Mondo is cool, but it is not platform independent and usable on a live CD. It requires installation and compilation on each box to build its own rescue environent.
The objective of my procedure was to come up with a method that could back up ANY Linux distribution at a low level in a very standardized and platform independent way, without sacrificing storage or performance OR TOUCHING THE OS ON THE BOX IN ANY WAY. This works with any Linux system you boot the System Rescue CD on, it requires no package or software installs or copied files onto the system you are backing up. None of the other methods that exist right now (System Imager, Mondo, even dd) will do -all- of the things I wanted it to do. My method is klunky as it is a manual process, but some enterprising programmer could probably script this in perl or python with a simple text menuing system pretty easily and have this built into the next System Rescue CD. That's what I was hoping for.
"Re: NTFS-3G as primary file system
Yes, we're working on this targeted for over a year. For instance full file permission and ownership support has been implemented, we have ported and maintain a POSIX file system test suite (http://ntfs-3g.org/pjd-fstest.html) which we pass 100% using the improved driver.
Apparently there is definitely need for this from a major (multi-million) user segment who are not interested in the technical implementation but just expect things to work out of the box fast, reliably, smoothly interoperating with the already used and common computing environment and without making intrusive, complicated, time consuming changes (partitioning, backup, reinstalling, etc).
The main NTFS-3G driver has been tested as primary file system using an ordinary Gentoo install and it worked surprisingly well. Though this is not yet recommended due to several technical reasons what we are going to address by new public driver releases in the coming months. We have significant amount of unreleased code still in rigid testing.
The open source community is enormous and incredible help in development, improving quality, and defining priorities. Enterprises interested in reliable interoperability on the desktop and in consumer appliances enable us to keep working on the driver and pay our bills.
I think on the long term (10+ years) the main Linux file system could be ZFS or brtfs. File systems mature in many years, especially the very complex ones. For instance the NTFS-3G code base is basically 8 years old already.
You wanna take a poll on how many XFS users there are?
Look, I am not belittling what XFS can do. It has a large amount of success on computing clusters, particularly those for hollywood-style CGI render farms. But this is highly specialized and is somewhat bolstered by the fact that those environments are already familiar with XFS, because they came from SGI environments already. People like to use what is familiar to them. The fact that the SGI IRIX platform is dead doesn't exactly speak well for the software.
I have it on pretty good authority that the reason why OS/2 cannot be open sourced is that in addition to the fact that the OS contains a decent portion of code originating from Microsoft, is that THEY DON'T KNOW WHERE A LOT OF THE CODE IS.
As it turns out, when IBM closed IBM Boca Raton, where most of the development of OS/2 Warp and OS/2 for PowerPC took place, they "lost" a lot of the backup tapes containing the source to a lot of the system components.
If anyone has any code at all, its Serenity Systems, but I believe they don't have the source to the core OS, just the stuff they added or modified.
the best elements of OS/2 worth saving are the versatile REXX scripting language (which already exists in some form for Linux, but its not IBM's implementation -- http://www-306.ibm.com/software/awdtools/obj-rexx/ linux/) and SOM, the System Object Model. WPS itself is useless because it would have to be completely rewritten for X and GTK or Qt instead of PM -- instead, a REXX with a GUI implementation for KDE and GNOME, along with hooks into SOM would give Linux the best aspects of OS/2.
Actually, the first UNIX to run on an x86 platform was SCO XENIX, way, way back in the early 80s. SCO also delivered the first 32-bit Unix to run on a 386 chip as well.
"Turion" sounds very, very similar to the way "Durian" is pronounced in several southeast asian languages.
A Durian, is a large, porcipine-like fruit that when ripe and cut open, smells and tastes like a mix of a sulphur factory (or rotten garlic or onions) and a lush, creamy custard. The smell is so pungent that you can smell it from a hundred feet away. Many people are so put off by the smell that in public areas like airports and hotel lobbies they have signs that say "No Durians"
I was the Developer Liasion at Sharp during 2002-2003 and I wrote a peice in Linux Magazine in May of 2004 about why the Zaurus failed. The short of it? Sharp had no clue about dealing with the Open Source community.
http://www.linux-mag.com/2004-05/hard_01.html
As another person mentioned upthread, there are -other- Linux handhelds in the works, some coming out by major companies. Lets just hope they aren't doomed to repeat the same mistakes.
As I understand it, the POWER architecture can handle 32 and 64 bit executables when running a 64-bit Linux kernel, so I would think any 64-bit limitations in MOL could be easily corrected with some extra development work.
Fedora Core 2 64-bit -- crappy 32-bit support
on
Fedora Core 2 Review
·
· Score: 2, Informative
I've been using Fedora Core 2 thru Test 3 on a brand new IBM Intellistation A-Pro -- thats a dual Opteron 2.2Ghz workstation. And while Fedora Core 2 is a pretty polished product, I have serious issues with how 32-bit libraries and plugins are handled in it.
For starters, EVERY SINGLE APPLICATION is compiled for 64-bit -- that includes stuff that can use 3rd-party plugin modules like Mozilla, GAIM, Mplayer, XINE, etc. Mozilla is one of the worst problems because you can't run pre-compiled 32-bit plugins on the 64-bit browser -- it uses a totally separate/lib64 and/usr/lib64 tree. So stuff like Realplayer, Mplayer (which uses 32-bit dll codecs swiped from Windows to make audio and video work for stuff like Quicktime and Windows Media), Flash won't work. To make matters worse you can't install the 32-bit Mozilla RPMs because the/lib and/usr/lib pre-requisites are not there, and theres no easy way to install them.
SuSE 9.1 handled this differently -- in their 64-bit version they provided duplicate libaries for 32-bit stuff, so you can RPM install 32-bit mozilla, gaim, openoffice, 3rd party apps, whatever, and all of their plugins.
Fedora Core could easily remedy this by doing what SuSE did. I hope they do, because otherwise Fedora Core 2 is a good distro.
These are courses composed by the eGullet.com membership, many of whom are culinary professionals and highly accomplished ameteur cooks. Many of the courses have amazing accompanying digital photos as well.
The site has been nominated for a James Beard Award and has been shortlisted for the GlenFiddich Food and Wine Awards in the UK, and has been featured in many international newspapers and magazines, including TIME.
Disclaimer: I am the founder of said website, so your mileage may vary.
In this horrendous economy you should be glad you have a job in IT and not spending your time surfing Careerpath.com, Monster and DICE.
Keep your mouth shut, do your job to your best of your ability, follow orders and do not concern yourself with making a "difference" in corporate IT. In the 15 years I've consulted and worked in MIS and IT, I've learned that being the confrontational "lets make a difference" guy means that you get identified as the Squeaky Wheel. Even if they brought you in as a consultant to help "fix" the problems, chances are they like to talk the talk, but not walk the walk.
IT departments on the whole are considered expenses, not integral to business at many companies. While this is a dysfunctional and unproductive attitude for businesses to have, its now the norm.
If you want to make a "differnce" become self-employed and start your own company. Even CIO's and MIS directors these days have to listen to the Man. They're just paid more than you are.
When times are good, IT is a tool for improving business. When times are bad, its maintenance mode.
As I said, be happy you're working. Many skilled IT types aren't. Collect your paycheck and be happy, at least until you can move on and call the shots.
This is a post about getting it to work on Xandros Linux, but people using other distros with newer versions of libstdc++ may have the same problem, so you may need to symbolically link the libstdc library on your unit to the version the software calls for as well.
The software is a PIG. Its a Java application and even on my 3.0Ghz HP graphics workstation with a Quadro4 graphics card, its slow and a major memory hog. Still, Its pretty cool.
Its only LOOSELY based on the Power4. There are some architectual similarities, but its kinda like saying a current model Chevy Camaro is based on GM's current NASCAR platform. Sure they share a heritage and the Camaro benefits from Chevys' experience at NASCAR, but they are way different.
The 970 only has a 512K L3 cache using a single CPU core, whereas the current Power4 is a dual-core (meaning two CPUs on a single wafer) using a shared 8MB L3 cache. Its got sixteen times the cache and two CPU's per wafer, not to mention proprietary bus extensions only the IBM Power4 machines have.
If you want to read more about this, check out the march edition of Linux Magazine for a review of the new IBM pSeries 615 -- or go check out the latest Power4 (and 970-based) linux server systems that IBM will be demoing at LinuxWorld Expo next month.
Server chip versus desktop processor? Duh.
on
G5 vs Opteron, Finally
·
· Score: 3, Interesting
This article strictly compares the 970 to the G5 using GAMING benchmarks.
The Opteron and the G5 (IBM PowerPC 970) are two disparately different chips meant to serve two different purposes. The Opteron is AMD's server chip designed to handle for the most part, 64 bit high performance database applications and applications which require large memory models in which the 64 bit memory addressing is needed-- NOT 64 bit desktop applications or games. That's reserved for the Athlon 64 which is clocked significantly higher and has a much smaller L2 cache than the Opteron. On Gaming and desktop content creation benchmarks the Athlon64 is a much better match for the 970.
If you want to compare apples to apples I would compare the IBM Power4 to the Itanium2 to the Opteron, hook them all up to an EMC storage array using fiberoptic SAN connections, and run a few million row length Oracle and DB2 databases and some SQL database benchmarks -- and for load up a few gigantic thermodynamic simulations up into main memory and see how quickly they can run through them. THAT would be an appropriate test for these server chips.
Slashdot Mirror
http://tinyurl.com/5lsa2x
I took these in March of 2007.
http://offthebroiler.wordpress.com/2008/07/06/bletchley-park-the-forge-of-computer-creation/
"I'll agree to this... a computer is nothing more then a machine that performs programmed/specified repetative tasks really fast."
By that definition... the DNA molecule which replicates itself is the oldest peice of computer code running.
I'll look into this and correct if necessary. It appears to be redundant because the config backup files contain the UUID.
Mondo is cool, but it is not platform independent and usable on a live CD. It requires installation and compilation on each box to build its own rescue environent.
The objective of my procedure was to come up with a method that could back up ANY Linux distribution at a low level in a very standardized and platform independent way, without sacrificing storage or performance OR TOUCHING THE OS ON THE BOX IN ANY WAY. This works with any Linux system you boot the System Rescue CD on, it requires no package or software installs or copied files onto the system you are backing up. None of the other methods that exist right now (System Imager, Mondo, even dd) will do -all- of the things I wanted it to do. My method is klunky as it is a manual process, but some enterprising programmer could probably script this in perl or python with a simple text menuing system pretty easily and have this built into the next System Rescue CD. That's what I was hoping for.
http://talkback.zdnet.com/5208-10532-0.html?forumID=1&threadID=47095&messageID=876261&start=0
"Re: NTFS-3G as primary file system
Yes, we're working on this targeted for over a year. For instance full file permission and ownership support has been implemented, we have ported and maintain a POSIX file system test suite (http://ntfs-3g.org/pjd-fstest.html) which we pass 100% using the improved driver.
Apparently there is definitely need for this from a major (multi-million) user segment who are not interested in the technical implementation but just expect things to work out of the box fast, reliably, smoothly interoperating with the already used and common computing environment and without making intrusive, complicated, time consuming changes (partitioning, backup, reinstalling, etc).
The main NTFS-3G driver has been tested as primary file system using an ordinary Gentoo install and it worked surprisingly well. Though this is not yet recommended due to several technical reasons what we are going to address by new public driver releases in the coming months. We have significant amount of unreleased code still in rigid testing.
The open source community is enormous and incredible help in development, improving quality, and defining priorities. Enterprises interested in reliable interoperability on the desktop and in consumer appliances enable us to keep working on the driver and pay our bills.
I think on the long term (10+ years) the main Linux file system could be ZFS or brtfs. File systems mature in many years, especially the very complex ones. For instance the NTFS-3G code base is basically 8 years old already.
Regards, Szabolcs Szakacsits "
You wanna take a poll on how many XFS users there are?
Look, I am not belittling what XFS can do. It has a large amount of success on computing clusters, particularly those for hollywood-style CGI render farms. But this is highly specialized and is somewhat bolstered by the fact that those environments are already familiar with XFS, because they came from SGI environments already. People like to use what is familiar to them. The fact that the SGI IRIX platform is dead doesn't exactly speak well for the software.
See, you forget one little file system, and someone gets pissed. XFS is nice, but its not exactly in widespread use.
Its in the patent map, clear as day.
http://www.centrify.com/downloads/public/microsoft_protocol_to_patent_map_courtesy_of_centrify.xls
"Now mate, look what happens when I shove my whole body up this Jurrasic croc's cloaca. She gets really grumpy, But not as grumpy as my wife!"
I have it on pretty good authority that the reason why OS/2 cannot be open sourced is that in addition to the fact that the OS contains a decent portion of code originating from Microsoft, is that THEY DON'T KNOW WHERE A LOT OF THE CODE IS.
As it turns out, when IBM closed IBM Boca Raton, where most of the development of OS/2 Warp and OS/2 for PowerPC took place, they "lost" a lot of the backup tapes containing the source to a lot of the system components.
If anyone has any code at all, its Serenity Systems, but I believe they don't have the source to the core OS, just the stuff they added or modified.
the best elements of OS/2 worth saving are the versatile REXX scripting language (which already exists in some form for Linux, but its not IBM's implementation -- http://www-306.ibm.com/software/awdtools/obj-rexx/ linux/) and SOM, the System Object Model. WPS itself is useless because it would have to be completely rewritten for X and GTK or Qt instead of PM -- instead, a REXX with a GUI implementation for KDE and GNOME, along with hooks into SOM would give Linux the best aspects of OS/2.
Actually, the first UNIX to run on an x86 platform was SCO XENIX, way, way back in the early 80s. SCO also delivered the first 32-bit Unix to run on a 386 chip as well.
http://forums.egullet.org/index.php?showtopic=3559 5
He's one of the members of eGullet.
"Turion" sounds very, very similar to the way "Durian" is pronounced in several southeast asian languages.
A Durian, is a large, porcipine-like fruit that when ripe and cut open, smells and tastes like a mix of a sulphur factory (or rotten garlic or onions) and a lush, creamy custard. The smell is so pungent that you can smell it from a hundred feet away. Many people are so put off by the smell that in public areas like airports and hotel lobbies they have signs that say "No Durians"
http://www.bbc.co.uk/dna/h2g2/classic/A255575
I was the Developer Liasion at Sharp during 2002-2003 and I wrote a peice in Linux Magazine in May of 2004 about why the Zaurus failed. The short of it? Sharp had no clue about dealing with the Open Source community.
http://www.linux-mag.com/2004-05/hard_01.html
As another person mentioned upthread, there are -other- Linux handhelds in the works, some coming out by major companies. Lets just hope they aren't doomed to repeat the same mistakes.
As I understand it, the POWER architecture can handle 32 and 64 bit executables when running a 64-bit Linux kernel, so I would think any 64-bit limitations in MOL could be easily corrected with some extra development work.
http://freshmeat.net/projects/mol/
http://www.maconlinux.org
I've been using Fedora Core 2 thru Test 3 on a brand new IBM Intellistation A-Pro -- thats a dual Opteron 2.2Ghz workstation. And while Fedora Core 2 is a pretty polished product, I have serious issues with how 32-bit libraries and plugins are handled in it.
/lib64 and /usr/lib64 tree. So stuff like Realplayer, Mplayer (which uses 32-bit dll codecs swiped from Windows to make audio and video work for stuff like Quicktime and Windows Media), Flash won't work. To make matters worse you can't install the 32-bit Mozilla RPMs because the /lib and /usr/lib pre-requisites are not there, and theres no easy way to install them.
For starters, EVERY SINGLE APPLICATION is compiled for 64-bit -- that includes stuff that can use 3rd-party plugin modules like Mozilla, GAIM, Mplayer, XINE, etc. Mozilla is one of the worst problems because you can't run pre-compiled 32-bit plugins on the 64-bit browser -- it uses a totally separate
SuSE 9.1 handled this differently -- in their 64-bit version they provided duplicate libaries for 32-bit stuff, so you can RPM install 32-bit mozilla, gaim, openoffice, 3rd party apps, whatever, and all of their plugins.
Fedora Core could easily remedy this by doing what SuSE did. I hope they do, because otherwise Fedora Core 2 is a good distro.
If you want to learn how to cook, the eGullet Culinary Institute provides free online courses for cooking here:
0 8
http://forums.egullet.com/index.php?showforum=1
These are courses composed by the eGullet.com membership, many of whom are culinary professionals and highly accomplished ameteur cooks. Many of the courses have amazing accompanying digital photos as well.
The site has been nominated for a James Beard Award and has been shortlisted for the GlenFiddich Food and Wine Awards in the UK, and has been featured in many international newspapers and magazines, including TIME.
Disclaimer: I am the founder of said website, so your mileage may vary.
Linux Magazine reviews the IBM p615
In this horrendous economy you should be glad you have a job in IT and not spending your time surfing Careerpath.com, Monster and DICE.
Keep your mouth shut, do your job to your best of your ability, follow orders and do not concern yourself with making a "difference" in corporate IT. In the 15 years I've consulted and worked in MIS and IT, I've learned that being the confrontational "lets make a difference" guy means that you get identified as the Squeaky Wheel. Even if they brought you in as a consultant to help "fix" the problems, chances are they like to talk the talk, but not walk the walk.
IT departments on the whole are considered expenses, not integral to business at many companies. While this is a dysfunctional and unproductive attitude for businesses to have, its now the norm.
If you want to make a "differnce" become self-employed and start your own company. Even CIO's and MIS directors these days have to listen to the Man. They're just paid more than you are.
When times are good, IT is a tool for improving business. When times are bad, its maintenance mode.
As I said, be happy you're working. Many skilled IT types aren't. Collect your paycheck and be happy, at least until you can move on and call the shots.
http://forums.xandros.com/viewtopic.php?t=4233
This is a post about getting it to work on Xandros Linux, but people using other distros with newer versions of libstdc++ may have the same problem, so you may need to symbolically link the libstdc library on your unit to the version the software calls for as well.
The software is a PIG. Its a Java application and even on my 3.0Ghz HP graphics workstation with a Quadro4 graphics card, its slow and a major memory hog. Still, Its pretty cool.
Its only LOOSELY based on the Power4. There are some architectual similarities, but its kinda like saying a current model Chevy Camaro is based on GM's current NASCAR platform. Sure they share a heritage and the Camaro benefits from Chevys' experience at NASCAR, but they are way different.
The 970 only has a 512K L3 cache using a single CPU core, whereas the current Power4 is a dual-core (meaning two CPUs on a single wafer) using a shared 8MB L3 cache. Its got sixteen times the cache and two CPU's per wafer, not to mention proprietary bus extensions only the IBM Power4 machines have.
If you want to read more about this, check out the march edition of Linux Magazine for a review of the new IBM pSeries 615 -- or go check out the latest Power4 (and 970-based) linux server systems that IBM will be demoing at LinuxWorld Expo next month.
This article strictly compares the 970 to the G5 using GAMING benchmarks.
The Opteron and the G5 (IBM PowerPC 970) are two disparately different chips meant to serve two different purposes. The Opteron is AMD's server chip designed to handle for the most part, 64 bit high performance database applications and applications which require large memory models in which the 64 bit memory addressing is needed-- NOT 64 bit desktop applications or games. That's reserved for the Athlon 64 which is clocked significantly higher and has a much smaller L2 cache than the Opteron. On Gaming and desktop content creation benchmarks the Athlon64 is a much better match for the 970.
If you want to compare apples to apples I would compare the IBM Power4 to the Itanium2 to the Opteron, hook them all up to an EMC storage array using fiberoptic SAN connections, and run a few million row length Oracle and DB2 databases and some SQL database benchmarks -- and for load up a few gigantic thermodynamic simulations up into main memory and see how quickly they can run through them. THAT would be an appropriate test for these server chips.