Red Hat Enterprise Linux 6.5 Preps New Capabilities

[garegin]

The security benefits are of neglible use since enterprise is running only stuff either from the repos or other certified software.
Who says RH doesn't care about good guis? They employee UX people that work on design patterns for ALL RH products except Gnome (as I understand it), however, since those same UX people would probably shoot the gnome designers first
Why are you getting so worked up? Who are you talking to?
RH wants to have guis for most of those sysadmin duties but they take time to make. So, no, RH doesn't want sysadmins to be high priests. They want their systems to be as reliable as possible and guis help with that by creating many fewer paths to fix a problem.
Scripting will still be available for the more unusual stuff just as you have in windows, but guis just make things less error prone as compared to scripts.

hahhah. the GUI on RHEL is a joke. it doesn't expose ANY administrative functions. besides, copying files around or editing text files in gedit all administration is only done in the CLI. you can't even share a folder in the GUI. not even change ACLs. my $20 tp-link router has more advanced network config settings in the GUI than RHEL. at least Suse tried hard with yast, and I give them props for that. And really an X11 GUI is a security risk on a server. More surface area to attack.

[Ericg]

you can't even share a folder in the GUI. not even change ACLs. my $20 tp-link router has more advanced network config settings in the GUI than RHEL.

system-config suite...? system-config-samba and system-config-firewall come to mind. Can't vouch for how EFFECTIVE system-config-samba is, but I know the firewall one works just fine

[liam]

hahhah. the GUI on RHEL is a joke. it doesn't expose ANY administrative functions. besides, copying files around or editing text files in gedit all administration is only done in the CLI. you can't even share a folder in the GUI. not even change ACLs. my $20 tp-link router has more advanced network config settings in the GUI than RHEL. at least Suse tried hard with yast, and I give them props for that. And really an X11 GUI is a security risk on a server. More surface area to attack.

I said WANTS TO HAVE, not HAS CURRENTLY.
For the old guis look at system-config-* (there's one for selinux too).
For something a bit newer there's firewall-config.
BTW, why would X be installed on an externally facing server? That seems like a real corner-case.

@ericg

I didn't see your response before i posted mine. I decided to keep the post as is since I wanted to address a specific case he mentioned (acls).

[tomato]

GUI configs in RHEL don't matter for a very simple reason: RHEL is not a distro for SMBs with Steve, the accounting guy, doing the administration, it's for enterprises that have dedicated, knowledgeable people that do the administration. And guess what? People like that don't care for GUI configuration tools. GUI configuration tools don't scale for dozens or hundreds of servers. CLI does, and CLI does that in many different ways, and you can pick which one you want to use.

[garegin]

GUI configs in RHEL don't matter for a very simple reason: RHEL is not a distro for SMBs with Steve, the accounting guy, doing the administration, it's for enterprises that have dedicated, knowledgeable people that do the administration. And guess what? People like that don't care for GUI configuration tools. GUI configuration tools don't scale for dozens or hundreds of servers. CLI does, and CLI does that in many different ways, and you can pick which one you want to use.

I agree. Which makes having a good GUI pointless. I don't think you should run a GUI on a webserver. As a matter of fact the kernel.org was hacked through a vulnerability in x.org, right?

[RahulSundaram]

I agree. Which makes having a good GUI pointless. I don't think you should run a GUI on a webserver. As a matter of fact the kernel.org was hacked through a vulnerability in x.org, right?

They didn't publish any such details so I don't think you can make that claim. For the people arguing about servers, you forget that RHEL is used by several major organizations including Dreamworks and so forth for thousands of workstations. Some servers certainly need a GUI because the admin panel etc are graphical. Ex: Oracle stuff

[MartinN]

They didn't publish any such details so I don't think you can make that claim. For the people arguing about servers, you forget that RHEL is used by several major organizations including Dreamworks and so forth for thousands of workstations. Some servers certainly need a GUI because the admin panel etc are graphical. Ex: Oracle stuff

Amazon loves RHEL too

[liam]

GUI configs in RHEL don't matter for a very simple reason: RHEL is not a distro for SMBs with Steve, the accounting guy, doing the administration, it's for enterprises that have dedicated, knowledgeable people that do the administration. And guess what? People like that don't care for GUI configuration tools. GUI configuration tools don't scale for dozens or hundreds of servers. CLI does, and CLI does that in many different ways, and you can pick which one you want to use.

I take it you're not aware of RHs movement into infrastucture (big data, cloud management, paas). For those building really good (web) guis is part of the process. Is expected admins will be using those.
For the more traditional desktop/server admin they are, increasingly, having to compete with windows since the old unix market is coming to resemble an increasingly dessicated, say, fruit. Those windows admins make heavy use of guis for many, many tasks. Of course they have the wonderful powershell to let them handle the cases not covered by the guis, but the vast majority of work they'll be doing will have a gui.

I agree. Which makes having a good GUI pointless. I don't think you should run a GUI on a webserver. As a matter of fact the kernel.org was hacked through a vulnerability in x.org, right?

You say that but above you were talking about running X on a server

[Serge]

Exactly... and on Unix (Linux rather), an 'appealing' alternative is to install things like ajax-enabled/wannabe-rich-GUIs like Webmin, or cPanel for administering the box (cpanel isn't even free, AFAIK)... but not a native GUI, like what you see in OS X's Preferences or Gnome's/KDE's attempt at supplying easy to use configuration GUIs...

garegin nailed it - had it not been for this ingrained, entrenched elitist attitude of many Unix CLI kings..... Linux, or even UNIX in general would've been a different beast today. Jobs seized on this - putting on a nice GUI (NeXT) on a UNIX-based OS, and look where that got Apple... whereas Linux stayed in its server niche for the longest time.

Thankfully, this is all past-based, and the momentum to bring a viable linux gui without the crufty 30+ year old X legacy that can benefit corporations and end-users/developer alike has picked up more steam in the last 2 years or so than in the last 5-10 years prior. Intel, RedHat, gnome, kde/qt... there's definite consensus and motivation of where things on the desktop (or mobile) are headed, and even some positive reinforcement in the shape of Mir and Shuttleworth labeling the Wayland effort "a repeat of past (X) mistakes"...

The lack of a solid GUI for Linux has been (and still currently is, though nowhere near to the same extent) the sole, major demoralizing factor for wider Linux adoption, beyond the server room.

This is revisionist history. Your analysis gives the impression that NeXTStep and Linux-based OS'es were developed largely in parallel, and that Steve Jobs's emphasis on graphical experience caused NeXTStep to pull ahead of Linux-based OS'es. This is incorrect in two ways:
•NeXTStep predates the Linux kernel by about three years. The last version of NeXTStep was released before Linux-based OS'es reached a functionality level that could make fair comparisons possible.
•NeXTStep and NeXT Computer were market failures. Focusing on the graphical experience was not some silver bullet that bred success and desktop adoption.

I think an argument could be made that Apple's brand recognition and marketing competence had more to do with OS X's success than OS X's NeXTStep base did. However, Apple continued to struggle to sell computers until the iPod and iPhone made Apple trendy again. I have read in the past that Apple was experiencing steady decline in desktop computer sales before they released the iPhone. I wish I could find hard numbers to back this up, but unfortunately my Google skills are not up to par. The oldest stats I found Googling put OS X at about 3% in 2007 (the release year of the first iPhone). Keep in mind that OS X was released in 2001. 3% by 2007 is not exactly a sign of great, metioric rise. Since then, however, pretty much all usage share stats indicate consistent growth in OS X adoption. You can draw whatever conclusion you want to from that. The one that I draw is that iPhone and other iOS devices had more to do with OS X becoming appealing than OS X itself did. The iPhone was a success, and Apple's other products got to tag along for the ride.

I agree that a polished and intuitive graphical experience is very important to driving adoption of Linux-based operating systems. However, I believe that there are many other factors that are also very important and should not be overlooked. Pointing to OS X as an example of an operating system that was successful primarly because it had a better graphical experience is factually wrong and misleading. There is no single simple solution that will cause Linux-based OS'es to break into the desktop mainstream (except, maybe, a large and very expensive marketing campaign), and arguing that there is does not help the situation.

[liam]

This is revisionist history. Your analysis gives the impression that NeXTStep and Linux-based OS'es were developed largely in parallel, and that Steve Jobs's emphasis on graphical experience caused NeXTStep to pull ahead of Linux-based OS'es. This is incorrect in two ways:
?NeXTStep predates the Linux kernel by about three years. The last version of NeXTStep was released before Linux-based OS'es reached a functionality level that could make fair comparisons possible.
?NeXTStep and NeXT Computer were market failures. Focusing on the graphical experience was not some silver bullet that bred success and desktop adoption.

I think an argument could be made that Apple's brand recognition and marketing competence had more to do with OS X's success than OS X's NeXTStep base did. However, Apple continued to struggle to sell computers until the iPod and iPhone made Apple trendy again. I have read in the past that Apple was experiencing steady decline in desktop computer sales before they released the iPhone. I wish I could find hard numbers to back this up, but unfortunately my Google skills are not up to par. The oldest stats I found Googling put OS X at about 3% in 2007 (the release year of the first iPhone). Keep in mind that OS X was released in 2001. 3% by 2007 is not exactly a sign of great, metioric rise. Since then, however, pretty much all usage share stats indicate consistent growth in OS X adoption. You can draw whatever conclusion you want to from that. The one that I draw is that iPhone and other iOS devices had more to do with OS X becoming appealing than OS X itself did. The iPhone was a success, and Apple's other products got to tag along for the ride.

I agree that a polished and intuitive graphical experience is very important to driving adoption of Linux-based operating systems. However, I believe that there are many other factors that are also very important and should not be overlooked. Pointing to OS X as an example of an operating system that was successful primarly because it had a better graphical experience is factually wrong and misleading. There is no single simple solution that will cause Linux-based OS'es to break into the desktop mainstream (except, maybe, a large and very expensive marketing campaign), and arguing that there is does not help the situation.

We need more than marketing. We need standards. Alsa needs to be fixed. We need a solution to configuration and management that doesn't include bash scripts or, in general, opening a cli. Wayland will help termendously with graphical stability, but we are technically lacking in some areas and no amount of marketing will fix that.