I am presuming that PDN will not like the OpenBSD install process, and that PDN will not like using the ports tree.
I don't mean to be insulting in any way. I don't like the ports tree, nor do I like the OpenBSD installer. I'm just trying to extrapolate from my own experience and the (admittedly shallow) impression I have of PDN's likes and dislikes.
As for getting pissed of, chill. It's only software. Most software will piss someone off at some time or another because all software sucks. The whole key for me is to stay mellow when things go bad. You know, take a walk for 15 minutes, clear your head, and come back with a fresh face to fix the damn thing.
This is really the traditional way that UNIX does things, and it's been refined to a high degree. It works _very_ well. However, is it such a good idea to route latency-sensitive data through the file subsystem.
Atheos claims that it's messaging system is both fast and responsive making the GUI much faster than X overall. How does the Atheos scheme compare to Beos and QNX/Neutrino? How much faster than X? Do we have anything to make apple2apple comparisons?
It seems to me that using (exclusively) a low latency setup like this would allow you to optimize the dispatch of your draw commands. Would this not make the system much snappier overall?
For Office '03 docs, it may turn out that this is so, However, a lot of the push - and controversy - behind.NET was it's XML storage and messaging facilities for web apps. If you believe that was going to be completely non-proprietary, I could probably sell you some prime tundra up here in Canada.
"Embrace and Extend" is such a wonderfully misleading concept. On the surface, it appears that Microsoft is a big proponent of open standards. They're members of some of the more influential standards boards, and have actually deigned to submit the C# spec to the ECMA.
However, the pivotal falsity of the whole enterprise is that underneath a thin vineer, every Microsoft app is locked tightly by the virtue of legality and obfuscated design.
Possibly it's a tension between management (who want to protect sales) and the younger stafflings (who want to work with the Next Best Thing). I prefer to imagine a small cadre of techno-literates whose job is to peruse worldwide resources, looking for something up and coming to latch onto and connect Windows with in a superficial, candyass way.
I believe it's these people who are also responsible for product naming: DirectX ActiveX XBox Windows XP
Wtf?
But I digress. The point is that every product produced by Microsoft has exactly the same nuts and bolts of half a dozen competitors, but these constituent parts are used in completely non-compatible ways. So while Microsoft can claim that they are "Standards Based" or whatever, their products will never gain the benefit of following standards -- interoperability.
1) By many popular definitions, Linux and BSD are unices. The announcement that "UNIX is dead" is too sweeping a term to be safely used.
2) We have no way to quantify what differences in performance are attributable to software rather than hardware in the given example, nor does one anecdotal application constitute a complete comparison between Solaris/SPARC and RHL/ia32.
This article seems to have more to do with squabbles between Dell and the traditional iron peddlers over market share in the enterprise sector than anything else.
As other's have pointed out, this guy partially duplicated the functionality of a PVR for more money than he would have spent. Why is this a good thing?
What using a computer gains you over a PVR is extensibility. If you add an optical drive (DVD or combo DVD/CDR) then your machine doubles as a DVD, CD, and VCD player. You can even start gaming on the machine rather than buying an expensive PS2, Xbox, or Cube.
Instead of having shelves of black boxes to get all your jollies, you can integrate most of the capabilities you want into a single, clean aluminum case and even add some new ones.
Re:How does it compare on windows?
on
Mesa 5.0 Released
·
· Score: 1
Yes, that would be interesting. The nVidia drivers support a larger feature set than many other commercial OpenGL implementations.
I think the more interesting question will be 'when will DRI begin to develop with Mesa 5?' Only when libGL.so is hardware accelerated and, therefore, usable does anyone care about feature sets. Unfortunately, the DRI with X4.2 is based on Mesa 3, which doesn't cut the mustard these days.
Hopefully X4.3 will be released with a very recent trunk build of DRI. TCL support for R100s, at least, should be supported.
p.s. Does anyone from Deb or perhaps the XSF currently package CVS snapshots of the XFree86 tree. Sure, packaging X is not nice, but an unstable package of X for i386 would still be kind of cool for the lazy twiddlers.
The guys on the Beowulf mailing list dealt with this a while ago. The general consensus was HT would be effective in applications which have, for instance, threads which do I/O and others which do computation, as well as programs in which memory latency is an issue. Additionally, it was postulated that HT would enable simultaneous int and fp operations. Generally, they found compute bound problems received no benefit -- there is physically no significant mathematical capability added to Hyperthreading CPUs.
The thread begins here. There are some posts which aren't in the thread, but are relevant. Find them here
Down in the basement of my University, there's still a MicroVax in service. The IT guys claim it's tape drive is still useful or something.
The thing is the size of a small refrigerator, as robustly built, and uses a 220V line. I was amazed when I learned that it used 220 -- it sucks power like a furnace.
Ace's Hardware reviews are clearly the best and most researched, but they are few and far between. Want an excellent review of current and future memory technologies written with the help of actual engineers? Read Ace's Hardware. Ars rarely has hardware reviews, but when they do the reviews are good. Anandtech is a good all-around major review site that as far as I can tell has never been biased, but is a little bit too PC for me. (that's Politically Correct, not the other one)
I was totally impressed with Ace's coverage of the various memory technologies. They explained the actual labyrinthian workings of all the modern memory systems in terms any technically literate person could understand, unlike Engineerese white papers. Great stuff. They're not the biggest on massive sets of benchmarks, nor the fastest to review a new product, but I can live with that.
Anandtech is good, but [H]ardOCP, surprisingly, is quite non-biased. Although they DO review from the point of view of the aftermarket, their conclusions are generally backed up by empirical evidence, and they make sure to qualify their results with reminders of what a particular benchmark is stressing. Every review is generally conducted with a fresh face, giving everyone who wants to play their game a chance.
Honestly, I've never followed Ars Technica all that much. It always seemed more of a trade magazine to me than a site about hardware, and I've got/. for my general purpose news.
Sen. Danforth: "There is nothing on the face of the album which would
notify you if the record has pornographics material or
material glorifying violence?" Tipper Gore: "No, there is nothing that would suggest that to me." Frank Zappa: "I would say that a buzz saw blade between the guy's legs on
the album cover is good indication that it's not for little
Johnny."
-- The Senate Commerce Committee hearing on rock
lyrics, from The Village Voice, 6 Oct 1985
Except for a few rare exceptions, politicians and newspersons never provide unbiased information to the public at large. Finding easy answers is too profitable for them to ever stop.
How many times have I heard this? "I bought an ATI card fifty years ago, and the driver updates were slow, and the drivers were bad, and..."
This is an industry where silicon is updated every 6 months. Drivers are released continually. Most importantly, NVidia was hardly a blip on the map when ATI was producing Mach64s in quantity. Now they're one of the top chip suppliers of the PC graphics industry, cutting into ATI's OEM lifeblood.
The arrival of NVidia forced ATI to reexamine their game. They've made genuine improvements in all aspects of their company -- including a vastly improved driver team, and a product cycle competitive with NVidias.
The reality is that, for the time being, ATIs card IS the latest and greatest, (free) driver updates from ATI produce meaningful improvements in quality and performance and, in general, the company has transformed.
ATI is a contender, and ignoring ATI cards as a possible component in your computer based on an outdated, mostly anecdotal experience serves only to hurt yourself in the long run.
I've seen OpenBSD hotswap identical disks. Once, at a Unix User's Group meeting, we decided to experiment.
The disk was unplugged, the plugged back in. OBSD said it was reinitializing the controller. Poof! back to normal. Still, this probably would not be a good general practice to follow.
New iBook units are lighter and have a longer battery life, excluding firewire, not to mention cheaper than most Intel machines. IMHO, they're the best product in the Apple line.
As noted above, a 333 MHz front side bus is not really stock nowadays.
However, if you want to void your warranty, Epox 8h3a's with recent firmware have a 1/5 and a (rumoured) 1/6 pci divider. This means that you can overclock the FSB to 166Mhz (333 DDR) on your motherboard while still maintaining a stock rate for the PCI bus (166/5=33), thereby improving the stability/survival rate of your PCI and AGP components.
With a good AXP 1.6 and the right cooling, 333 should be within reach.
Heh, there should be another article on/. about practical overclocking and the benefits of water cooling for the mainstream user.
The real problem, I think, lies with the inability of business to reconcile itself with an information economy. Groups like the MPAA (and Unisys, and RSA Data Sec.) want to provide information as a product. To do so, they have to control the availability of this 'product' to the end user.
In every other part of the Universe, it is the specific product which is patented. If you make a carbon copy of a Honda Accord, then you'll get sued. If you make a vehicle with four wheels, four doors and an engine, you won't. In academia, if you make a discovery, then some time later another person claims to have made the same discovery, that person -- except in rare cases -- you will be laughed out of town, but reproducing the same result via independent work is okay.
So, reproducing the DeCSS algorithm via independent work is okay via logical extension. As for breaking copyright protection, this is really governed by two laws: the so called "Betamax decision" from which the fair-use concept is derived, and by the DMCA. Although I don't know much about the DMCA, fair-use says that any particular consumer of media content can copy it limitlessly for backup, personal storage, alternate viewing, or whatever. Ripping a DVD to DIVX is perfectly legal, as long as you don't redistribute it and merely use it for personal viewing.
Slashdot Mirror
Point taken. I will rephrase.
I am presuming that PDN will not like the OpenBSD install process, and that PDN will not like using the ports tree.
I don't mean to be insulting in any way. I don't like the ports tree, nor do I like the OpenBSD installer. I'm just trying to extrapolate from my own experience and the (admittedly shallow) impression I have of PDN's likes and dislikes.
As for getting pissed of, chill. It's only software. Most software will piss someone off at some time or another because all software sucks. The whole key for me is to stay mellow when things go bad. You know, take a walk for 15 minutes, clear your head, and come back with a fresh face to fix the damn thing.
OpenBSD/NetBSD is definitely not up her alley, and the FreeBSD ports tree would probably scare the bejesus out of her.
Using Doppler Imaging (DI), it is possible to map the surface distribution of various elements on rapidly rotating * CVn stars like E UMa.
Check here, or here for more nfo. Check here for a pretty film.
Well, when you send data over an Unix socket, that data does not go over the file system, any more that write() to a TCP socket does.
Sockets are certainly exposed for access in the VFS layer. It's cool and handy, but I reckon performance will be attenuated by the VFS layer.
Or do I reckon wrong?
It's pre-TOS, with old fashioned nukes and stuff. Should be around the right time period.
Oh, wait. Berman probably has shellshock from Nemesis still.
>It communicates over a local socket.
This is really the traditional way that UNIX does things, and it's been refined to a high degree. It works _very_ well. However, is it such a good idea to route latency-sensitive data through the file subsystem.
Atheos claims that it's messaging system is both fast and responsive making the GUI much faster than X overall. How does the Atheos scheme compare to Beos and QNX/Neutrino? How much faster than X? Do we have anything to make apple2apple comparisons?
It seems to me that using (exclusively) a low latency setup like this would allow you to optimize the dispatch of your draw commands. Would this not make the system much snappier overall?
-G
For Office '03 docs, it may turn out that this is so, However, a lot of the push - and controversy - behind .NET was it's XML storage and messaging facilities for web apps. If you believe that was going to be completely non-proprietary, I could probably sell you some prime tundra up here in Canada.
"Embrace and Extend" is such a wonderfully misleading concept. On the surface, it appears that Microsoft is a big proponent of open standards. They're members of some of the more influential standards boards, and have actually deigned to submit the C# spec to the ECMA.
However, the pivotal falsity of the whole enterprise is that underneath a thin vineer, every Microsoft app is locked tightly by the virtue of legality and obfuscated design.
Possibly it's a tension between management (who want to protect sales) and the younger stafflings (who want to work with the Next Best Thing). I prefer to imagine a small cadre of techno-literates whose job is to peruse worldwide resources, looking for something up and coming to latch onto and connect Windows with in a superficial, candyass way.
I believe it's these people who are also responsible for product naming:
DirectX
ActiveX
XBox
Windows XP
Wtf?
But I digress. The point is that every product produced by Microsoft has exactly the same nuts and bolts of half a dozen competitors, but these constituent parts are used in completely non-compatible ways. So while Microsoft can claim that they are "Standards Based" or whatever, their products will never gain the benefit of following standards -- interoperability.
Two problems with this article:
1) By many popular definitions, Linux and BSD are unices. The announcement that "UNIX is dead" is too sweeping a term to be safely used.
2) We have no way to quantify what differences in performance are attributable to software rather than hardware in the given example, nor does one anecdotal application constitute a complete comparison between Solaris/SPARC and RHL/ia32.
This article seems to have more to do with squabbles between Dell and the traditional iron peddlers over market share in the enterprise sector than anything else.
As other's have pointed out, this guy partially duplicated the functionality of a PVR for more money than he would have spent. Why is this a good thing?
What using a computer gains you over a PVR is extensibility. If you add an optical drive (DVD or combo DVD/CDR) then your machine doubles as a DVD, CD, and VCD player. You can even start gaming on the machine rather than buying an expensive PS2, Xbox, or Cube.
Instead of having shelves of black boxes to get all your jollies, you can integrate most of the capabilities you want into a single, clean aluminum case and even add some new ones.
Yes, that would be interesting. The nVidia drivers support a larger feature set than many other commercial OpenGL implementations.
I think the more interesting question will be 'when will DRI begin to develop with Mesa 5?' Only when libGL.so is hardware accelerated and, therefore, usable does anyone care about feature sets. Unfortunately, the DRI with X4.2 is based on Mesa 3, which doesn't cut the mustard these days.
Hopefully X4.3 will be released with a very recent trunk build of DRI. TCL support for R100s, at least, should be supported.
p.s.
Does anyone from Deb or perhaps the XSF currently package CVS snapshots of the XFree86 tree. Sure, packaging X is not nice, but an unstable package of X for i386 would still be kind of cool for the lazy twiddlers.
The guys on the Beowulf mailing list dealt with this a while ago. The general consensus was HT would be effective in applications which have, for instance, threads which do I/O and others which do computation, as well as programs in which memory latency is an issue. Additionally, it was postulated that HT would enable simultaneous int and fp operations. Generally, they found compute bound problems received no benefit -- there is physically no significant mathematical capability added to Hyperthreading CPUs.
The thread begins here. There are some posts which aren't in the thread, but are relevant. Find them here
Down in the basement of my University, there's still a MicroVax in service. The IT guys claim it's tape drive is still useful or something.
The thing is the size of a small refrigerator, as robustly built, and uses a 220V line. I was amazed when I learned that it used 220 -- it sucks power like a furnace.
Ace's Hardware reviews are clearly the best and most researched, but they are few and far between. Want an excellent review of current and future memory technologies written with the help of actual engineers? Read Ace's Hardware.
/. for my general purpose news.
Ars rarely has hardware reviews, but when they do the reviews are good.
Anandtech is a good all-around major review site that as far as I can tell has never been biased, but is a little bit too PC for me. (that's Politically Correct, not the other one)
I was totally impressed with Ace's coverage of the various memory technologies. They explained the actual labyrinthian workings of all the modern memory systems in terms any technically literate person could understand, unlike Engineerese white papers. Great stuff. They're not the biggest on massive sets of benchmarks, nor the fastest to review a new product, but I can live with that.
Anandtech is good, but [H]ardOCP, surprisingly, is quite non-biased. Although they DO review from the point of view of the aftermarket, their conclusions are generally backed up by empirical evidence, and they make sure to qualify their results with reminders of what a particular benchmark is stressing. Every review is generally conducted with a fresh face, giving everyone who wants to play their game a chance.
Honestly, I've never followed Ars Technica all that much. It always seemed more of a trade magazine to me than a site about hardware, and I've got
Sen. Danforth: "There is nothing on the face of the album which would
notify you if the record has pornographics material or
material glorifying violence?"
Tipper Gore: "No, there is nothing that would suggest that to me."
Frank Zappa: "I would say that a buzz saw blade between the guy's legs on
the album cover is good indication that it's not for little
Johnny."
-- The Senate Commerce Committee hearing on rock
lyrics, from The Village Voice, 6 Oct 1985
Except for a few rare exceptions, politicians and newspersons never provide unbiased information to the public at large. Finding easy answers is too profitable for them to ever stop.
How many times have I heard this? "I bought an ATI card fifty years ago, and the driver updates were slow, and the drivers were bad, and..."
This is an industry where silicon is updated every 6 months. Drivers are released continually. Most importantly, NVidia was hardly a blip on the map when ATI was producing Mach64s in quantity. Now they're one of the top chip suppliers of the PC graphics industry, cutting into ATI's OEM lifeblood.
The arrival of NVidia forced ATI to reexamine their game. They've made genuine improvements in all aspects of their company -- including a vastly improved driver team, and a product cycle competitive with NVidias.
The reality is that, for the time being, ATIs card IS the latest and greatest, (free) driver updates from ATI produce meaningful improvements in quality and performance and, in general, the company has transformed.
ATI is a contender, and ignoring ATI cards as a possible component in your computer based on an outdated, mostly anecdotal experience serves only to hurt yourself in the long run.
long enough to be shoved out an airlock.
Seriously, he was the most irritating part about any tv show ever.
I've seen OpenBSD hotswap identical disks. Once, at a Unix User's Group meeting, we decided to experiment.
The disk was unplugged, the plugged back in. OBSD said it was reinitializing the controller. Poof! back to normal. Still, this probably would not be a good general practice to follow.
re: the Supermicro. It's probably a Serverworks chipset. Make sure the board uses a 33Mhz PCI bus, not 66Mhz.
I question if the card is truly 64 bit considering it works fine in a 32 bit slot.
A RAM-based drive, or ?RAMDrive? would not be persistent without some sort of external power source, and electronics to keep it refreshed.
What you're thinking of is an EEPROM or Flash drive. Looked at the price of a Nomad 2 cartridge lately?
The real knuckle busters are really tight power connectors on drives. You know, the kind that you need a crowbar to remove.
But aren't items 1, 2, 5, 6, and 7 really features of Sun Fire servers, not the OS.
Hey, if I'm wrong, that's great. I'd like dynamic reconfiguration and "easy hot swapping" on my k6/2 machine just by switching to Solaris.
New iBook units are lighter and have a longer battery life, excluding firewire, not to mention cheaper than most Intel machines. IMHO, they're the best product in the Apple line.
I believe the tactic was to blind IR devices on nuclear launch detection satellites and such. The helo pilot probably just got in the way.
As noted above, a 333 MHz front side bus is not really stock nowadays.
/. about practical overclocking and the benefits of water cooling for the mainstream user.
However, if you want to void your warranty, Epox 8h3a's with recent firmware have a 1/5 and a (rumoured) 1/6 pci divider. This means that you can overclock the FSB to 166Mhz (333 DDR) on your motherboard while still maintaining a stock rate for the PCI bus (166/5=33), thereby improving the stability/survival rate of your PCI and AGP components.
With a good AXP 1.6 and the right cooling, 333 should be within reach.
Heh, there should be another article on
The real problem, I think, lies with the inability of business to reconcile itself with an information economy. Groups like the MPAA (and Unisys, and RSA Data Sec.) want to provide information as a product. To do so, they have to control the availability of this 'product' to the end user.
In every other part of the Universe, it is the specific product which is patented. If you make a carbon copy of a Honda Accord, then you'll get sued. If you make a vehicle with four wheels, four doors and an engine, you won't. In academia, if you make a discovery, then some time later another person claims to have made the same discovery, that person -- except in rare cases -- you will be laughed out of town, but reproducing the same result via independent work is okay.
So, reproducing the DeCSS algorithm via independent work is okay via logical extension. As for breaking copyright protection, this is really governed by two laws: the so called "Betamax decision" from which the fair-use concept is derived, and by the DMCA. Although I don't know much about the DMCA, fair-use says that any particular consumer of media content can copy it limitlessly for backup, personal storage, alternate viewing, or whatever. Ripping a DVD to DIVX is perfectly legal, as long as you don't redistribute it and merely use it for personal viewing.