🚀 go-pugleaf

Red Hat Linux 6.0
Just released!

Why spend hours downloading it???
Why spend $70.00???
Get this exclusive offer:  2 CDs for only $7.95 at:
http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?ViewItem&item=104067473
 

Specifications as specified by Red Hat at http://www.RedHat.com


Technical Features of Red Hat Linux 6.0 

Improved SMP support allows better performance in multi-processor machines. 
SMP - Symmetric Multi-Processor
SMP describes a computer architecture that makes use of multiple, identical CPUs sharing common resources such as memory, busses, and devices. By having the kernel properly tuned, as is the case with the 2.2 kernel, nearly linear scalability can be achieved in some types of system loads. 

The 2.2 Linux kernel provides a great amount of improvement when compared to the SMP support that is present in the 2.0 Linux kernel. The improvements come from a rearchitecting of the way in which interupts are handled and the coresponding locks are called and exectured. 

Intel - scales well to 4 CPUs, works on up to 16 CPUs.
SPARC32 - scales well to 4 CPUs.
SPARC64 - scales well to 12 CPUs, works on up to 64 CPUs.
Alpha - works on dual 21264 based machines.


SMP is better overall from 2.0, but the benefits are most obvious when doing a lot of network I/O or to a lesser extent a lot of block I/O. The reason why there are no Intel 16-CPU boxes out there is that the 'simple' APIC configuration supports only 15 CPUs. (0xf is a broadcast ID). Since 15 CPUs is such an ugly number of CPUs, it seems unlikely that anything greater than 8-way x86 boxes will ever exist. 

Another 'big change' in 2.2 is that interrupts now get dynamically routed to all CPUs, and interrupt handlers are 'fully threaded', ie. they can go in parallel with everything (no kernel lock). So while it's true that some of the kernel subsystems have pretty rough locking, about 30-50% of the real work in a busy server is done in interrupts, and those are pretty parallel already. 


The GNOME desktop environment provides a user friendly and intuitive desktop for users.

Advanced New User Interface with a Desktop Metaphor
Red Hat Linux 6.0 ships with the GNOME Desktop. GNOME (which stands for GNU Network Object Model Environment) is an open source project designed to create a standard graphical desktop based entirely on free software. With a File Manager, Control Center, and many robust applications, users will feel "at home" while using GNOME. 

The GNOME Desktop utilizes GTK+, which provides a robust, mature widget set with a modern appearance. GTK+ supports themes, which allows users to change the look of all of their GTK+ applications with a few clicks of the mouse. Users will also enjoy GTK's Drag and Drop functionality, which can work with both Motif and Qt applications. 

Also included is the Enlightenment window manager which works seamlessly with the GNOME Desktop Environment and provides a graphical configuration tool to help create an environment that is right for any working condition. 

Not only is GNOME supplying a desktop for end users, it is laying down a strong foundation for creating modern applications. Developers will find that their work will be enhanced with a base of libraries and tools that help create good looking, easy to use, extensible applications. Many technologies are included in the GNOME development framework: 

Ease of Use - A standard desktop that has a set of consistent user interface elements leading to shorter learning curves for new users. 
Internationalization - GNOME has been coded with various locales in mind. The GTK+ widget set supports the X Input Method Extension to allow for international text entry. 
Language Bindings for Many Programming Languages - The GNOME Application Programming Interface (API) is designed to support many scripting languages as well as compiled languages. Support for C, C++, Perl, Python, Guile, Ada and others already exists and these bindings are constantly being maintained. 
GNOME Metadata - Metadata provides a system to attach arbitrary information to a file. For example, a developer can customize the icon or even a behavior for a particular file. This can be used to override the global settings that bind actions to MIME-types for files. 
GNOME Components - GNOME comes with ORBit, a CORBA ORB implementation that allows code reuse, component programming, and scripting within the desktop. 
GNOME Canvas - The Canvas is a powerful display engine for structured graphics. Applications can use the basic drawing primitives in the canvas to create interactive displays, or define their custom canvas item types for complex displays. 

Software RAID allows for multiple hard drives to be used in combination for increased performance and reliability.

RAID - Redundant Array of Independent Disks
RAID is a technology that allows multiple disks to be combined in ways that provide redundancy and the creation of larger virtual volumes. There are two types of RAID support offered in Red Hat Linux 6.0. The first, is hardware, which will not be discussed here. Please see the hardware compatibility list. The second type is software RAID. Software RAID is disk array management done entirely within the kernel and supporting user-land tools. Software RAID is on par with hardware RAID when speed and reliability are concerned. 

Red Hat Linux 6.0 has the following RAID features: 

Full RAID levels 0, 1, 4, 5, and linear support. 
Threaded rebuild process. 
Fully kernel based configuration. 
Arrays can be moved between Linux boxes without reconstruction. 
Array reconstruction is "backgrounded" using idle system resources. 
Hot-swappable drive support. 
Automatic CPU detection to take advantage of certain CPU optimizations. 
The various RAID levels are defined as: 

linear - 2 or more disks are appended end to end.

0 (a.k.a. striping) - 2 or more disks are "interleaved" together.

1 (a.k.a. mirroring) - provides duplication of data to 2 or more disks.

4 - uses a single disk in the array for storing the parity information (not commonly used because parity disk is a bottleneck).

5 - most common type of RAID. Distributes parity information throughout disks in array.

New version of glibc, the core system library providing NIS+ support, increased i18n awareness, and POSIX compliance.

Now uses the EGCS compiler as the standard compiler for better optimization and improved support for C++.

EGCS - An improved compiler system
Red Hat Linux 6.0 includes the compilers part of the EGCS (Experimental GNU Compiler System), version 1.1.2. The new feature comes with many improvements in the areas of code generation, platform support and compilation speed. A partial list of new features includes: 

Vast improvements in the C++ compiler: name spaces are fully supported, massive template improvements (member template classes; template friends; template template parameters; local classes in templates now supported). Also, exception handling is now thread-safe and protected virtual inheritance is supported. 
Integrated GNU Fortran (g77) compiler and runtime library with improvements. 
Enhanced platform support (code generation with improved performance on Pentium(R) processors and better support for 64bit operations; SPARC port now includes V8 plus and V9 support as well as performance tuning for Ultra class machines; Alpha port has been tuned for the EV6 processor and has an optimized expansion of memcpy/bzero). 
Code optimizations and bug fixes; the compiler now has options to prefer optimizing for code space over optimizing for code speed; memory footprint for the compiler has been significantly reduced and the scheduler is tuned to improve performance of generated code. 
In previous releases of Red Hat Linux we have also shipped GCC, the GNU C compiler as a backup compiler. With Red Hat 6.0, we have switched over completely to EGCS 1.1.2, confident that EGCS has proved its maturity and versatility as a sophisticated compiler system. 


The RPM package technology allows users and administrators to easily upgrade and install software.

RPM - Red Hat Package Manager
RPM 3.0 includes significant new features for system administrators. One of the most noticeable features is disk space checking. Before installing any packages, RPM ensures that the system has enough free space for all of the new packages which are being installed. If there is not enough free space available, RPM reports which filesystems need more space, and how much. 

RPM now examines the entire set of install, upgrade, or uninstall operations before executing any of them. By performing system-wide consistency checks on the entire operation set, RPM can report errors which will occur before starting any of the work. 

RPM has advanced symbolic link handling, which allows it to recognize that two file paths refer to a single file in many cases. This prevents poorly built packages from overwriting files owned by other packages without RPM being able to detect the situation. 

To help system administrators more closely manage their systems, RPM now lets them specify where individual files should be placed when packages are installed, as well as which files should be left out of the filesystem. 

For developers who are building RPM packages, the macro facility has been expanded to permit passing per-platform parameters into the build process. The addition of macros will permit the same packaging to be built on a larger variety of systems without the need for implementing system specific policies in the source package. 

Finally, the RPM build process now supports build prerequisites. The use of build prerequisites can assist in identifying missing components on a build system. 


Latest versions of XFree86 for support of all the latest video cards.

XFree86 - X Server
XFree86 Gets TrueType Font Support, Better Hardware Compatibility, and Tighter Security 

Several enhancements to XFree86, the Linux implementation of the X Window System, have taken place since the 5.2 version. Some of the changes were brought about by the upgrade from XFree86 3.3.2.3 to XFree86 3.3.3.1, others were added to the standard XFree86 distribution. 

XFree86 3.3.3.1 adds support for a much larger variety of popular video cards and chipsets available today. The list includes the popular Matrox G100/G200, NVIDIA-chipset based laptop displays, Imagine 128 cards, Rendition V1000 and V2x00 cards, Weitek P9100 cards, 3Dlabs GLINT 500 series and Permedia series cards, NeoMagic laptop displays, NVidia Riva TNT cards, newer Trident cards, and ATI Rage cards. Many other drivers have been updated as well. 

In addition to these improvements, we have improved the default security settings so that Xauthority is used instead of the older, less capable xhost system. Our version of the "su" command automatically determines if Xauthority is running under X and adds your authorization information. With this improvement, you should never again receive an "X client connection refused" message. 

TrueType font support has been added and enabled right out of the box. Thousands of freeware and shareware TrueType fonts are available on the Internet for you to download and use. The TrueType font support is based on the excellent open source TrueType rendering library FreeType, which generally produces better looking output than the Type1 PostScript font system which is also included in X. 

Additionally, we now enable the X Font Server (xfs) by default, which has several benefits. First of all, fonts are now rendered independently of the main X Server, and thus the noticeable system sluggishness previously associated with font rendering is gone. Second, we have created a simple to use tool, "chkfontpath," which lets a system administrator quickly and easily add or remove fonts to the xfs configuration file. RPMs which install fonts need to simply make a call to chkfontpath in their post-configuration stage, and the fonts will immediately be available to users; no further configuration in the XF86Config file is necessary. Last, xfs has had some important security holes in it plugged. 

We have greatly enhanced the Xconfigurator tool to support the new video cards available, as well as added a "test X screen" display at the end of the installation process to insure that you have properly configured the system. It also offers you the choice to have an X-based login screen at bootup versus the standard, traditional, console-based login. 


Includes Apache, the worlds most popular web server. 
Installation is now even easier than before and adds the ability install via HTTP, NIS configuration, and installation to RAID partitions.

The Installer: Bringing efficient, robust, flexible, installation to Red Hat Linux
The Red Hat Linux 6.0 installer has several new features and improvments over previous versions. Much work has been done to make the Red Hat Linux installer efficient, robust, flexible, and easy to use. These changes fall into three main categories: hardware support, package installation, and installation methods. 

Hardware Support 

Most Symmetric Multi Processor (SMP)-capable motherboards are detected by the installer. A SMP-enabled kernel is installed automatically if a SMP motherboard is detected. 
Several new RAID controllers are now supported in the Red Hat Linux installer. You are now able to install onto AMI MegaRaid, DAC 960 and Compaq SmartArray controllers. 
Many drivers have been added to the Red Hat installer, including new Ethernet drivers and improved Adaptec SCSI driver support. 
Package Installation 

Package scoring in the installer allows the distribution of processor optimized binary packages. During installation, the installer selects the package best suited for your computer. There are kernel packages compiled for i386, Pentium, and Pentium Pro or better processors and the best match will be installed on the system. 
Invididual package selection has been rewritten in Red Hat Linux 6.0. The individual package selection screen is now in a tree format sorted by group. Easy to use navigation keys allow quick selection or deselection of 600+ packages in Red Hat Linux. 
The installer now checks the available disk space on the target system before performing any upgrade or installation operations. This prevents errors caused by running out of disk space. 
Installation Methods 

Installation methods now require only one boot disk. A second disk is only needed if you require PCMCIA support during the installation process. Previously, FTP and hard drive installs required a supplementary disk. Now the installer retrieves the supplementary disk from the FTP server or hard drive image, removing the requirement of two disks. 
Installations over HTTP are now supported. The installer can retrieve files it needs from a Web server. 

The 2.2 kernel is included, and includes improved SMP support, more hardware support, and enhanced NFS server for increased performance.

Notable items in 2.2 Kernel
The following items are either vastly improved in the Linux 2.2 kernel or are completely new. 

Serial Consoles - New on the x86 and Alpha versions. The kernel now echoes all text through a defined serial port. By cleverly attaching a getty to that same serial port, 2 way communications can be achieved over a serial link with the Linux operating system. These configurations are typically known as "headless" machines and have been supported for a long time in the SPARC kernel.


Larger memory support - Native support for larger memory on x86 and Alpha. On x86, the memory system will autodetect up to 1G of RAM and through kernel configurations, can be made to support up to 2G of RAM. Alpha will support up to 1G of RAM. On SPARC32, the kernel will support up to 3.5G of RAM and on SPARC64 the kernel will support up to 64G of RAM.


Larger swap device support - the Linux 2.2 kernel now supports a single swap partition device size up to 2G. Swap space is used by the virtual memory subsystem as storage for storing pages of memory that are not presently being actively used.


General file system improvements


large file descriptor support - this allows for a significantly large number of file descriptors to be open at the same time, needed for large data bases.

CODA file system support - this is a large, network aware file system that is capable of maintaining coherency of data in a distributed environment as well as offering a much higher level of access security control. CODA also provides transparent caching, server replication and off-line operation.

sendfile() support - allows for very fast movement of data within the kernel by redirecting pointers rather than actually copying data between open file descriptors.

Networking enhancements


traffic shaping - allows a Linux router to control the amount of bandwidth that is used by various network devices and ports.

ipchains - new, hierarchal firewall rule-set tool. This is a much improved replacement for ipfwadm.

GigEthernet support

IPv6 support

Generic Alpha kernel


Hardware differences that previously required specific kernels for each unique Alpha architecture have now been combined into a self-detection schema that automatically configures the kernel at boot time.

Frame buffer support on x86 and Alpha


PC99 spec dictates that VGA modes be deprecated on Intel based hardware. Frame buffer support has been integrated to provide support for machines that will not have VGA support as well as provide accelerated video support in non X modes.

SPARC has always used frame buffers, this is only new to x86 and Alpha kernels.

New console group for easier user access to local devices like floppy drives and CD-ROMs. 

Enhanced System Device Accessibility
Red Hat Linux 6.0 introduces the concept of a "console user." Those using the machine locally (that is, not remotely via network, modem or serial connection) can take advantage of many features more easily than before. 

In the past, to access the floppy drive, sound system and some other devices, users had a few options, all of which were slightly inconvenient. One option was to allow every user on the system to access all these devices; another was letting only one user (regardless of login location) have access to the devices, and a third option was to use the su command to become superuser in order to use the devices. 

When you log into a Red Hat Linux 6.0 system locally, you are considered a console user, and (if no one else is currently logged in locally) you are given ownership of all "console devices" such as floppies, the sound card. This allows you to to play sounds, look at files on the floppy, removable drive, and use other devices, with no further effort. When you log out, those devices become ready for the next user who logs in locally. 

Similarly, while users at the console have always been able to reboot with control-alt-delete (or, if they are in X, something like control-alt-F1 control-alt-delete), that is not precisely intuitive. Now, users at the console can use menu items within the GNOME desktop or the shutdown command at the command line to easily shut the system down cleanly. GNOME even asks console users at logout time if they wish to reboot or halt the computer. 

System administrators who are unnerved by this new feature can relax--it is all configurable, and you can even disable it entirely if you wish. Advanced Linux users can also customize the console user system in myriad ways, even redefining what is considered a console. For more information, man 8 pam_console. 

Portions of this description are direct quotes from Red Hat software (http://www.RedHat.com) and portions are copyright Red Hat software.  All other copyrights are the property of their respective owners.