Red Hat Enterprise Linux 7 Ideas Are Needed

[phoronix]

Phoronix: Red Hat Enterprise Linux 7 Ideas Are Needed

Red Hat is beginning to look for feedback and ideas from their enterprise customers about what they would like to see from Red Hat Enterprise Linux 7, the next major release of their flagship Linux operating system...

http://www.phoronix.com/vr.php?view=OTg1OA

[Shining Arcanine]

Here is my wishlist:

• Official repositories that rival other distributions in having up to date software and a large variety (e.g. >15,000 packages).
• The ability to have the package manager recompile installed software with -march=native to take better advantage of the CPU. (e.g. AVX Support)
• SAMBA 4 so that legacy Windows 7 systems can use RHEL servers as domain login servers.
• First party support for the installation of Microsoft Office in a multiple user configuration via a compatibility layer (e.g. CrossOver, WINE)
• An open source flash plugin like what IcedTea is for Java so that legacy software can be used without compromising security.
• RDP support in KVM-QEMU to enable access to virtual infrastructure from legacy Windows machines on which installing the SPICE client is not always an option.
• Proper KVM PCIe Passthrough Support so PCI devices behind PCIe-to-PCI bridges will work.
• The ability to run the KVM SPICE server on a UNIX domain socket.
• Improvements in the KVM vmware graphics device so that Windows Aero works via the VMWare WDDM 3D Windows driver, both in situations where the host lacks a GPU and in situations where one is installed.
• Virtualized GPGPU functionality in KVM (e.g. CUDA, OpenCL).
• Better POSIX support (e.g. implementation of gettid() in libc, proper implementation of UNIX System V syscall codes, etcetera).
• Better written manpages that explain how signal handling works in conjunction with threads.
• An IDE for UNIX systems that functions similarly to Microsoft Visual Studio in both performance and ease of use.
• First party support for alternative compilers such as LLVM/Clang.
• The ability for system administrators to use GRUB with UEFI systems, without any need for the much more cumbersome GRUB2 (e.g. do the work to push the Fedora UEFI GRUB patches upstream)

I am not a RedHat customer, but I hope that my wishlist will make it to RedHat. I know some organizations that do not use RHEL specifically because of items on my wishlist.

[Shining Arcanine]

Here is an addendum:

• Generic improvements to power management so that the air conditioners in datacenters do not have to work as hard and the electric bills better correspond to the actual work done.

[yogi_berra]

Proper NFS4 support.
Legal mp3 and h.264 codecs.
Stop enabling Lennart Poettering.
Keep Ulrich Drepper away from the public.

[cl333r]

Here is my wishlist:

• An IDE for UNIX systems that functions similarly to Microsoft Visual Studio in both performance and ease of use.

Keep dreaming. Eclipse/NetBeans/QtCreator are the tallest leprechauns in the Unix/Linux world.

As to the System V compatibility - all the stuff that was worth supporting is already supported in Linux. Linux was meant as a Unix-like platform, not as a free Unix swap solution, so the policy is - change your source code to use the (better) Linux solution. On the desktop - Linux is already _the_ OS from which the "true" Unixes like BSD and Solaris copy code and adapt to their needs (graphics drivers, KMS). Get used to it - upgrade your code to the new reality, not the OS (Linux) to the old API.

[DavidNielsen]

I'd love it if they would stop their FUD and shipped Mono with RHEL7. Preferably packaged in a manner more consistent with Upstream (and what debian/Ubuntu does) rather than the backwards Fedora packaging.

[Shining Arcanine]

Keep dreaming. Eclipse/NetBeans/QtCreator are the tallest leprechauns in the Unix/Linux world.

As to the System V compatibility - all the stuff that was worth supporting is already supported in Linux. Linux was meant as a Unix-like platform, not as a free Unix swap solution, so the policy is - change your source code to use the (better) Linux solution. On the desktop - Linux is already _the_ OS from which the "true" Unixes like BSD and Solaris copy code and adapt to their needs (graphics drivers, KMS). Get used to it - upgrade your code to the new reality, not the OS (Linux) to the old API.

You probably have never programmed on Linux. The kernel supports gettid(), but gettid() is not available in userland. It is something that makes threaded programming a pain on Linux and it also introduces unnecessary issues when porting software to Linux.

http://www.kernel.org/doc/man-pages/.../gettid.2.html

With that said, I would like to ask that you refrain from making such ignorant comments, or better yet, that you use Windows. Software developers don't have time to deal with this nonsense.

[allquixotic]

Here is my wishlist:

• Official repositories that rival other distributions in having up to date software and a large variety (e.g. >15,000 packages).
• The ability to have the package manager recompile installed software with -march=native to take better advantage of the CPU. (e.g. AVX Support)
• SAMBA 4 so that legacy Windows 7 systems can use RHEL servers as domain login servers.
• First party support for the installation of Microsoft Office in a multiple user configuration via a compatibility layer (e.g. CrossOver, WINE)
• An open source flash plugin like what IcedTea is for Java so that legacy software can be used without compromising security.
• RDP support in KVM-QEMU to enable access to virtual infrastructure from legacy Windows machines on which installing the SPICE client is not always an option.
• Proper KVM PCIe Passthrough Support so PCI devices behind PCIe-to-PCI bridges will work.
• The ability to run the KVM SPICE server on a UNIX domain socket.
• Improvements in the KVM vmware graphics device so that Windows Aero works via the VMWare WDDM 3D Windows driver, both in situations where the host lacks a GPU and in situations where one is installed.
• Virtualized GPGPU functionality in KVM (e.g. CUDA, OpenCL).
• Better POSIX support (e.g. implementation of gettid() in libc, proper implementation of UNIX System V syscall codes, etcetera).
• Better written manpages that explain how signal handling works in conjunction with threads.
• An IDE for UNIX systems that functions similarly to Microsoft Visual Studio in both performance and ease of use.
• First party support for alternative compilers such as LLVM/Clang.
• The ability for system administrators to use GRUB with UEFI systems, without any need for the much more cumbersome GRUB2 (e.g. do the work to push the Fedora UEFI GRUB patches upstream)

I am not a RedHat customer, but I hope that my wishlist will make it to RedHat. I know some organizations that do not use RHEL specifically because of items on my wishlist.

Fantastic list. To this, I would add

• First party support for cutting-edge but well-tested open source graphics drivers (backport DRM, Mesa, Xorg drivers and components as needed) and continually update and backport from the latest drivers for each point release
• The ability to get paid Red Hat support to fix any application compatibility issues that arise with the open source graphics stack, and ensure that the Red Hat contributions are upstreamed
• Support for the latest mice, keyboards, motherboards, sound cards and peripherals in each point release
• Officially licensed support for an emulated Internet Explorer (wine) - perhaps a deal with Crossover
• Officially licensed support for non-free media decoders - perhaps a deal with Fluendo
• GUI management of container-based virtualization to compete with Parallels' proprietary offering
• Adopt a more liberal update policy for graphical desktop software, especially with point releases (e.g. a point release could upgrade from Gnome 3.0 to Gnome 3.2, but not from Gnome 2.30 to Gnome 3.0)
• Hand-select a small number of extremely popular packages that will be continually integrated into the RHEL update chain as they evolve, based on the quality of upstream QA & testing and level of backwards compatibility. Examples: Firefox, Virtualbox. Major new releases (e.g. Firefox 5.0 -> 6.0) would ship in point releases of RHEL, while point releases of the programs themselves (e.g. Firefox 5.1) would ship as normal updates through RHN without even waiting for a RHEL point release.

In a nutshell, I think RHEL is already the obvious winner for servers, and they should continue along the obvious linear path for improving and deepening their server-side support. What I would really love to see is RHEL being a relevant, up-to-date, aggressive competitor for the enterprise desktop, which is much more prevalent than you think. Practically every white collar office worker in America has a copy of Vista Enterprise, XP Professional or some similar Windows operating system. I want Red Hat to uproot some of that market share that has been owned by Microsoft for two decades. People who use Linux at work will be much more likely to be comfortable using it at home. The beautiful thing about most corporate work environments is that, often, you can't choose which OS you want to run on your computer. It can be a damning, miserable trap if your office insists on using Windows XP through 2050, but on the flip side, if you have an open-minded, forward-thinking IT department, it can also win over a lot of willing Linux users by forcing a standard RHEL Desktop image on all the users and making them get used to it. They will hate the transition, but by forcing them through it (and paying them for the time they spend figuring things out), they will eventually get familiar with it.

[dsmithhfx]

Official repositories that rival other distributions in having up to date software and a large variety (e.g. >15,000 packages).

What part of "Enterprise Linux" don't you understand?

If you want up-to-date packages, go with Fedora. If you want umpty-thousand packages, Debian might be a better bet.

[tomm3h]

What part of "Enterprise Linux" don't you understand?

If you want up-to-date packages, go with Fedora. If you want umpty-thousand packages, Debian might be a better bet.

+1, the OP hasn't really grasped one of the fundamental points of RHEL: provide a stable environment for Enterprises. This means you don't change major software versions, introduce major new features or support a tonne of software that only a tiny, tiny percentage of your customers will actually use.