Red Hat redeploys one of its main desktop developers A blog post from senior Red Hat developer Bastien Nocera indicates that the IBM-owned distro maker is further consolidating its development efforts on desktop Linux. The post, simply titled "New responsibilities", refers back to Red Hat's earlier decision to stop packing LibreOffice for RHEL back in June. That in turn followed …
Zero.
Wayland is a complete waste of developer energy and effort. Extending Xorg and providing good quality hardware support, especially open source hardware acceleration, and avoiding GTK past version 3, should be their #1 priority, and NOT time-wasting on what could easily be described as "The 'Supriority by Arthur C. Clarke' moment" also known as Wayland.
export DISPLAY=othermachine:0.0
This is the #1 BEST feature of Xorg, especially for embedded development.
[I go out of my way to ensure Wayland does NOT get installed]
You're welcome.
A single-threaded display server which (due to inherent technical limitations) prevented software like Adobe Photoshop from ever being ported over, even if Adobe wanted to. Without a replacement, there would never be decent open source hardware acceleration as it would forever remain bottlenecked in such a stupid way due to immutable design limitations. For similar reasons, it was also responsible for a number of stupid problems which would cause unnecessary full-screen lockups under high disk I/O, as demonstrated across Linux, OpenSolaris and FreeBSD. Then you have the lack of security which prevented simple mandatory access controls from being implemented to secure GUI apps (like macOS has done competently for many years) as Red Hat’s Dan Walsh commented on in his LiveJournal back in the Fedora Core days.
Feel free to bash the Wayland protocol to your hearts content (IMHO a Quartz clone as a replacement windowing server probably have worked out better) but please stop pretending that Xorg is worthy of extending or supporting, it really isn’t,
[Author here]
OK, multiple points here. I think you are wrong, but it's more complicated and I think there is benefit in spelling it out.
> Wayland is a complete waste of developer energy and effort.
Nope. It is another systemd: it's a legit answer to real issues that real Linux buyers/users/developers have. I will return to this point in a moment.
> Extending Xorg
Nope. IMHO the answer, if these is one, is giving it a drastic prune, not extending it.
> and providing good quality hardware support
Mostly already there.
> especially open source hardware acceleration,
Only 1 company hasn't got this message and frankly I suspect that that will only change in accordance with Max Planck's dictum: "science advances one funeral at a time".
> and avoiding GTK past version 3
Whole separate issue.
> should be their #1 priority
Whose? Why? Can you justify this bold claim?
> and NOT time-wasting on what could easily be described as "The 'Supriority by Arthur C. Clarke' moment" also known as Wayland.
Now, that is another argument.
The AC below you makes good points.
Linux is still mainly volunteer-driven. You just volunteered yourself. Well done. Let me explain how and why and what you need to do.
* Linux is a huge commercial tool and billions go into its R&D and basically none of that is on the desktop.
* The desktop stuff is a tiny corner case and you need to acknowledge that.
* This bit still is mostly volunteer driven. None of those big bucks go into this.
* KDE screwed itself 2 ways out of the door in 1997/98.
[a] It uses C++
[b] It uses a non-FOSS toolkit.
* That means it was the first all-FOSS-all-the-way-down competitor that runs the game and sets the rules. That is GNOME. That's how we got here.
* The devs don't want to work on X11 because it's too big and hairy and complicated and weird and old fashioned.
That is _why_ the old timers like it but let's be real here.
They are scratching their itch, they are making something newer that is easier and cleaner and more fun.
They've done it by ditching all the legacy cruft they didn't need. Unfortunately that included 1 core function that the old timers desperately want and need.
If you want to keep the old system alive, then the only way to do it will be to find a way to ditch all that legacy cruft as well, and just keep the bits that you desperately want that the Wayland folks did not.
Which is why I keep suggesting waking up the X12 project and get together and write some docs and decide what it is that you really need to keep and what you can afford to lose.
There are no mono displays any more. There are no X terminals. Compatibility with 1980s kit does not matter any more.
All that matters is FOSS xNix on little-endian kit.
Stop bitching on forums and start a conversation about what you *need* to *keep* in the X window system that Wayland doesn't do, or within 2-4 Ubuntu LTS generations it will be gone.
> Wayland is a complete waste of developer energy and effort.
Nope. It is another systemd: it's a legit answer to real issues that real Linux buyers/users/developers have. I will return to this point in a moment.
The entire original premise of systemd was to "make it boot faster" [1] and do away with init scripts. It's since grown to an obscene tentacled monster, trying to pervade every facet of linux [2]. There's a reason why on all my server VMs, the only one running a distro that uses systemd (Ubuntu LTS) was because I needed to spin up Mastodon and none of the systemd-free distribution had it in their package lists.
So, mostly, I run FreeBSD or Devuan.
[1] How often do you need to boot your server? Or even a laptop - just close the lid and let it sleep. It's not Windows dammit!
[2] Binary logs requiring a viewer to read them? *Really* Which idiot thought that was a good idea? What happens when the VM goes badly wrong and you need to read the logs but the binary no longer works?
The X.org days may be numbered, but I suspect it's not a small number.
Years, not months, as seems to be the impression that some authors and commentators are trying to create. Whether that is happening for clicks and headlines, or maybe in some cases to try to fulfill the "prophesy" by repeating it a lot, I don't know.
Seems like X.org as a project and code base has been pretty quiet for a while, but that's not really the same thing as defunct or abandoned, which other projects have fallen to over the years.
We do, but I agree with fellow commentard, there is power in the language we use to describe reality, so it should at least be a style convention so we never forget those times when this company was spearheading Linux.
Also, for those who may just see the headline on a search result.
[Author here]
> Shouldn't the headline read "IBM's Red Hat division redeploys ... "?
FWIW, the big boss has already replied above, but ISTM that he took your commend as being 100% straight, whereas I am taking it as deadpan.
So, yes, Chris is right: we all know IBM now owns RH.
*However...*
There is a lot of rhetoric about the IBM/RH deal and it's a nod to that that I refer to them as "Big Purple" or "Purple Hat" or other facetious nicknames.
I can't find it again now, but before I was a staffer at the Reg, as a former RH staffer who had also interviewed at IBM, when I heard about the deal, my summary was something like this:
"Bloated, sclerotic industry giant, long devoid of innovation and dependent on declining subscription revenues from legacy customers, in shock purchase by IBM."
While it is true that, over time, most companies which get acquired do gradually get subsumed into their parent and lose their distinct identity, and this is particularly true of Lotus for instance, it hasn't happened yet with RH.
I generally do not believe company PR statements. This comes as a nasty shock to many large companies, especially ones that are American culturally dominated. They generally react extremely poorly to British journos doubting their words.
But inside and out, officially and privately off the record, I am consistently and repeatedly told that IBM is exerting no substantial control over RH.
Provisionally. Yet. So far.
But still the case.
I do think that the big controversial moves by the company of late are entirely its own choices.
> But inside and out, officially and privately off the record, I am consistently and repeatedly told that IBM is exerting no substantial control over RH. ... I do think that the big controversial moves by the company of late are entirely its own choices.
Yes. Blaming IBM is at least implicitly giving Red Hat management a pass for the bad decisions they've been making. Red Hat executives seem quite capable, proficient even, at foot shooting, even if they don't recognize it at such.
> Provisionally. Yet. So far.
Again, yes. I expect when IBM gets around to really mucking with Red Hat, it will be very obvious, and presumably much, much worse.
But I suspect we would prefer they forfeited the red fedora.
A lilac homburg might be more in keeping these changed circumstances. Or at least a trilby.
Tophats would make this outfit look like a procesion of undertakers (lp ~ funeral directors) which might indeed be prophetic.
The big enterprise vendors long ago decided there's little to no money to be made from Linux on the desktop, which is why it's left up to free distros and non-profit organizations
This really isn't good for us. Our Linux workstations are setup with the same libraries are present on HPC and scientific servers[0] to enable developers & data scientists to deploy and/or scale out their code.
[0] as opposed to database or other webservers which have many fewer libraries for obvious reasons
[Author here]
> This really isn't good for us.
Please do not get me wrong. I entirely agree, and I am most definitely *not* saying this is a good thing.
Also, I think they are wrong.
But it is the case, at least for now.
Only one company has successfully shipped Linux on the desktop in mass volumes, and that is Google.
And it doesn't charge for it.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
