Part of the reason Rambus memory and the Pentium III processor didn't fare so well was that the Pentium III's front side bus was only 100/133Mhz (800MB/s and 1.06GB/s respectively) and it had to contend with a higher latency memory bus and mismatched speeds. The Pentium 4 resolves the problem by running the FSB to the same speed (but not really the same frequency) as the dual-channel Rambus memory bus.
I'm still holding out on the Pentium 4 solution until Northwood (ie: Pentium 4 in 0.13 micron and new packaging) proves itself and there is enough applications/server services that are optimized for the P4. Unfortunately, I'm stuck with a dual-channel Rambus setup at work:(
So far, I'm relatively impress with the nVidia chipset and hope that it will help increase the acceptance of the AMD Athlon/Palimino processors.
There was also the NEC TurboGraphix Express... I have one sitting in my room collecting dust. It also had a TV Tuner and was a 16-bit with really nice backlit LCD. The thing was fairly heavy, ate 6 batteries (if I can remember correctly) and a relatively good built-in speaker.
I would have been nice if the TubroGraphix system would have had a longer run in the market since it was based on a nice card (and later CD) system, responsive controllers, and a small form factor. Sigh... now I have a craving for Bonk and the original Bomberman...:)\
Correct, the BSD license is not proprietary... but whoever decides to use a piece of code or software that is under the BSD license, they have the choice to keep the source open (either in the sense of for gratis or for a monetary about) or close it off completely. If the software company decides to close off the source, then yes, their modifications would be considered proprietary but at least the original source is still available.
Webmin was released under the BSD license from the start and was then acquired by Caldera. Caldera left the license used by Webmin alone and continues to contribute to the project.
Webmin is included in most of Caldera's ``server'' Linux offerings and can be installed on almost every UNIX derivative.
Most of Webmin's modules are licensed under the GPL rather than BSDL.
The URL is www.webmin.com/webmin and information about Caldera's involvement and the bit about the license is under the Introduction to Webmin page.
Samba includes a web-based administration called SWAT (Samba Web Administration Tool) that can be added to inetd and services. Webmin also has a module for Samba.
p-code... boy, that brings back memories of Visual Basic 4.0 and 5.0:)
If I wanted to a cross-platform soluton, I'm still sticking with Python (interpreted kind of like p-code but bytecode instead like Java)... it's nice and not nearly as messy as Java.
I think.NET in general is more hype than substance, but that's just me... hehe
C# is the Microsoft.NET'd version of C/C++/Java/et al. Visual Studio.NET is still in beta and there probably won't be another beta release until sometime in the summer. I'm not sure when the final release is supposed to be out.
I would personally program in Python than in C#, mostly when Active State is working on a Python.NET model/compiler/etc. for Visual Studio.NET (or 7.0).
Re:Somehow this isn't much of a surprise....
on
Gamecube In Danger?
·
· Score: 1
The quality of the Sega Dreamcast is very good (the graphics capability is really nice, mostly when coupled with a computer monitor... or even a high-quality LCD monitor)... what ``killed'' the Dreamcast was the Playstation (and in part, the N64).
Sony has a very, very good line-up of game software companies (EA, Square, Capcom, etc.) where Sega couldn't get into the sights of those companies (other than Capcom, but that's because Capcom is willing to spread themselves across several console platforms).
It is really nice to see that Samba is getting a lot closer to a real Windows NT server! I have been using Samba 2.0.7 where I work and it has worked very well (although we don't have a lot of people using it, but it works).
DFS is a really nice feature to have, since a sysadmin can create a single SMB share that links to all of the other network shares... less to remember and less support calls like, ``where's such-and-such folder again?''.
Also being able to edit the ACL directly from a Windows NT/2000 workstation is nice... probably won't have to do a lot of chown/chmod-ing again:)
The the concern is that as Intel increases the speed on the Pentium 4 processor, the thermal diode could then be tripped off sooner and diminish the performance quite a bit more than the slower variants of the Pentium 4. Or is Intel actually increasing the thermal threshold as they up the processor speed?
It would be pointless to pay a lot more for a 1.7Ghz processor that would throttle to 775Mhz rather than for a 1.5Ghz that would throttle to 750Mhz. I would rather have Intel would provide a switch to turn off/on the feature in the BIOS than force the feature to be on all of the time.
Intel is usually known to play around to make the power consumption of their chips to look really good. One instance is the mobile Pentium III processor with SpeedStep; there have been several claims that Intel intentionally reduces the voltage down further (using their best samples to do so) and only run the most simple apps... then when idle, Intel takes a snapshot of the approximate power consumption.
In reality, it's very hard to believe that a 1.5Ghz processor with a die that is almost twice the size of the Athlon T-bird can have a lot less power consumption than a 1.2Ghz T-bird. Intel may have a more refined process, but the numbers still don't add up.
Partly the reason why the higher memory bandwidth overhead is that it uses the dual-channel Rambus memory subsystem. I could be wrong, but isn't the Rambus memory architecture packet-based? As with TCP/IP packets, there is overhead that cannot be avoided and could be higher than SDRAM's overhead. The [possible] overhead would in turn increase the latency memory access even more.
Belkin [www.belkin.com] and Orange Micro [www.orangemicro.com] have announced and are possibly shipping USB 2.0 devices. I think the USB 2.0 cards use OHCI instead of UHCI (which is the interface style used by the on-board USB ports found on most of the newer chipsets).
I think the next revision of the IEEE 1394 (IEEE 1394b I think... The Register has some articles on Microsoft, 1394b and CPRM) which is supposed to be faster than the current version of 1394, but includes CPRM and some other additions that we may not know about.
I think it's the other way around... FireWire is Apple's version of IEEE 1394 (the actual standard). That is why hardware manufacturers have to pay Apple for each FireWire port that is on a device; be it a FireWire PCI card, a FireWire hard drive, etc.
Sony's i.Link is their version of the IEEE 1394 standard. I'm not sure how or if licensing is paid for i.Link.
Think of it as the W3C's HTML standard... each vendor takes that standard and tries to implement it so that it tries to meet the standards.
I personally prefer FireWire over USB for high speed devices, but for keyboards, mice, PDA cradles and printers... USB 1.x is definitely good. It gives me a chance to disable the PS/2, parallel and serial ports on my machine:)
Via has also announced and has samples of their new ITX standard, which competes with the Flex ATX standard, but I think will also fit into a MicroATX case. I think the sample Via had included an on-board S3 video chipset (since the two are working together).
NFS mounting shares would probably be the best way to do it, unless if you want to store static stuff onto a CD (but I'm not sure how well the CD-ROM drive will fair.. hehe).
I wonder if there is anyway to hack the Dreamcast to upgrade the memory (like some of the guys did with the HP 48G/GX calculators to double the memory... drool:)
For a simple web server, you can use tHTTPD or maybe some tight C code that can call to a backend server to generate the HTML. Dunno.
It would be interesting to see someone or some company use the Dreamcast as web servers behind a load balancer considering that the Dreamcast is fairly inexpensive compared to the cheapest rackmount servers out there.
Another interesting function that the Dreamcast servers could provide would be a mini-gaming server. It might work as a MUD server or as a simple backend server for web-based games, etc.
But wouldn't that exclude a home PC? Maybe I'm wrong, but I've always thought of a workstation as a client on a network that connects to a server to run apps and do additional processing that isn't done at the server.
Although marketing via hype tends to be effective at the beginning, but quite a few people get jaded and sick by the hype. I, for one, got sick of the Linux,.com, XML and now the B2B hype quite quickly. Why? It's marketing fluff and that's what it will be unless people take it seriously enough so that the product can and will live without the media hype.
My question is how does one define a desktop operating system? Is a desktop operating system one that is used at home by Joe User? Or is a desktop operating system one that is installed on any machine that sits on or below a desk?
A server operating system is fairly easily to understand... it's an operating system that is built to serve services (be it files, web pages, MP3s, databases, printers, mail, etc.) and has multiple users connected at once via terminals, desktops and/or workstations.
So what do you consider to be the merits of a desktop OS?
FreeBSD can be both a desktop and a server operating system, mostly since FreeBSD is the only BSD OS (available for free, as in beer and as in open source) that supports SMP. OpenBSD would be a great server operating system since it's security audits on the core code and it's ``secure by default'' legend. And because of that, OpenBSD could also be a network appliance OS since it would make a perfect firewall/router device for any network (okay, maybe not for an ISP or a conglomerate).
But I wouldn't just go silly and give only one label to each BSD OS.
Slashdot Mirror
Part of the reason Rambus memory and the Pentium III processor didn't fare so well was that the Pentium III's front side bus was only 100/133Mhz (800MB/s and 1.06GB/s respectively) and it had to contend with a higher latency memory bus and mismatched speeds. The Pentium 4 resolves the problem by running the FSB to the same speed (but not really the same frequency) as the dual-channel Rambus memory bus.
:(
I'm still holding out on the Pentium 4 solution until Northwood (ie: Pentium 4 in 0.13 micron and new packaging) proves itself and there is enough applications/server services that are optimized for the P4. Unfortunately, I'm stuck with a dual-channel Rambus setup at work
So far, I'm relatively impress with the nVidia chipset and hope that it will help increase the acceptance of the AMD Athlon/Palimino processors.
There was also the NEC TurboGraphix Express... I have one sitting in my room collecting dust. It also had a TV Tuner and was a 16-bit with really nice backlit LCD. The thing was fairly heavy, ate 6 batteries (if I can remember correctly) and a relatively good built-in speaker.
:)\
I would have been nice if the TubroGraphix system would have had a longer run in the market since it was based on a nice card (and later CD) system, responsive controllers, and a small form factor. Sigh... now I have a craving for Bonk and the original Bomberman...
Correct, the BSD license is not proprietary... but whoever decides to use a piece of code or software that is under the BSD license, they have the choice to keep the source open (either in the sense of for gratis or for a monetary about) or close it off completely. If the software company decides to close off the source, then yes, their modifications would be considered proprietary but at least the original source is still available.
Webmin was released under the BSD license from the start and was then acquired by Caldera. Caldera left the license used by Webmin alone and continues to contribute to the project.
Webmin is included in most of Caldera's ``server'' Linux offerings and can be installed on almost every UNIX derivative.
Most of Webmin's modules are licensed under the GPL rather than BSDL.
The URL is www.webmin.com/webmin and information about Caldera's involvement and the bit about the license is under the Introduction to Webmin page.
Samba includes a web-based administration called SWAT (Samba Web Administration Tool) that can be added to inetd and services. Webmin also has a module for Samba.
p-code... boy, that brings back memories of Visual Basic 4.0 and 5.0 :)
.NET in general is more hype than substance, but that's just me... hehe
If I wanted to a cross-platform soluton, I'm still sticking with Python (interpreted kind of like p-code but bytecode instead like Java)... it's nice and not nearly as messy as Java.
I think
I think the color scheme for the programming section is blue, where as the Apache section is purple and funny yellow/green.
C# is the Microsoft .NET'd version of C/C++/Java/et al. Visual Studio .NET is still in beta and there probably won't be another beta release until sometime in the summer. I'm not sure when the final release is supposed to be out.
.NET (or 7.0).
I would personally program in Python than in C#, mostly when Active State is working on a Python.NET model/compiler/etc. for Visual Studio
The quality of the Sega Dreamcast is very good (the graphics capability is really nice, mostly when coupled with a computer monitor... or even a high-quality LCD monitor)... what ``killed'' the Dreamcast was the Playstation (and in part, the N64).
Sony has a very, very good line-up of game software companies (EA, Square, Capcom, etc.) where Sega couldn't get into the sights of those companies (other than Capcom, but that's because Capcom is willing to spread themselves across several console platforms).
It is really nice to see that Samba is getting a lot closer to a real Windows NT server! I have been using Samba 2.0.7 where I work and it has worked very well (although we don't have a lot of people using it, but it works).
:)
DFS is a really nice feature to have, since a sysadmin can create a single SMB share that links to all of the other network shares... less to remember and less support calls like, ``where's such-and-such folder again?''.
Also being able to edit the ACL directly from a Windows NT/2000 workstation is nice... probably won't have to do a lot of chown/chmod-ing again
Does anyone know the speed of the unladen African Python?
The sorta same thing happened with the 386SX/DX and the 486SX/DX bit.
The the concern is that as Intel increases the speed on the Pentium 4 processor, the thermal diode could then be tripped off sooner and diminish the performance quite a bit more than the slower variants of the Pentium 4. Or is Intel actually increasing the thermal threshold as they up the processor speed?
It would be pointless to pay a lot more for a 1.7Ghz processor that would throttle to 775Mhz rather than for a 1.5Ghz that would throttle to 750Mhz. I would rather have Intel would provide a switch to turn off/on the feature in the BIOS than force the feature to be on all of the time.
Intel is usually known to play around to make the power consumption of their chips to look really good. One instance is the mobile Pentium III processor with SpeedStep; there have been several claims that Intel intentionally reduces the voltage down further (using their best samples to do so) and only run the most simple apps... then when idle, Intel takes a snapshot of the approximate power consumption.
In reality, it's very hard to believe that a 1.5Ghz processor with a die that is almost twice the size of the Athlon T-bird can have a lot less power consumption than a 1.2Ghz T-bird. Intel may have a more refined process, but the numbers still don't add up.
Partly the reason why the higher memory bandwidth overhead is that it uses the dual-channel Rambus memory subsystem. I could be wrong, but isn't the Rambus memory architecture packet-based? As with TCP/IP packets, there is overhead that cannot be avoided and could be higher than SDRAM's overhead. The [possible] overhead would in turn increase the latency memory access even more.
Belkin [www.belkin.com] and Orange Micro [www.orangemicro.com] have announced and are possibly shipping USB 2.0 devices. I think the USB 2.0 cards use OHCI instead of UHCI (which is the interface style used by the on-board USB ports found on most of the newer chipsets).
I think the next revision of the IEEE 1394 (IEEE 1394b I think... The Register has some articles on Microsoft, 1394b and CPRM) which is supposed to be faster than the current version of 1394, but includes CPRM and some other additions that we may not know about.
I think it's the other way around... FireWire is Apple's version of IEEE 1394 (the actual standard). That is why hardware manufacturers have to pay Apple for each FireWire port that is on a device; be it a FireWire PCI card, a FireWire hard drive, etc.
:)
Sony's i.Link is their version of the IEEE 1394 standard. I'm not sure how or if licensing is paid for i.Link.
Think of it as the W3C's HTML standard... each vendor takes that standard and tries to implement it so that it tries to meet the standards.
I personally prefer FireWire over USB for high speed devices, but for keyboards, mice, PDA cradles and printers... USB 1.x is definitely good. It gives me a chance to disable the PS/2, parallel and serial ports on my machine
Via has also announced and has samples of their new ITX standard, which competes with the Flex ATX standard, but I think will also fit into a MicroATX case. I think the sample Via had included an on-board S3 video chipset (since the two are working together).
NFS mounting shares would probably be the best way to do it, unless if you want to store static stuff onto a CD (but I'm not sure how well the CD-ROM drive will fair.. hehe).
:)
I wonder if there is anyway to hack the Dreamcast to upgrade the memory (like some of the guys did with the HP 48G/GX calculators to double the memory... drool
For a simple web server, you can use tHTTPD or maybe some tight C code that can call to a backend server to generate the HTML. Dunno.
It would be interesting to see someone or some company use the Dreamcast as web servers behind a load balancer considering that the Dreamcast is fairly inexpensive compared to the cheapest rackmount servers out there.
Another interesting function that the Dreamcast servers could provide would be a mini-gaming server. It might work as a MUD server or as a simple backend server for web-based games, etc.
Sounds good to me :)
But wouldn't that exclude a home PC? Maybe I'm wrong, but I've always thought of a workstation as a client on a network that connects to a server to run apps and do additional processing that isn't done at the server.
Although marketing via hype tends to be effective at the beginning, but quite a few people get jaded and sick by the hype. I, for one, got sick of the Linux, .com, XML and now the B2B hype quite quickly. Why? It's marketing fluff and that's what it will be unless people take it seriously enough so that the product can and will live without the media hype.
Oh well...
My question is how does one define a desktop operating system? Is a desktop operating system one that is used at home by Joe User? Or is a desktop operating system one that is installed on any machine that sits on or below a desk?
A server operating system is fairly easily to understand... it's an operating system that is built to serve services (be it files, web pages, MP3s, databases, printers, mail, etc.) and has multiple users connected at once via terminals, desktops and/or workstations.
So what do you consider to be the merits of a desktop OS?
FreeBSD can be both a desktop and a server operating system, mostly since FreeBSD is the only BSD OS (available for free, as in beer and as in open source) that supports SMP. OpenBSD would be a great server operating system since it's security audits on the core code and it's ``secure by default'' legend. And because of that, OpenBSD could also be a network appliance OS since it would make a perfect firewall/router device for any network (okay, maybe not for an ISP or a conglomerate).
But I wouldn't just go silly and give only one label to each BSD OS.