Well, impactfulness is easily measured... It's a simple ratio of the amount of powdered substance in the packet to the total pieces of pasta multiplied by a scaling factor to adjust for the intensity and color of the aforementioned powdered cheese substance. With this simple formula, we see that Kraft Macaroni and Cheese is far more impactful than anything written on Yahoo.
In terms of security, I still wonder why certain aspects of OpenBSD aren't brought in to Linux (TCP/IP especially). I certainly don't want to encourage any more holy wars over licensing (we've had plenty lately), but really - sometimes, you should just adopt 'proven' solutions, and work your way through others.
I love linux, I want to run it full time everywhere, but boy - some of the more advanced multi-proc OS/s (Solaris/UltraSparc, Hp-ux, AIX, OS400) are just so much better at large scale projects...
Well, technically, it's 133.333333333(etc)x7.5 = 1GHz - properclock generators are supposed to be accurate to many decimals - PCI is 33.333333, not 33MHz. It makes a difference... (even if this is a hoax, which would be dissapointing).
Here's a link to a project that already has some function at 2GHz, and is planning to step to 4GHz with some of the new technologies. Cool stuff, and a great group of guys.
This includes some chip shots, as well as descriptions of several of the new technologies that are being used for the project.
Yeah - maybe I can get a Sony AIBO to follow me around and conform to all the various wireless standards - that way, I can just talk to my dog instead of using a phone. Or the dog can carry the phone for me... or maybe (sound of mind snapping)
More than ever, man's best friend (now with less clean up!) 8^)
So is this just a buffer / latency issue where they reduce the total protocol layers between the modem and the (PCI) bus? It seems that you should be able to do some nifty optimizations that you wouldn't necessarily need for browsing/downloading (aside from streaming video), and these would help cut the time. Of course, real modems use less CPU than the winmodems, and that never hurts either, since things have to go much further into the system before they get returned... I always got better d/l speeds with my old USRs than any others at the time - but presently I'm on a cable modem 8^) which gives me an even better ping;o)
Hmm, seems as if my old alpha 533MHz used only something like 90 W (including the MB) - only enough to fry one egg, no bacon 8^) But way hotter than any other chip, that's for sure... too much heat = bad. The multi-proc Alphas are quite toasty (keeps the room^H^H^H^Hhouse warm in winter).
As for performance/MHz, the FPU in the Alpha is generally considered far superior to the Intel FPU at the same clockrate. Plus, my 533 was built in the spring of 1997 - (insert rambling about RISC vs CISC) - (retort about how chips are no longer CISC or RISC)...
Well, I'll have to get rid of any Ultra Sparcs that I run across - silly attempts at an 8-way box that scales better than NT or Linux on an x86. Don't get me wrong, SMP on Linux is getting better, but my personal experience with Solaris on multi-proc Sparc boxes tells a story of a more mature SMP implementation. Solaris x86 isn't nearly as cool overall as it is for the Sparcs, but that should be somewhat expected - more of an afterthought...
The rest of your post reeks too much of moldy trolls and flamebait, so I won't qualify it with much of a response, aside from this: I think Linux is the Greatest Thing(tm) since sliced bread, but I have to admit that it isn't perfect (and certainly far more hackable by default than OpenBSD). Advocate all you want, but the position you are taking sounds pretty uniformed and inflammatory to me...
True, many here are higher level programmers, but those who get interested in kernel development should really understand the lower level architectures (not necessarily down to the gate level) so they can more effeciently program. This is a major concern with device drivers, where you might have to worry about latching data (there, now you've seen it) before you do something else with it, and that can be down at the gate level. Think about firmware - you need an explicit knowledge of what's going on under the covers...
While some may be software "artists" (a well deserved title for some), some of us also happen to be hardware folk (I'm in chip design at the moment, in a large blue building), who love to play with programming, but really love the lower levels. I think that there is a pretty good mix here overall, and like I said, folks worrying about device drivers or certain areas of kernel developement need to have this background.
Computer Organization and Design : The Hardware/Software Interface by David A. Patterson, John L. Hennessy
This book is a well written foray into the internals of processor architecture, memory management structures, etc... one of the best books I've seen on the subject(s). Walks through examples based on a MIPS processor, including pipelining, and really does a good job of making it 'easy' to understand. Kind of expensive for the casual reader 8^) but well worth it for anyone who wants to really learn it.
Yeah - I nearly gave up on it all together when I first tried to port a few net apps from Linux, and I got an error because it didn't have unistd.h... on second thought, I should have left well enough alone.
C++ is (IMHO) a pretty poor vehicle for a lot of everyday programming. I remember showing some of my friends who had to get certain performance specs on their programs that by simply changing a few things from C++ to C calls (cin/cout -> scanf/printf), he nearly tripled the speed of several subroutines, even with -O2 (or equivalent) set for both versions. There wasn't as dramatic a difference in gcc as there was in VC++, but... oh well, what can you do...
My old place of higher learning has converted many of the CS courses to VC++ on 95/8/NT from gcc on SunOS/AIX/System V. Pretty new Dell boxes with "Donated by Intel" stickers all over them insted of RS/6000 workstations. At least they still have few labs with SGI Indys...
The Soul of a New Machine (Tracy Kidder) should be required reading for everyone who really wants to see an example of a how, instead of just a historical view (this was actually required for a course at my college this past year). A great account of Data General's efforts to build, what else, a new machine. Well worth the $11 8^)
I hadn't realized it before either... I'm so shocked. I'm not only going to throw away all my linux cds and diskettes, I'm going to print all of the kernel and X source, so I can burn while waving my big M$ Standards list in the air. Oh, it will be a joyous day when the penguin and the BSDaemon finallly end their foolish attempts to be a real OS. I mean really, I can't even get a nice GPF bluescreen from KDE or GNOME, and I *know* that C/C++ included with M$ VC++ is SO much more standard than any of that gcc ansi crap. Repent all ye CLI users, your day is through. Last but not least.... aiiiieeeee
Only an idiot administrator would expect there to be some magical command to "fix my computer and make it secure" (yes sir, right away, sir!)
How true... installing OpenBSD takes a few commands;-) Seriously, though - a central resource is fairly difficult to maintain, and I think that the efforts at BUGTRAQ have been well above par - I can remember severeal instances where patches came out in record time for security holes, as opposed to the normal M$ - wait for the SP theory.
UNIX people prefer to pay attention and take responsibility for their own security.
Then there's the continuing trend for Windows hot fixes and service packs - they install, without telling you what they do, or offering any version checking - I know a couple of NT admins who have "fixed" security problems by adding an older hotfix that may have kept the one bug they were worried about closed, but re-opened another one that was fixed later. There's a good reason that DLLs, etc. all have verion info. There's something to be said for proper revision control, and doing a little bit to prevent people from backtracking...
>Why AOL would want Excite, which is barely a second-rate portal these days, is beyond me. They already have a >second-rate portal: Netcenter.
COMING SOON!!!! - The AOLNetHomeExciteCenter!!! Links to anywhere and eveywhere we want you to go! Don't be fooled by those other 'portals' The first fully integrated telepathic AIM!!!
Bill G. just purchased the AOLNetHomeExciteCenter, which will now be renamed MSNetHomeAOLExcenteritement. View only with Ayyyyeeee(5) on Windoze 2000^H^H^H^H2001!!!!
Unfortunately, not all of us are in such a nice situation. ADSL in my area ends up being ~$65/mo for 256k access (756k is ~$250/mo), and that's only if you are within a certain distance of the teleco (someone please fill in the details). I know that many people just outside of the city are too far away to get it, or some of the hardware needs to be upgraded between their house and the teleco (not gonna happen -$$).
Cable is far more ubiquitous (and ~$40/mo here), and while @Home clogs or screeches to a halt (the mail server is dead slow regardless of anything else). I would prefer dedicated bandwidth (the technical aspects of DSL are just cooler than cable modem), but in many areas, it just isn't cost effective. Hence the rush for the ever slowing cable modems... though I can usually get a few concurrent 25 - 50k/s downloads.
A mail server that doesn't suck would be great, but at least I have a resonably cheap/fast connection...
95 may run (slightly) quicker on a box with 32MB or RAM, but my experience (on a iP200MMX / 32MB) has shown that 98, NT4 and especially win2000 are *significantly* slower than KDE on that box. Since most people are searching for a *new* desktop, they would probably be using 98 on the machine so they ould play their games / use the new APIs, have the latest, greatest stability that m$ can provide them with. Gnome/E is still really slow... hopefully that can be sped up some.
As for configuring things in the GUI, you couldn't be more right there... but a lot of that has to do with X framework. It sure would be nice to open the Kcontrol Panel and be able to swap resolution / color depths on the fly there, instead of configuring them before hand from a CLI (though with some of the commercial X-servers - Accelerated X in particular - things are vastly nicer than with Xfree...
Yeah - my space Legos have had EMP capabilities since as long as I can remember - I just didn't have the power! My lego army is preparing to take over the world... one square centimeter at a time...
>The NSA'll be wanting to recruit them!! NSA Legos - No Such Android...
Now that there's the Rage 128 support, I'm gpoing to have to get one... I wonder if I can get a MacOS rebate (like the wintel rebate scheme)... or else try to get someone who is building those boards that IBM laid out the spec for.
I've always liked the PowerPC architecture(s), but this actually makes it worth my trouble. Whoo-hoo!
This is a good point - I have obtained much better results from ripping with my plextor SCSI than with any IDE drive out there (and for you windows folks, Plextor provides a nice ripping tool with the drive)... As for soundcards - many times you have more noise intrduced by other components on the motherboard or in the case than by the soundcard itself - hiss and hum in your audio out can sometimes be fixed by relocating harddrives, changing the slot for your sound card, or finding a motherboard with better power regulation (not to mention the power supply in your case!)
Slashdot Mirror
Well, impactfulness is easily measured...
It's a simple ratio of the amount of powdered substance in the packet to the total pieces of pasta multiplied by a scaling factor to adjust for the intensity and color of the aforementioned powdered cheese substance. With this simple formula, we see that Kraft Macaroni and Cheese is far more impactful than anything written on Yahoo.
Or maybe I need to get out more...
In terms of security, I still wonder why certain aspects of OpenBSD aren't brought in to Linux (TCP/IP especially). I certainly don't want to encourage any more holy wars over licensing (we've had plenty lately), but really - sometimes, you should just adopt 'proven' solutions, and work your way through others.
I love linux, I want to run it full time everywhere, but boy - some of the more advanced multi-proc OS/s (Solaris/UltraSparc, Hp-ux, AIX, OS400) are just so much better at large scale projects...
Well, technically, it's 133.333333333(etc)x7.5 = 1GHz - properclock generators are supposed to be accurate to many decimals - PCI is 33.333333, not 33MHz. It makes a difference... (even if this is a hoax, which would be dissapointing).
Here's a link to a project that already has some function at 2GHz, and is planning to step to 4GHz with some of the new technologies. Cool stuff, and a great group of guys.
This includes some chip shots, as well as descriptions of several of the new technologies that are being used for the project.
Yeah - maybe I can get a Sony AIBO to follow me around and conform to all the various wireless standards - that way, I can just talk to my dog instead of using a phone. Or the dog can carry the phone for me... or maybe (sound of mind snapping)
More than ever, man's best friend (now with less clean up!) 8^)
So is this just a buffer / latency issue where they reduce the total protocol layers between the modem and the (PCI) bus? It seems that you should be able to do some nifty optimizations that you wouldn't necessarily need for browsing/downloading (aside from streaming video), and these would help cut the time. Of course, real modems use less CPU than the winmodems, and that never hurts either, since things have to go much further into the system before they get returned... I always got better d/l speeds with my old USRs than any others at the time - but presently I'm on a cable modem 8^) which gives me an even better ping ;o)
Hmm, seems as if my old alpha 533MHz used only something like 90 W (including the MB) - only enough to fry one egg, no bacon 8^) But way hotter than any other chip, that's for sure... too much heat = bad. The multi-proc Alphas are quite toasty (keeps the room^H^H^H^Hhouse warm in winter).
As for performance/MHz, the FPU in the Alpha is generally considered far superior to the Intel FPU at the same clockrate. Plus, my 533 was built in the spring of 1997 - (insert rambling about RISC vs CISC) - (retort about how chips are no longer CISC or RISC)...
just my $.01999999999 (Intel)
$.02 (Alpha)
Well, I'll have to get rid of any Ultra Sparcs that I run across - silly attempts at an 8-way box that scales better than NT or Linux on an x86. Don't get me wrong, SMP on Linux is getting better, but my personal experience with Solaris on multi-proc Sparc boxes tells a story of a more mature SMP implementation. Solaris x86 isn't nearly as cool overall as it is for the Sparcs, but that should be somewhat expected - more of an afterthought...
The rest of your post reeks too much of moldy trolls and flamebait, so I won't qualify it with much of a response, aside from this: I think Linux is the Greatest Thing(tm) since sliced bread, but I have to admit that it isn't perfect (and certainly far more hackable by default than OpenBSD). Advocate all you want, but the position you are taking sounds pretty uniformed and inflammatory to me...
jm$.02
True, many here are higher level programmers, but those who get interested in kernel development should really understand the lower level architectures (not necessarily down to the gate level) so they can more effeciently program. This is a major concern with device drivers, where you might have to worry about latching data (there, now you've seen it) before you do something else with it, and that can be down at the gate level. Think about firmware - you need an explicit knowledge of what's going on under the covers...
While some may be software "artists" (a well deserved title for some), some of us also happen to be hardware folk (I'm in chip design at the moment, in a large blue building), who love to play with programming, but really love the lower levels. I think that there is a pretty good mix here overall, and like I said, folks worrying about device drivers or certain areas of kernel developement need to have this background.
Just my $.02
Computer Organization and Design : The Hardware/Software Interface by David A. Patterson, John L. Hennessy
This book is a well written foray into the internals of processor architecture, memory management structures, etc... one of the best books I've seen on the subject(s). Walks through examples based on a MIPS processor, including pipelining, and really does a good job of making it 'easy' to understand. Kind of expensive for the casual reader 8^) but well worth it for anyone who wants to really learn it.
Yeah - I nearly gave up on it all together when I first tried to port a few net apps from Linux, and I got an error because it didn't have unistd.h ... on second thought, I should have left well enough alone.
C++ is (IMHO) a pretty poor vehicle for a lot of everyday programming. I remember showing some of my friends who had to get certain performance specs on their programs that by simply changing a few things from C++ to C calls (cin/cout -> scanf/printf), he nearly tripled the speed of several subroutines, even with -O2 (or equivalent) set for both versions. There wasn't as dramatic a difference in gcc as there was in VC++, but... oh well, what can you do...
My old place of higher learning has converted many of the CS courses to VC++ on 95/8/NT from gcc on SunOS/AIX/System V. Pretty new Dell boxes with "Donated by Intel" stickers all over them insted of RS/6000 workstations. At least they still have few labs with SGI Indys...
The Soul of a New Machine (Tracy Kidder) should be required reading for everyone who really wants to see an example of a how, instead of just a historical view (this was actually required for a course at my college this past year). A great account of Data General's efforts to build, what else, a new machine. Well worth the $11 8^)
I hadn't realized it before either... I'm so shocked. I'm not only going to throw away all my linux cds and diskettes, I'm going to print all of the kernel and X source, so I can burn while waving my big M$ Standards list in the air. Oh, it will be a joyous day when the penguin and the BSDaemon finallly end their foolish attempts to be a real OS. I mean really, I can't even get a nice GPF bluescreen from KDE or GNOME, and I *know* that C/C++ included with M$ VC++ is SO much more standard than any of that gcc ansi crap. Repent all ye CLI users, your day is through. Last but not least.... aiiiieeeee
Only an idiot administrator would expect there to be some magical command to "fix my computer and make it secure" (yes sir, right away, sir!)
;-) Seriously, though - a central resource is fairly difficult to maintain, and I think that the efforts at BUGTRAQ have been well above par - I can remember severeal instances where patches came out in record time for security holes, as opposed to the normal M$ - wait for the SP theory.
How true... installing OpenBSD takes a few commands
UNIX people prefer to pay attention and take responsibility for their own security.
Then there's the continuing trend for Windows hot fixes and service packs - they install, without telling you what they do, or offering any version checking - I know a couple of NT admins who have "fixed" security problems by adding an older hotfix that may have kept the one bug they were worried about closed, but re-opened another one that was fixed later. There's a good reason that DLLs, etc. all have verion info. There's something to be said for proper revision control, and doing a little bit to prevent people from backtracking...
Just my $.015 (I always come up a little short)
>Why AOL would want Excite, which is barely a second-rate portal these days, is beyond me. They already have a
>second-rate portal: Netcenter.
COMING SOON!!!! - The AOLNetHomeExciteCenter!!!
Links to anywhere and eveywhere we want you to go!
Don't be fooled by those other 'portals'
The first fully integrated telepathic AIM!!!
Bill G. just purchased the AOLNetHomeExciteCenter, which will now be renamed MSNetHomeAOLExcenteritement. View only with Ayyyyeeee(5) on Windoze 2000^H^H^H^H2001!!!!
Where you want to go today...
Unfortunately, not all of us are in such a nice situation. ADSL in my area ends up being ~$65/mo for 256k access (756k is ~$250/mo), and that's only if you are within a certain distance of the teleco (someone please fill in the details). I know that many people just outside of the city are too far away to get it, or some of the hardware needs to be upgraded between their house and the teleco (not gonna happen -$$).
Cable is far more ubiquitous (and ~$40/mo here), and while @Home clogs or screeches to a halt (the mail server is dead slow regardless of anything else). I would prefer dedicated bandwidth (the technical aspects of DSL are just cooler than cable modem), but in many areas, it just isn't cost effective. Hence the rush for the ever slowing cable modems... though I can usually get a few concurrent 25 - 50k/s downloads.
A mail server that doesn't suck would be great, but at least I have a resonably cheap/fast connection...
95 may run (slightly) quicker on a box with 32MB or RAM, but my experience (on a iP200MMX / 32MB) has shown that 98, NT4 and especially win2000 are *significantly* slower than KDE on that box. Since most people are searching for a *new* desktop, they would probably be using 98 on the machine so they ould play their games / use the new APIs, have the latest, greatest stability that m$ can provide them with. Gnome/E is still really slow... hopefully that can be sped up some.
As for configuring things in the GUI, you couldn't be more right there... but a lot of that has to do with X framework. It sure would be nice to open the Kcontrol Panel and be able to swap resolution / color depths on the fly there, instead of configuring them before hand from a CLI (though with some of the commercial X-servers - Accelerated X in particular - things are vastly nicer than with Xfree...
Just my $.02...
Yeah - my space Legos have had EMP capabilities since as long as I can remember - I just didn't have the power! My lego army is preparing to take over the world... one square centimeter at a time...
>The NSA'll be wanting to recruit them!!
NSA Legos - No Such Android...
Now that there's the Rage 128 support, I'm gpoing to have to get one... I wonder if I can get a MacOS rebate (like the wintel rebate scheme)... or else try to get someone who is building those boards that IBM laid out the spec for.
I've always liked the PowerPC architecture(s), but this actually makes it worth my trouble. Whoo-hoo!
Just imagine - you could crack RC5 and capture all of the pokemon at the same time 8^)
This is a good point - I have obtained much better results from ripping with my plextor SCSI than with any IDE drive out there (and for you windows folks, Plextor provides a nice ripping tool with the drive)... As for soundcards - many times you have more noise intrduced by other components on the motherboard or in the case than by the soundcard itself - hiss and hum in your audio out can sometimes be fixed by relocating harddrives, changing the slot for your sound card, or finding a motherboard with better power regulation (not to mention the power supply in your case!)