing along a curb (at least that's what my driving instructor said - I'm not 16 yet and so haven't taken the test).
Consequently overloading that node and causing the latter to fall off the mesh as well.
I think you're associating a wireless network a little too much with a power grid. Routing everything through one node won't cause it to "fall off the mesh" - it will just start dropping the excess packets. What do you think happens when you send a 100mbps stream of Ethernet packets to a 256k upload cable modem? Same thing. The connection speed of all nodes funneled through a single bottleneck would merely suffer somewhat.
Starting up the network (after a power outage, say) wouldn't necessarily need a certain order either. It just wouldn't reach full speed until all the critical nodes (ones with lots of links to other nodes) came up.
Yes - theoretically (as mentioned elsewhere) it can be solved in 21 (I think) moves. I simply offered the reversing moves strategy as proof that a method always exists.
Actually I would think -1 : less than 0 : equal to 1 : greater than Would make sorting interesting.
You've reinvented a wheel! C's strcmp (string compare) returns exactly those values.
I never knew they were actually solvable
Actually, the proof (informal) of that is quite easy - it starts solved (when you buy it), and to scramble you rotate the faces, so to solve you simply reverse the moves you did to scramble. If you start flipping corners or edges (an odd number of flips) then it becomes unsolvable, however.
DB connection failed ().
Um, that would be the Slashdot effect most likely...
Could you couple WebDAV with SSH to achieve the requisite security?
You read the part about the preferability of no extra client software to install, right? Last I checked, PuTTY on Windows was a separate download. SSL is built into Windows.
Sorry if I'm going against the conventional Slashbot wisdom here, but what exactly about building a standard PC (note the original question - comparable specs, not an exact replica) is illegal under the DMCA?
Last I checked, it was quite easy to legally pick up a few PC parts and slap them in a case...
Not sure if RH9 has NTFS in the stock kernel (if not you'll have to recompile) - but it's just a normal filesystem driver. try:
/dev/hdXX /mnt_point
/etc/fstab too.
# mount -t ntfs
Of course you can add the mountpoint to your
Hardware floating point is only so accurate - if you need more floating point (or integer) precision, use GNU MP - a library for C with bindings for many other languages too. It came in quite handy when I wrote some cryptography code with very large numbers.
Note "_an_ open source OS" - to keep as even as possible they should have either used Linux on both i386 and ppc, or Darwin on i386 (I'm pretty sure it runs on i386) and ppc.
How about taking that a step further, and benchmarking with an open source OS as well? That way the only variable is the processor/system itself. The Linux kernel runs on both x86 and ppc64...
I have to say that this is one of the more fun areas of development (having been involved for the last 3 yrs or so). To anyone wanting to get started I would highly recommend the newsgroup alt.os.development - many of the members there also have very informative pages.
(sorry for posting kinda late, I was busy working on my OS!)
constant of proportionality is resistance
Sorry, should be 1/resistance - higher resistance == less current.
Thus, a high amperage at a low voltage is just as deadly as a low amperage at a high voltage
Right - current will cause fibrillation, but even low current at high power will cause the dissipated heat to burn your flesh.
But - remember Ohm's law? Current is directly proportional to voltage - the constant of proportionality is resistance. Human skin has a fairly high resistance, thus 12v or even 48v won't cause very much current to flow (for dry skin, less than a milliamp).
I can't speak for the UK, but at least here in the US it's actually closer to 240V (for the total voltage coming into the house). I just put a voltmeter across an outlet here, and I got 118V (and a single outlet is one-half of the total voltage - so about 236V).
Smp systems?
Symmetric multiprocessor - two or more processors sharing a common memory space and peripherals (sometimes a common FSB too). Basically, all multiprocessor PCs.
If you think about it, how is this any different than, say, an Apache log, or an intrusion detection system log? Well, the data is recorded on the server and not on the client, but other than that it's basically the same. What happens when someone attempts to crack a system? The logs are evidence. With black boxes it would be exactly the same in a car accident.
It is equally true that "If we can access something, we can copy it". No one ever seems to make that logic leap, either. Especially as applies to DVD's - If we have to decrypt them to watch them, we can also copy them. Or those copy protected audio CD's - if we can hear them, we can copy them.
You're assuming, by saying that anything heard can be copied, that the destination medium doesn't require things like digital signatures. Although your two examples, CDs and DVDs, don't require signing, things like console games do (thus the need for a modchip). The only way to get around a restriction like this would be to copy a disc bit-for-bit, which is many times not an option with proprietary formats.
OK, let me expand on that: it's specifically bad for open source software, because most open source licenses (the GPL and the BSD license at least) contain a "no guarantee" clause.
If this goes through, it could set a precedent of liability for software bugs... that's bad, of course.
Here's an interesting thought: maybe closed source software could be hit harder by this because keeping the source closed could be considered hiding the vulnerability? IANAL, of course.
Another thing - aren't there liability issues for engineers in other fields as well - like holding a bridge engineer accountable if the whole thing falls down? Of course, a software bug isn't quite that serious, but still...