This seems like major horse poo, there simply is not reasonable general purpose architecture available for 128bit and I even fail to see a requirement for the far future. Sooner we will see shard architectures based on massive parallel multiprocessing like today GPUs but they will be per core a lot simpler than today 64bit systems, not more complex.
I have never ever thought about buying a low budget indy game.
First, there are thousands of free and nice games out there. Free as Free Beer and also some as Free Speech. I simply do not look for another Tetris-Clone or Lemming Clone or whatever lame remake of a 80 arcade game, they are available for free in thousand versions.
Second, I can get full blown studio games for one to three bucks.
I bought Pirates! two years ago for two bucks, a surprisingly complete Tycoon-Collection last year for seven bucks, Battlefield2+AllExpansions for seven Bucks just a week ago.
There is no room for DRM-shit, it is really that easy.
They are free like in free beer, have no limits, no binaries, just create an account on arcor.de and use the account and password to join their nntp-server. They are professional, doing it for over ten years, their servers are powerfull, what else do you want?
I build a faster computer which only costed a small percentage of that Opteron-Beast:
An Intel5400-Board with two 4Ghz QuadXeons and two GT280. Ok, it only has 16GB and four drives but this would be easily corrected nowadays.
I personally use an Intel5000XVN-Board with two 2,5Ghz QuadXeons, 16GB RAM, two 500 Harddrives and an Geforce 8800, waiting for the GT2xx-Line to become mature and passive cooled ---- BECAUSE MY WHOLE SYSTEM IS PASSIVELLY COOLED BESIDES ONE SINGLE 40CM FAN RUNNING AT 50RPM.
The whole System did cost a lot less than â2000 and consists of standard hardware allover.
Yeah I still remember PCS but I played a lot more "Racing Destruction Set" which is still a very formidable game though it has its lengths.
Also I remember "Lode Runner" and "Mister Robot and his Facroty" and lets not forget "Seven Cities of Gold" (though you had little control over your random world there were editors available for fine manipuation) and others...
It is a series about a bunch of spaceships full of retarded idiots, so stubborn, unpleasant, unfocused, no planning, blind and arrogant that I would join the cylons in the instance to whipe out that insult to intelligent life.
The old Galactica had its moments, was fancy and had surprises, but Galactica NT is just a remake of a brasilian telenovela which added spaceships.
The fusion cells are here, the definite answer to asymetric terrorism, the "blitzkrieg" of the 21st century.
Its been all over the net the last year (militaryphoto, strategypage, longwarjournal, sicherheitspolitik and others) and centers around a new geek approach about hunting the bad guys down: Small teams with lots of freedom to move and as many toys to play with as they like. And also more secrecy than anything ever before. Think of "Mission Impossible", the classic series from the seventies, just with hundreds of teams in operation.
I am pretty surprised that the geek world hasn't had fusion cells in their focus earlier, it is the ultimate geek military unit.
There are quite some variants of AmigaOS alive and kicking today (AROS, AmigaOS4, Syllabel etcpp) and they usually have nuch higher bounties for ported software. And have ported a lot more than the OS/2 community too!
In the early 90th my team thought about realizing a multi core design with 16 to 64 6502 cores on a chip with the same complexity as a 68020. Some rough estimates showed that this system could outperform an equal complex 68020 system by ten times. Later this drifted towards using an AMD29000 core but it all failes as all ready designs weren't available for licensing back then.
Back to topic, a 6502 has around 4 000 micro elements, compared to a GT-280 with 1 400 000 000 elements. Given that a nowadays 6502 would include some additional circuits this would mean we could include 1400000000/5000=280 000 cores running at 1,2Ghz each, resulting in 150-200 Tera Integer Operations per second. Given that many long floating point operations can be reprogrammed as short integer operations with less than 100 cycles per op this would outperform every recent solution.
> The remarkable thing is the bargain they give on > the base MacPro system - last time I priced > equivalent 8 core Linux boxes, you'd have to pay > 25% more to get equivalent hardware with no OS > installed.
Where do you shop? I can get an eight core linux machine for half the price of a decent eight core mac.
Look here, 8 cores, 8gb ram, raid1 hd, highend opengl, running absolutely silent and a cool case for just 1490Euro including 19% vat (that is, without vat, 1250 Euro)...http://geizhals.at/deutschland/?cat=WL-3811
Ok, you have the right to buy an overhyped, gaudy mac running a dull unix with some standard unix applets and pay ten times more than for a standard linux system running exactly the same applications.
And I have the right to call you n00bish. But well, maybe your ego just needs some external extension to your self respect;-)
Ok, you have the right to buy an overhyped, gaudy mac running a dull unix with some standard unix applets and pay ten times more than for a standard linux system running exactly the same applications.
And I have the right to call you n00bish. But well, maybe your ego just needs some external extension to your self respect;-)
This is just another case where Moores Law bypasses another Law, in this case the Micromechanical Law.
Microchips are developing faster than Micromechanic so by leaving behind harddrives we might see flashbased storage become bigger and bigger.
But even more interesting, I think we will see SATA-attached flash being a short lived exception. It is much cheaper, faster, more flexible and more direct to attach it directly into address space (doesn't matter if NAND or NOR, the memory controller will take care of that). With that done we will also see DRAM becoming some sort of cache for flash and flash replacing DRAM. And as memory and inner circuit bandwidth will increase even further there really is no reason to safe bandwidth or storage any more - who cares if your 100GB database uses 200GB because you deflate/multi-index it for faster access times? The only limit here is that 64bit address space may be a little narrow but then we just see 128bit architectures ten years earlier.
Ah yes, the future of SQL... well, I think a small SQL-to-DBM interface should do the trick until all applications map the whole db into real address space. There is little use for a overcomplicated interpreting language if you can just put your whole 200GB db into an associative array.
Today you can buy a machine with eight cores, 8gb memory and 1tb harddrive for less than 2000. And most software will only use one core and a maximum of 2gb memory.
WE NEED MULTITHREADING NOW BIG AND EVERYWHERE.
Multithreading is maybe the biggest change in software development. In contrast to advanced command sets like MMX, SSE and so on it is not about some peep hole optimization, about replacing a bunch of x86_32 commands with some SSE commands, it is about changing the whole approach, finding new algorithms and redevelop much if not all software we are used to work with.
And we are lightyears away from using appropiate software on on todays systems.
See compressors: Most can't multithread at all. Those who can have more issues than the SCO buisness plan, eg reduced efficency, scale very bad, can't accept input from stdio or can't output on stdout. Basically compression should be a good starting point but even here todays solutions are incredible far behind the hardware.
Also consider network communication, graphical interfaces, games, printing, non-enterprise realtime presentation and so on and so forth.
The revolution is NOT here. People aren't even talking about it.
Thats exactly whats BSD made for. Get everything you like and then throw back into the developers face everything you hate. No need to say thanks. Apple did it, Linux did it, dozends of others do it all day.
Seriously, bragging about this is a sign of total ignorance about the BSD philosophy: Giving away everything without asking for anything. They should feel honored that they are getting ripped like they wanted always to be.
You simply can't install modern distributions on the first system. And if it would be possible it would look totally stupid. You can not even install a textmode-only system with anything else than Slackware (very slow) and Basic Linux (very basic).
The second system can be made to run Ubuntu but sucks increadibly big time - it takes 8 minutes to boot and then crawls to no end. Same for Suse 9.1 but even slower.
Enter Basic-Linux:
Boots on both systems in 20 seconds. Installation is a piece of cake. On the amiga I had to recompile a lot of stuff but then works fine too.
Slackware: Painfull to install but at least usable on system 2. Not for Amiga.
NetBSD: Installs and runs acceptable on all systems.
Debian: Painfull to install and painfull to use except on system 1). Debian is much to fat for old systems.
My experience tells me that there are several caveats:
1. PnP - modern distributions expect PnP, otherwise they don't run at all. Be ready to use isapnp and/or/proc/pnp to do this by hand and or edit complicated options for kernel and modules.
2. booting from CD - modern distributions expect that they are booted from CD. Sorry but a 15 years old 486 doesn't even have a CD. There are no disks for Ubuntu and in most other cases the disks don't contain all modules needed especially some weird old 486-scsi-vlb-drivers are missing.
3. Giant RAM-disks. All systems use very large RAM-disks for installation except Basic Linux. How do I load a 24MB RAM-Disk on a 16MB Amiga? Or a 48MB-RAM-Disk on a 4MB 486?
Ah, btw, Windows95 installed without much hazzles on the 486-systems and runs quite well.
brandtc@sword:~$ apt-cache search rexx crashmail - JAM and *.MSG capable Fidonet tosser exuberant-ctags - multi-language reimplementation of ctags regina-rexx - The Regina REXX interpreter. regina2 - The Regina REXX interpreter, run-time library. regina2-dev - The Regina REXX interpreter, development files. rexxtk - Interface to Tcl/Tk for Regina REXX rxsock - Socket function library for Regina REXX
And btw, REXX was best integrated into AmigaOS and not into OS/2. You could for example syncronize Processes, use it for interprocess-communication and the best of all, it was the absolute standard for everything. Nearly all Software features a REXX-port which made the Amiga the ultimate OS for "use several small and nifty pieces to make something really great".
In 26 years of computer gaming only Transport-Tycoon was able to survive ten years of attention and still rock.
Ok, today I am playing the Open-Source-Clone OpenTTD but it is still the same gameplay, just with more depth in resources.
The original TT was rather blunt (needed a 386dx40) but already TT Deluxe was pretty deep in gameplay (fun at 486dx2-66). The Windows-Version TTDW wasn't a big step but is nowadays the base of all further versions (starting at a 486dx4-133).
There is TTDW-Patch which adds hundreds of features to the old original game by doing binary patches at runtime, especially likeable are the very advanced train-signals. I suggest a Pentium-200.
OpenTTD is a clean-room-reimplementation of TTDW+TTDWPatch with source-code and a working internet-multiplayer-system. Needs a pretty tough machine though, at least a PentiumIII-400 but even my Athlon-2100+ has a hard time when playing at 1600x1200 with some thousand vehicles on screen.
My Tapes never survive more than 2-3 years but most if not all my CDRs from ten years ago are still fine and so are my CDRs from five years and my few DVDR from two years ago.
But I have something which surrelly outlives them all and is much cheaper also:
A RAID5.
I have a RAID5 from 1997 around, 4x4GB SCSI. It was switched off most time to reduce stress on the parts but I expect it work for many, many more years. And if one drive fails... hey, its a RAID5.
Then I have another RAID5 made from 5x9GB from 2001. Still working like a charm. I have left them inside the computer from that age so I just have to power up and telnet to the box to access my backup.
And my nowadays RAID5, 4x160GB from 2003, soon going to be update with a "RAID6" with 6x250GB. The old 4x160GB-drives will be put either inside a USB-case or into another computer for easy powerup+telnet-access.
Also I think ethernet and telnet/ssh will be the best option for a longliving standard.
One of the major drawbacks of OpenSSH was the lack of a per-account/per-key based traffic-accounting. I always had the impression that the developer of OpenSSH opposed the basic idea of getting precise data about how much everyone did even if it still is possible for the admin to track everything from outside SSH.
-The GUIs are much too overloaded already. Every button which can be removed should be removed.
-The GUI is not following the Unix-mantra of "everything is a file". Too much stuff in deeply nested menues instead of properties of a file (or an icon). Even Gnome/Ubuntu are not able to place a Dial-Up-Icon on the Desktop.
-Too many distributions of files and drivers. When I last looked for a nv-driver it came in as many as 50 instances: four tar.gz, five debs and a whooooping 78 rpms (starting with suse 7.3 and ending with fedora 10.whatever)
!!!Thats insane!!!
At least within ONE distribution major release I expect only ONE driver (say one RPM for SuSE7, one for Debian3).
Slashdot Mirror
This seems like major horse poo, there simply is not reasonable general purpose architecture available for 128bit and I even fail to see a requirement for the far future. Sooner we will see shard architectures based on massive parallel multiprocessing like today GPUs but they will be per core a lot simpler than today 64bit systems, not more complex.
This stinks. A critical part of windows becoming unsupported.
What goes around comes around, I will remember that.
I have never ever thought about buying a low budget indy game.
First, there are thousands of free and nice games out there. Free as Free Beer and also some as Free Speech. I simply do not look for another Tetris-Clone or Lemming Clone or whatever lame remake of a 80 arcade game, they are available for free in thousand versions.
Second, I can get full blown studio games for one to three bucks.
I bought Pirates! two years ago for two bucks, a surprisingly complete Tycoon-Collection last year for seven bucks, Battlefield2+AllExpansions for seven Bucks just a week ago.
There is no room for DRM-shit, it is really that easy.
I have been using arcor.de for decades.
They are free like in free beer, have no limits, no binaries, just create an account on arcor.de and use the account and password to join their nntp-server. They are professional, doing it for over ten years, their servers are powerfull, what else do you want?
Besides you'll find more free providers at http://www.google.de/search?q=free+nntp
I build a faster computer which only costed a small percentage of that Opteron-Beast:
An Intel5400-Board with two 4Ghz QuadXeons and two GT280. Ok, it only has 16GB and four drives but this would be easily corrected nowadays.
I personally use an Intel5000XVN-Board with two 2,5Ghz QuadXeons, 16GB RAM, two 500 Harddrives and an Geforce 8800, waiting for the GT2xx-Line to become mature and passive cooled ---- BECAUSE MY WHOLE SYSTEM IS PASSIVELLY COOLED BESIDES ONE SINGLE 40CM FAN RUNNING AT 50RPM.
The whole System did cost a lot less than â2000 and consists of standard hardware allover.
Yeah I still remember PCS but I played a lot more "Racing Destruction Set" which is still a very formidable game though it has its lengths.
Also I remember "Lode Runner" and "Mister Robot and his Facroty" and lets not forget "Seven Cities of Gold" (though you had little control over your random world there were editors available for fine manipuation) and others...
What are people liking about Galactica NT?
It is a series about a bunch of spaceships full of retarded idiots, so stubborn, unpleasant, unfocused, no planning, blind and arrogant that I would join the cylons in the instance to whipe out that insult to intelligent life.
The old Galactica had its moments, was fancy and had surprises, but Galactica NT is just a remake of a brasilian telenovela which added spaceships.
Oh, Slashdot is last on the news...
The fusion cells are here, the definite answer to asymetric terrorism, the "blitzkrieg" of the 21st century.
Its been all over the net the last year (militaryphoto, strategypage, longwarjournal, sicherheitspolitik and others) and centers around a new geek approach about hunting the bad guys down: Small teams with lots of freedom to move and as many toys to play with as they like. And also more secrecy than anything ever before. Think of "Mission Impossible", the classic series from the seventies, just with hundreds of teams in operation.
I am pretty surprised that the geek world hasn't had fusion cells in their focus earlier, it is the ultimate geek military unit.
There are quite some variants of AmigaOS alive and kicking today (AROS, AmigaOS4, Syllabel etcpp) and they usually have nuch higher bounties for ported software. And have ported a lot more than the OS/2 community too!
> I've sent several donations already this year,
> but I won't be sending more.
Proof or STFU
In the early 90th my team thought about realizing a multi core design with 16 to 64 6502 cores on a chip with the same complexity as a 68020. Some rough estimates showed that this system could outperform an equal complex 68020 system by ten times. Later this drifted towards using an AMD29000 core but it all failes as all ready designs weren't available for licensing back then.
Back to topic, a 6502 has around 4 000 micro elements, compared to a GT-280 with 1 400 000 000 elements. Given that a nowadays 6502 would include some additional circuits this would mean we could include 1400000000/5000=280 000 cores running at 1,2Ghz each, resulting in 150-200 Tera Integer Operations per second. Given that many long floating point operations can be reprogrammed as short integer operations with less than 100 cycles per op this would outperform every recent solution.
> The remarkable thing is the bargain they give on
> the base MacPro system - last time I priced
> equivalent 8 core Linux boxes, you'd have to pay
> 25% more to get equivalent hardware with no OS
> installed.
Where do you shop? I can get an eight core linux machine for half the price of a decent eight core mac.
Look here, 8 cores, 8gb ram, raid1 hd, highend opengl, running absolutely silent and a cool case for just 1490Euro including 19% vat (that is, without vat, 1250 Euro) ...http://geizhals.at/deutschland/?cat=WL-3811
Ok, you have the right to buy an overhyped, gaudy mac running a dull unix with some standard unix applets and pay ten times more than for a standard linux system running exactly the same applications.
And I have the right to call you n00bish. But well, maybe your ego just needs some external extension to your self respect ;-)
Ok, you have the right to buy an overhyped, gaudy mac running a dull unix with some standard unix applets and pay ten times more than for a standard linux system running exactly the same applications.
And I have the right to call you n00bish. But well, maybe your ego just needs some external extension to your self respect ;-)
It is very easy to go with very little fuel:
Get a fuel efficient motor bike and drive very carefully.
You can easily drive 100km with a lot less than one litre.
This is just another case where Moores Law bypasses another Law, in this case the Micromechanical Law.
Microchips are developing faster than Micromechanic so by leaving behind harddrives we might see flashbased storage become bigger and bigger.
But even more interesting, I think we will see SATA-attached flash being a short lived exception. It is much cheaper, faster, more flexible and more direct to attach it directly into address space (doesn't matter if NAND or NOR, the memory controller will take care of that). With that done we will also see DRAM becoming some sort of cache for flash and flash replacing DRAM. And as memory and inner circuit bandwidth will increase even further there really is no reason to safe bandwidth or storage any more - who cares if your 100GB database uses 200GB because you deflate/multi-index it for faster access times? The only limit here is that 64bit address space may be a little narrow but then we just see 128bit architectures ten years earlier.
Ah yes, the future of SQL... well, I think a small SQL-to-DBM interface should do the trick until all applications map the whole db into real address space. There is little use for a overcomplicated interpreting language if you can just put your whole 200GB db into an associative array.
Today you can buy a machine with eight cores, 8gb memory and 1tb harddrive for less than 2000. And most software will only use one core and a maximum of 2gb memory.
WE NEED MULTITHREADING NOW BIG AND EVERYWHERE.
Multithreading is maybe the biggest change in software development. In contrast to advanced command sets like MMX, SSE and so on it is not about some peep hole optimization, about replacing a bunch of x86_32 commands with some SSE commands, it is about changing the whole approach, finding new algorithms and redevelop much if not all software we are used to work with.
And we are lightyears away from using appropiate software on on todays systems.
See compressors: Most can't multithread at all. Those who can have more issues than the SCO buisness plan, eg reduced efficency, scale very bad, can't accept input from stdio or can't output on stdout. Basically compression should be a good starting point but even here todays solutions are incredible far behind the hardware.
Also consider network communication, graphical interfaces, games, printing, non-enterprise realtime presentation and so on and so forth.
The revolution is NOT here. People aren't even talking about it.
Thats exactly whats BSD made for. Get everything you like and then throw back into the developers face everything you hate. No need to say thanks. Apple did it, Linux did it, dozends of others do it all day.
Seriously, bragging about this is a sign of total ignorance about the BSD philosophy: Giving away everything without asking for anything. They should feel honored that they are getting ripped like they wanted always to be.
can't they make up a new and original name?
Sorry to say but Microsoft ist right if we allow a strange comparison, that is, comparing today Linux-Distributions to older Window-Versions.
/proc/pnp to do this by hand and or edit complicated options for kernel and modules.
My systems, all without APM/PCI/PnP:
1. 486dx1-33Mhz/4MB-RAM/120MB-IDE
2. 486dx4-160Mhz/64MB-RAM/2GB-IDE
3. Amiga3000/030-25Mhz/16MB-RAM/2GB-SCSI
You simply can't install modern distributions on the first system. And if it would be possible it would look totally stupid. You can not even install a textmode-only system with anything else than Slackware (very slow) and Basic Linux (very basic).
The second system can be made to run Ubuntu but sucks increadibly big time - it takes 8 minutes to boot and then crawls to no end. Same for Suse 9.1 but even slower.
Enter Basic-Linux:
Boots on both systems in 20 seconds. Installation is a piece of cake. On the amiga I had to recompile a lot of stuff but then works fine too.
Slackware: Painfull to install but at least usable on system 2. Not for Amiga.
NetBSD: Installs and runs acceptable on all systems.
Debian: Painfull to install and painfull to use except on system 1). Debian is much to fat for old systems.
My experience tells me that there are several caveats:
1. PnP - modern distributions expect PnP, otherwise they don't run at all. Be ready to use isapnp and/or
2. booting from CD - modern distributions expect that they are booted from CD. Sorry but a 15 years old 486 doesn't even have a CD. There are no disks for Ubuntu and in most other cases the disks don't contain all modules needed especially some weird old 486-scsi-vlb-drivers are missing.
3. Giant RAM-disks. All systems use very large RAM-disks for installation except Basic Linux. How do I load a 24MB RAM-Disk on a 16MB Amiga? Or a 48MB-RAM-Disk on a 4MB 486?
Ah, btw, Windows95 installed without much hazzles on the 486-systems and runs quite well.
brandtc@sword:~$ apt-cache search rexx
crashmail - JAM and *.MSG capable Fidonet tosser
exuberant-ctags - multi-language reimplementation of ctags
regina-rexx - The Regina REXX interpreter.
regina2 - The Regina REXX interpreter, run-time library.
regina2-dev - The Regina REXX interpreter, development files.
rexxtk - Interface to Tcl/Tk for Regina REXX
rxsock - Socket function library for Regina REXX
And btw, REXX was best integrated into AmigaOS and not into OS/2. You could for example syncronize Processes, use it for interprocess-communication and the best of all, it was the absolute standard for everything. Nearly all Software features a REXX-port which made the Amiga the ultimate OS for "use several small and nifty pieces to make something really great".
In 26 years of computer gaming only Transport-Tycoon was able to survive ten years of attention and still rock.
Ok, today I am playing the Open-Source-Clone OpenTTD but it is still the same gameplay, just with more depth in resources.
The original TT was rather blunt (needed a 386dx40) but already TT Deluxe was pretty deep in gameplay (fun at 486dx2-66). The Windows-Version TTDW wasn't a big step but is nowadays the base of all further versions (starting at a 486dx4-133).
There is TTDW-Patch which adds hundreds of features to the old original game by doing binary patches at runtime, especially likeable are the very advanced train-signals. I suggest a Pentium-200.
OpenTTD is a clean-room-reimplementation of TTDW+TTDWPatch with source-code and a working internet-multiplayer-system. Needs a pretty tough machine though, at least a PentiumIII-400 but even my Athlon-2100+ has a hard time when playing at 1600x1200 with some thousand vehicles on screen.
My Tapes never survive more than 2-3 years but most if not all my CDRs from ten years ago are still fine and so are my CDRs from five years and my few DVDR from two years ago.
But I have something which surrelly outlives them all and is much cheaper also:
A RAID5.
I have a RAID5 from 1997 around, 4x4GB SCSI. It was switched off most time to reduce stress on the parts but I expect it work for many, many more years. And if one drive fails... hey, its a RAID5.
Then I have another RAID5 made from 5x9GB from 2001. Still working like a charm. I have left them inside the computer from that age so I just have to power up and telnet to the box to access my backup.
And my nowadays RAID5, 4x160GB from 2003, soon going to be update with a "RAID6" with 6x250GB. The old 4x160GB-drives will be put either inside a USB-case or into another computer for easy powerup+telnet-access.
Also I think ethernet and telnet/ssh will be the best option for a longliving standard.
One of the major drawbacks of OpenSSH was the lack of a per-account/per-key based traffic-accounting. I always had the impression that the developer of OpenSSH opposed the basic idea of getting precise data about how much everyone did even if it still is possible for the admin to track everything from outside SSH.
The mainproblems of desktop linux:
-The GUIs are much too overloaded already. Every button which can be removed should be removed.
-The GUI is not following the Unix-mantra of "everything is a file". Too much stuff in deeply nested menues instead of properties of a file (or an icon). Even Gnome/Ubuntu are not able to place a Dial-Up-Icon on the Desktop.
-Too many distributions of files and drivers. When I last looked for a nv-driver it came in as many as 50 instances: four tar.gz, five debs and a whooooping 78 rpms (starting with suse 7.3 and ending with fedora 10.whatever)
!!!Thats insane!!!
At least within ONE distribution major release I expect only ONE driver (say one RPM for SuSE7, one for Debian3).