What is it about message boards that make normally civilized people turn into stereotypical razor-tounged neophytes? I guess the lack of needing to stare your debatee in the face removes all tracese of politeness. Snideness aside, I enjoy debaters like you, as it helps to refines one's arguments on both sides. Most people at this point have been "X rocks, you suck!" without substantiating their claims. I thank you for an intelligent debate (peppered with cheap insults as they may be).
Looked at it. Your argument is full of inaccuracies, (for example, the Mozilla developers did not "develop their own widget sets"), enough to make it completely nonsensical.
Mozilla under X has it's own font control, theme control, widget set independent of X and KDE/Gnome. It's GTK-based, but independent of Gnome or even KDE. It requires extra coding and extra bloat as opposed to developing to an environment you already know is there (Mac OS X, MS Windows).
And users just flocked to it! You're also missing the point: X is not a GUI. Repeat after me: X is not a GUI.
What's a GUI? Graphical User Interface. X is graphical, users use it, to wait for it, interface with things. Computer-related things I would imagine. You're splitting hairs on trivial matters to try to make your point look more valid. Old trick.
As per the other point, users have really flocked to the multitude of attempts at a Linux desktop, haven't they.
While that is useful for the corporate and geek environment, it is absolutely useless for a consumer.
Where's the proof? What is it that makes it "useless" Spit it out, buddy! Don't keep us guessing!
The "useless" referred to network transparency from the origional comment that I was commenting to. Millions of people use GUIs/environments/desktops (whatever you feel like calling it) without network transparency. Hence, it's useless to the average user. I agree, it's very useful in the corporate/sysadmin setting.
What gets X "near the usablitiy and even asthetic (sic) qualities of other GUIs" is software like KDE and GNOME. These are not X.
No, they aren't X. But there aren't any Linux desktop distributions that come without one or both. Without KDE or Gnome, X is unusable for consumers, even with a good window manager. Hence why I said "X environment". Perhaps "X-based environment" would be even more accurate. And thanks for pointing out my spelling error with that [sic], it also adds quite a bit of validity to your argument.
I can only guess how you draw your extraordinary conclusion that "[X] becomes slow". I'm assuming that you mean that KDE is slow, or that GNOME is slow. You'd be right, of course - they are slow.
Unfortunately, your failure to understand why they are slow leads you to the wrong conclusion. You can do much more to speed up KDE by fixing Linux's dynamic linking mechanism than by replacing X. In fact, you could replace X completely, port KDE to your spiffy new X replacement, and KDE would still be slow.
It sounds like you're the one with a failure to understand why they are slow. X isn't that efficient to begin with, it sacrificed speed for network transparency by choice because of the specific requirements to be flexible for things like X-terms and other engineering-related applications. Processor speeds are now fast enough that X by itself is relatively fast despite this inefficiency. But what consumer is going to use X with twm?
KDE and Gnome are both bloated attempts to get X to do what it wasn't designed to do, at least not efficiently. You can fix linking systems all you want, Linux's dyanmic linking system has nothing to do with it. It's still slow on FreeBSD, and plenty slow on other operating systems.
Take any other windowing system on the same hardware, on hardware years old, and X will always be slower when you include modern features such as widget/theme control, centralized font control, docs, etc. (that bascially require KDE and Gnome). BeOS, NeXTSTEP, Windows, Mac OS, Mac OS X, all beat X in the speed game.
What many others? And you haven't actually listed any flaws.
I've listed plenty of other flaws in plenty of other posts, but for your sake, I'll list a few:
The X windowing system was designed to a completely different set of specifications than what consumers want and need. It's got ugly fonts, lack of configuraiton, splintered distributions, splintered environments, splintered specifications, and more. These aren't problems in the server realm. They aren't even problems in the geek desktop realm, but they are issues with the consumer desktop realm. And that's what we're discussing here, isn't it?
Various distributions and projects are addressing some of these issues, but not all. You say X is slow really because KDE and Gnome are slow. Well, it's hard to imagine a consumer-grade desktop without one of them. So again, that makes it an X-based issue.
KDE and Gnome are at war and casualty is Linux on the desktop. Kill one off, and you've still got a slow desktop that is at least can be universally developed for. Not in-itself an X problem, but X's lack of features necessitates some sort of all-encompasing environment, which then does make it an X problem.
There is a lack of any kind of unifying software installation. Lindows took care of that, quite nicely in fact. But you're stuck with the Lindows distribution to get it. Other distributions have the same issue. More splintering. Not an X issue? Perhaps not directly, but as X doesn't have a unified graphical software installation specification, leaving a void that has never been adequately filled.
What's more, you acknowledge that whatever flaws you imagine exist in X can actually be fixed. In light of that, why ditch X?
Maybe they can fix X, but with more and more and more hacks, patches, and extra addons. You've already got half a dozen graphics libraries, layer upon layer of innefficiency piling up and slowing things down. In there end, there are so many add ons to X that aren't X or close to X, why use X to begin with? Because it was there?
I say stop the madness. X had a good run, it's great for engineers and sysadmins, but it's just too much to try to get it patched up to modern specs, and even then, it's slow. Why not just start over?
BeOS, NeXTSTEP, OS X, all fast, modern, elegant, graphical environments: Everything that X and Gnome/KDE are not. It's been done before, it can be done again, and with open source, be done even better.
You also say:
I got tired of ugly fonts,
...One word: XRender
Yet another hack on top of X
poor performance,
...Not an X issue
Ahh, but since X isn't any good as a consumer desktop without KDE or Gnome, it does become and X issue.
battling environments,
...Get a decent distribution
How is any distribution going to solve the Gnome versus KDE issue? You're not making any sense there.
the need for tinkering,
... Get a decent distribution
Distributions now install beautifully, but deveating from the narrow scope of any distribution will most likely require tinkering.
and at the same time pretending it was somehow better than Windows, so I stopped using it.
Finally You make a coherent point! I thought I'd never see the day.
I'm sure it was a lapse of stupidity on my part, I apologize.
The interesting question, for the rest of us, is "will the Linux desktop ever be good enough for the likes of Picky Tokki?"
Nope, the real question for those that care about the future of open source on the desktop is will there ever be a Linux desktop good enough for the likes of the general masses. Obviously there is nothing out there now, and I don't see anything, even fancy Red Hat and Lindows betas, coming close anytime soon. The entire setup of X/KDE/Gnome/extra hacks is incredibly cumbersome. Geeks might tolerate it because they've got the time and experience, but I doubt consumers ever will.
I've seen indications that some companies are determined to drag Linux, kicking and screaming, and place it square on the desktop. RedHat's most recent beta impressed the hell out of me, for example, precisely because even after installing it, I still have no idea what its XF86Config file looks like, or even where it is. Out of the box, the fonts look nice, the system administration tasks I had to do (Setting up networking with DHCP) didn't require dropping to the command line, most error messages made sense... At this rate, they'll have something solid in a couple of releases.
Indications? Are you referring to outright statements from CEOs and advocates to get Linux kicking and screaming onto the desktop? For years pundits have been claiming that mainstream consumer-grade Linux desktops are right around the corner, a single Red Hat beta away, but it's never come. Keep on hoping.
There is so much out there that's more efficient, elegant, and easy to use than the X/Gnome/KDE/Whatever solution. Let's at least explore moving forward in another direction.
That's a non-sequitur. X is the functional equivalent of GDI, which hasn't changed that much since Windows 1.0. GDI has absolutely NO impact on the Windows UI. By the same token, X has no impact on the Unix GUI. Therefore, there's nothing for a user to like or not to like about X.
Yet still Windows is a much more elegant and easy to use environment than X with KDE or Gnome. Issues like configuration, font control, performance, media, two completely different APIs built on even older APIs, dozens of required package installations and libraries, and overall plumbing are all areas where X is behind Windows. And that's not a compliment to Microsoft. When Windows was just coming out with Windows 95, there were other GUIs/windowing environments that can compete today even years ago.
Take NeXT. I worked at a company where everyone had a NeXT workstation, everyone from the receptionist to the sysadmins to the CEO. Despite their expense, the graphical environment on these workstations were fast, all on technology almost 10 years old.
Everyone loved using them, from the receptionist to the hardcore sysadmins. They were easy to use, easy to configure, and easy to manage and install applications. At the same time, they provided all the flexibility and power that us techies needed.
They ran on 68040's running at 25 Mhz; 33 Mhz if you were lucky, on a wopping 32 MB of RAM. Yet still, the NeXTSTEP environment was elegant, easy to develop for, powerful, and fast, and with 12-bit video when most everyone, except perhaps high-end Apple and SGI, was runing 8-bit video at most. Window movements, redraws, where all amazingly fast on antiquated hardware, and it still had all the bells and whistles of a modern GUI. You could drag a file from your home directory and place it in an email, and it'd attach automatically. This is common place today of course, but this was before Microsoft could even get Windows 95 out the door.
All of this was on a modified BSD 4.3 underpinning running on a Mach kernel. Everyone claims X is light and fast, but NeXTSTEP puts it to shame on the same hardware in both speed, functionality, and overall usability.
NeXT wasn't the only company with a powerful, easy, and elegant GUI around that time either, BeOS was also a system under the healm of Jean-Louis Gassee.
Both of those systems were doing full motion video, gaming, 3D, and even millions of colors years before X systems even cared or even Microsoft.
Now why can't the open source community come up with something like that? Why wouldn't a brand new windowing system, designed from the ground up, without all the complicated libraries, without the slow network-based plumbing, without the battling environments, without the technical baggage, with capabilities that other even non-MS GUIs came up years before, be good for getting open source on the desktop?
Thus far the arguments pro-X11 come down to "X is awesome you suck" which is hardly an argument, "there's nothing wrong with X" which I agree with if you're a techie, but if you're trying to get Linux and others on the desktop then I respectfully believe you to be wrong, and the argument that there is too much running on X today.
The only argument that I've heard thus far that I think holds merrit is that there are thousands of apps running on X currently. Like I've said, no great leaps in technology has come without a price, without abandoning old code and old ways. There is always resistence, everyone likes the status quo. X is the establishment in the open source world. Not because it's the best, but because it's the only thing out there right now. Remind you of any other situations?
Porting apps to a new environment wouldn't be all that difficult. Even mozilla, which is a daunting project itself, has been ported to Mac OS X's Aqua environment with Chimera. Port Mozilla, an instant messenger app like GAIM, OpenOffice, and a few other widely used X apps and you're halfway there to kicking Microsoft's ass.
Creating a legacy environment within the new environment wouldn't be that difficult, except possibly that the old environment was so complicated (X11 libs, KDE, Gnome, ugly fonts). Still, X can run atop BeOS, Windows, NeXTSTEP, and Mac OS X, making for a functional transition period.
So that issue has been addressed.
This isn't a very popular opinion, but slashdot is a place to discuss various opinions, or so I'm lead to believe.
I have to disagree with you whole-heartadly. If you're a geek, and use to.xinitrc files and loading up the latest GTK libraries, sure X is easy for you and it's got great usability. But if you're a regular consumer, the slightest deviation from most distributions (such as Mandrake or Red Hat), then it's a nightmare. There is just too much that requires a command line, and not even an easy command line.
Personally, I prefer Mac OS X and BeOS as my preferred desktops. They are elegant, clean, and easy to use, so no, Windows isn't my golden rule. It is, however, better than X in many, many, many respects.
I've listed several reasons in other posts for other articles, hence my claim to knee-jerk reactions and my own reasons for not liking X. I don't want to be too redundant here.
Bascially, my main beefs with X are lousy fount control, KDE versus Gnome, lack of a standardized widget set and environment for developers to develop on (part of the Gnome/KDE issue), poor performance, the patchy, hacked nature of X11 (band aid upon band aid), the fact that it was developed with engineers in mind and is now being attempted to sold to consumers, lack of any kind of ease of configuration for things like adding fonts, upgrading packages, utterly poor performance, even with extensive band-aids (it's network based, which is great for sysadmins, but consumers don't care).
I got tired of ugly fonts, poor performance, battling environments, the need for tinkering, and at the same time pretending it was somehow better than Windows, so I stopped using it.
You say we're deluding ourselves, but yet you do absolutely nothing to back up your claim.
I've done nothing but back up my claim, pointing out the various problems with X and the reasons they are affecting open source desktop proliferation. Now you back up yours.
So, you want to ditch X, and get to keep it, with an X server running on top of whatever it is you pick to replace X. Of course, now you have to worry about how X applications would integrate with your New'n'Spiffy (tm) windowing system. Don't forget that at least initially (say, first 5-8 years or so), most of the applications would want to use will be X applications, so desktop integration with X clients IS going to be an issue.
It's a simliar issue with Mac OS X and Mac OS 9. They made a decision to drop the old system in favor of the new. It was the right thing to do, the old system wasn't allowing Mac OS to evolve anymore and offer features that consumers wanted. But at the same time they knew they couldn't completely abandon the older Mac OS, so they made the Classic environment, which allowed them to run old apps on top of.
That's because your idea is misguided. When you casually suggest throwing away ~200000 LOC (for the server alone) and at least ten times as much code in libraries, clients, desktop gizmos, screensavers, drivers, video rendering and 3D services - you name it, I would expect YOU to come up with a reason why you think that would be a good idea, not the other way around.
All great leaps in innovation in the operating system world required abanding old code and old systems. It wasn't an easy choice, but it's one that needs to be at least discussed.
How many patches, shortcumming, and complicated installs, and leaps in technology on competing platforms. How many more excuses and scapegoats to blame for X's lack of proliferation on the desktop are going to be made before is enough.
Before enough is enough, and a clean break is need to be made.
Linux itself has several such milestones, such as LibC, VM code, packet filtering, and more.
Here comes the "X sucks" post again. What is wrong with X? Why break all compatibility just to ditch X? Why ditch X at all?
Here comes the "X doesn't suck" post again;)
1) There are no good alternatives. Period. DirectFB doesn't support nearly as many cards, and Berlin isn't even ready.
That's hardly ever been a reason not to make a change. If that was the reasoning years ago, open source and Linux and FreeBSD never would have come to be. There are several good candidates to choose from, personally I'd like to see something done with OpenBeOS. BeOS was a wonderfully full featured and elegant GUI, as well as very easy to use.
2) Network transparency. Some people claim that it's useless today but that's just false. It's still being used in corporate environments and it's becoming more and more important in the embedded market. If you want to create an alternative, it better be network transparent.
While that is useful for the corporate and geek environment, it is absolutely useless for a consumer. There are several alternatives to network transparency while still mainting the capability, such as making a GUI that is capable of being network transparent, but not by default.
3) X is proven. It's more than 15 years old now. Don't think X sucks just because XFree86 isn't the best implementation.
X is proven in the sysadmin and engineering environments. However, it has failed in the desktop realm.
4) X is extensible. Nearly all shortcomings can be worked around using extensions. Take a look at XRender for example. Or DRI. Or DGA. And in the near future: translucent windows, screen resizing and rotation (RandR or something).
Nearly all of X's shortcomings perhaps, but not all. Developers still have to choose whether to develop in KDE, Gnome, or X by itself, or develop their own widget sets (ala Mozilla). They have little idea of what type of environment a consumer would have.
5) X is fast enough. No X isn't slow. Moving windows doesn't seem to be smooth, but that's because of the communication between the window manager and the window, not because X is slow. When I switched to Metacity, moving windows suddenly became *a lot* smoother.
X by itself might be fast, but to get it to anywhere near the usability and even asthetic qualities of other GUIs, it becomes slow. Thus I still say "X is slow", or perhaps more accurately, "the X environment is slow".
Yes, X communicates through sockets. But locally, pixmaps (95% of all traffic) are transferred through shared memory (at least XFree86 does). CPUs are becoming faster and faster, so socket overhead should become smaller and smaller.
Of course, assuming that the driver is good and fast.
Just another example of how X was designed with a completely different set of requirements than those that apply to a consumer desktop.
6) XFree86 configuration is currently complicated. But that won't stay that way. Why ditch XFree86 and replace it with something new and incompatible when you can just improve XFree86? The developers are already planning on getting rid of XF86Config completely and go for hardware autodetection.
That's a major step, but how will it handle driver updates? Will grandma have to recompile her kernel?
While not totally X-related, the split between KDE and Gnome is only making things more difficult. Competition generally benefits all, but it's creating a rift between an already niche market.
If the desire is to keep X and open source desktops in the realm of the geek, then these steps are fine. But if there is really a desire to get them onto the desktop and bust the Microsoft monopolies, these flaws I've listed and many others need to be addressed by X or by a new GUI.
X for geeks is fine. Geeks like to play, tinker, putter. But for the consumer desktop, KDE or Gnome aren't even enough to make the environment
There are so many things wrong with X that it would take alot more than a slashdot post to list them.
Please, go ahead and list any benefits to X to the consumer, I've never heard anyone from the X camp list any, so if there are, I'd love to hear them. The only arguments I've heard thus far are either knee-jerk rhetoric damning my anti-X heresy, or benefits of X that regular computer users could care less about.
I saw take a step back, look at what we've got with X and Gnome/KDE, why it isn't working, and look at projects like those you listed and projects like OpenBeOS and see if things are still going in the right direction. As it stands, Microsoft is sweating bricks over open source operating systems on the server side, but they are laughing themselves silly at the failed attempts to get them onto the desktop.
I doubt it would be hard to do backwards compatability, either through a legacy API or through an X client running on top of the new GUI.
It's necessary only if you think Linux or other open source operating systems (heck, even commercial) should be widely used on the desktop.
Everytime I bring this up and people come up with knee-jerk reactions to defend X, I've yet to see anyone actually come up with a compelling reason why X is as good as or better than the other GUIs out there (Mac OS X, BeOS, MS Windows) as far as a consumer-oriented desktop goes. It's goes some great features if you're a system administrator or an engineer, but the regular user any benefits of X are worthless to them.
Linux and open source needs to just drop X as the GUI, and come up with something new. Something developed with the general consumer in mind. Something that doesn't require KDE or Gnome to patch it's many shortcomings.
Why are we all deluding ourselves into thinking that X11 and it's KDE/Gnome companions are viable desktop environments for consumers when they really aren't. It does the open source movement a disservice to constantly hype up an inferior platform while ignoring it's many shortcomings, simply because it says "open source" or "Linux".
I don't like Microsoft any more than you do, but it's a 5 year old product, and monopoly or not, I dont' think they or anyone should be compelled to fix a product they've long since retired, $billions or not. I also don't believe anyone should be compelled to do something with something they own.
While you can patch it yourself in open source, how many open source users actually have the ability and skill necessary to patch code? Especially if that product is on the more obscure side, do you round up the origional developers with some software gestapo and force them to fix it? That prospect would turn off quite a few potential open source developers as well as commercial.
They shouldn't be legally required to do anything, it might be bad practice, but it's their choice. Open source has no obligation to fix bugs, an open source project could get abandoned, and years later a fatal security flaw is discovered. Should the old developers be compelled to spend their valuable time fixing it if no one else has the time or ability to do so? I don't think so, software should be about freedom (as in free) and not perpetual servitude to the project.
Kahn really hates Kirk because he gets all the girls. I'm sure on Kahn's wife was all "isn't Kirk dreamy?" and "Did you hear? Kirk saved Earth again! That probably drove Kahn mad by itself, and when Ceti Alpha VI exploded, it probably only made him only slightly more irritated at Kirk.
"Kirk! You get all the women!"
"Kahn, I'm laughing at your feeble attempts to get the ladies."
I'm waiting for slightly higher resolution LCD displays to come out, such as 17-inch displays that do 1600x1200, instead of 1280x1024 (which isn't even 4:3).
My laptop has a 14-inch display at 1400x1050, and I love it. I love the extra space, and the fact that it's an LCD makes up for the small type.
I started using LCD displays a year or so ago and I don't want to go back. I do a majority of my time on computers reading and writing, and the crispness and sharp lines of LCD's square pixels as opposed to a CRT's round dots (although Sony's dots are square, making Trinitron the best CRT IMHO). It's just much easier on the eyes, and with anti-aliasing/cleartype on XP, and Mac OS X, it's even more pleasant to read.
I've got my 1280x1024 display at home with a Nvidia GeForce 2 running Quake 3 at 1280x1024 at about 90 FPS, and the pixel latency is so low on my Dell LCD, it's as good or better than CRTs for games, which are typically LCD's weakness.
This isn't the most comprehensive test of a disk's I/O, but it's easy and will help determine is something is potentially wrong.
Run this command from the Mac OS X shell:
dd if=/dev/zero of=testfile bs=1024 count=102400
This should make you a 100 MB file. dd on some systems will give you how long it took to write out the file in seconds (I don't have access to a Mac OS X box, just a FreeBSD box right now). If it doesn't, just put the time command in front of it:
time dd if=/dev/zero of=testfile bs=1024 count=102400
Devide the number of seconds it took to write the file out by 100, and you'll have a rough estimate of how many megabytes per second it wrote to disk.
Again, not the most comprehensive metric of course, but it gives you a rough idea. For an ATA system, it should probably do between 15 and 25 MB/s without too much trouble for a fast drive. The interface can be much faster (ATA/33, ATA/66, ATA/100), but the UFS and HFS+ file systems aren't optimized for sequntial writing, and/or there are other factors involved, so about 20 MB/s is a good mark I think.
We need subspace Ethernet. That would let us play UT with our Martian neighbors. I wonder if communicating via gravity would be possible. Gravitational waves may travel only at the speed of light, but the effect of gravity is instantaneous. If the Sun were to disappear right now, we wouldn't see it disappear for another 8 minutes, but the Earth's orbit would change immediately.
Also, the development environment for Linux GUI apps is precarious at best. When a developer takes a look at Linux, what choices do they have? They could develop their app on the basic X wiget set. It would be ugly, and therefor have almost no consumer appeal, but at least it would work on all X11 installs.
Or, they could develop for KDE and/or Gnome, but a developer, commercial or open source, wouldn't know if the user would have KDE or Gnome installed.
Or, they could go the Mozilla route, and write their own font and widget sets. That would take a while.
Either way, Linux and other open sources operating systems are a basket case for desktop development currently.
Linux on the desktop has failed for a great number of reasons other than Microsoft's influence and the lack of ports in general. An aging GUI (X11) for one, and the lack of any kind of semblence of a "plug and play" setup that the other commercial OSes have long enjoyed (Mac OS X, pre-X, BeOS, M$ Windows). That is what consumers demand in a desktop.
It's one thing to get a Linux box up and running and looking pretty for Grandma, but as soon as you try to deviate from the narrow scope of a Linux install (and other open source operating systems), such as changing resolution, adding a scanner, plugging in printer, Linux just doesn't compare. Not yet anyway.
As I recall, IBM doesn't have a license to run the AltiVec stuff, which is kind of a hinderance to deploy machines the IBM Power processors, but maybe that doesn't affect the Xserve machines. I wonder if they'll finally get a license or resolve that. I could be wrong though.
Slashdot Mirror
What is it about message boards that make normally civilized people turn into stereotypical razor-tounged neophytes? I guess the lack of needing to stare your debatee in the face removes all tracese of politeness. Snideness aside, I enjoy debaters like you, as it helps to refines one's arguments on both sides. Most people at this point have been "X rocks, you suck!" without substantiating their claims. I thank you for an intelligent debate (peppered with cheap insults as they may be).
Looked at it. Your argument is full of inaccuracies, (for example, the Mozilla developers did not "develop their own widget sets"), enough to make it completely nonsensical.
Mozilla under X has it's own font control, theme control, widget set independent of X and KDE/Gnome. It's GTK-based, but independent of Gnome or even KDE. It requires extra coding and extra bloat as opposed to developing to an environment you already know is there (Mac OS X, MS Windows).
And users just flocked to it! You're also missing the point: X is not a GUI. Repeat after me: X is not a GUI.
What's a GUI? Graphical User Interface. X is graphical, users use it, to wait for it, interface with things. Computer-related things I would imagine. You're splitting hairs on trivial matters to try to make your point look more valid. Old trick.
As per the other point, users have really flocked to the multitude of attempts at a Linux desktop, haven't they.
While that is useful for the corporate and geek environment, it is absolutely useless for a consumer.
Where's the proof? What is it that makes it "useless" Spit it out, buddy! Don't keep us guessing!
The "useless" referred to network transparency from the origional comment that I was commenting to. Millions of people use GUIs/environments/desktops (whatever you feel like calling it) without network transparency. Hence, it's useless to the average user. I agree, it's very useful in the corporate/sysadmin setting.
What gets X "near the usablitiy and even asthetic (sic) qualities of other GUIs" is software like KDE and GNOME. These are not X.
No, they aren't X. But there aren't any Linux desktop distributions that come without one or both. Without KDE or Gnome, X is unusable for consumers, even with a good window manager. Hence why I said "X environment". Perhaps "X-based environment" would be even more accurate. And thanks for pointing out my spelling error with that [sic], it also adds quite a bit of validity to your argument.
I can only guess how you draw your extraordinary conclusion that "[X] becomes slow". I'm assuming that you mean that KDE is slow, or that GNOME is slow. You'd be right, of course - they are slow.
Unfortunately, your failure to understand why they are slow leads you to the wrong conclusion. You can do much more to speed up KDE by fixing Linux's dynamic linking mechanism than by replacing X. In fact, you could replace X completely, port KDE to your spiffy new X replacement, and KDE would still be slow.
It sounds like you're the one with a failure to understand why they are slow. X isn't that efficient to begin with, it sacrificed speed for network transparency by choice because of the specific requirements to be flexible for things like X-terms and other engineering-related applications. Processor speeds are now fast enough that X by itself is relatively fast despite this inefficiency. But what consumer is going to use X with twm?
KDE and Gnome are both bloated attempts to get X to do what it wasn't designed to do, at least not efficiently. You can fix linking systems all you want, Linux's dyanmic linking system has nothing to do with it. It's still slow on FreeBSD, and plenty slow on other operating systems.
Take any other windowing system on the same hardware, on hardware years old, and X will always be slower when you include modern features such as widget/theme control, centralized font control, docs, etc. (that bascially require KDE and Gnome). BeOS, NeXTSTEP, Windows, Mac OS, Mac OS X, all beat X in the speed game.
What many others? And you haven't actually listed any flaws.
I've listed plenty of other flaws in plenty of other posts, but for your sake, I'll list a few:
The X windowing system was designed to a completely different set of specifications than what consumers want and need. It's got ugly fonts, lack of configuraiton, splintered distributions, splintered environments, splintered specifications, and more. These aren't problems in the server realm. They aren't even problems in the geek desktop realm, but they are issues with the consumer desktop realm. And that's what we're discussing here, isn't it?
Various distributions and projects are addressing some of these issues, but not all. You say X is slow really because KDE and Gnome are slow. Well, it's hard to imagine a consumer-grade desktop without one of them. So again, that makes it an X-based issue.
KDE and Gnome are at war and casualty is Linux on the desktop. Kill one off, and you've still got a slow desktop that is at least can be universally developed for. Not in-itself an X problem, but X's lack of features necessitates some sort of all-encompasing environment, which then does make it an X problem.
There is a lack of any kind of unifying software installation. Lindows took care of that, quite nicely in fact. But you're stuck with the Lindows distribution to get it. Other distributions have the same issue. More splintering. Not an X issue? Perhaps not directly, but as X doesn't have a unified graphical software installation specification, leaving a void that has never been adequately filled.
What's more, you acknowledge that whatever flaws you imagine exist in X can actually be fixed. In light of that, why ditch X?
Maybe they can fix X, but with more and more and more hacks, patches, and extra addons. You've already got half a dozen graphics libraries, layer upon layer of innefficiency piling up and slowing things down. In there end, there are so many add ons to X that aren't X or close to X, why use X to begin with? Because it was there?
I say stop the madness. X had a good run, it's great for engineers and sysadmins, but it's just too much to try to get it patched up to modern specs, and even then, it's slow. Why not just start over?
BeOS, NeXTSTEP, OS X, all fast, modern, elegant, graphical environments: Everything that X and Gnome/KDE are not. It's been done before, it can be done again, and with open source, be done even better.
You also say:
I got tired of ugly fonts,
Yet another hack on top of X
poor performance,
Ahh, but since X isn't any good as a consumer desktop without KDE or Gnome, it does become and X issue.
battling environments,
How is any distribution going to solve the Gnome versus KDE issue? You're not making any sense there.
the need for tinkering,
Distributions now install beautifully, but deveating from the narrow scope of any distribution will most likely require tinkering.
and at the same time pretending it was somehow better than Windows, so I stopped using it.
Finally You make a coherent point! I thought I'd never see the day.
I'm sure it was a lapse of stupidity on my part, I apologize.
The interesting question, for the rest of us, is "will the Linux desktop ever be good enough for the likes of Picky Tokki?"
Nope, the real question for those that care about the future of open source on the desktop is will there ever be a Linux desktop good enough for the likes of the general masses. Obviously there is nothing out there now, and I don't see anything, even fancy Red Hat and Lindows betas, coming close anytime soon. The entire setup of X/KDE/Gnome/extra hacks is incredibly cumbersome. Geeks might tolerate it because they've got the time and experience, but I doubt consumers ever will.
I've seen indications that some companies are determined to drag Linux, kicking and screaming, and place it square on the desktop. RedHat's most recent beta impressed the hell out of me, for example, precisely because even after installing it, I still have no idea what its XF86Config file looks like, or even where it is. Out of the box, the fonts look nice, the system administration tasks I had to do (Setting up networking with DHCP) didn't require dropping to the command line, most error messages made sense... At this rate, they'll have something solid in a couple of releases.
Indications? Are you referring to outright statements from CEOs and advocates to get Linux kicking and screaming onto the desktop? For years pundits have been claiming that mainstream consumer-grade Linux desktops are right around the corner, a single Red Hat beta away, but it's never come. Keep on hoping.
There is so much out there that's more efficient, elegant, and easy to use than the X/Gnome/KDE/Whatever solution. Let's at least explore moving forward in another direction.
Yet still Windows is a much more elegant and easy to use environment than X with KDE or Gnome. Issues like configuration, font control, performance, media, two completely different APIs built on even older APIs, dozens of required package installations and libraries, and overall plumbing are all areas where X is behind Windows. And that's not a compliment to Microsoft. When Windows was just coming out with Windows 95, there were other GUIs/windowing environments that can compete today even years ago.
Take NeXT. I worked at a company where everyone had a NeXT workstation, everyone from the receptionist to the sysadmins to the CEO. Despite their expense, the graphical environment on these workstations were fast, all on technology almost 10 years old.
Everyone loved using them, from the receptionist to the hardcore sysadmins. They were easy to use, easy to configure, and easy to manage and install applications. At the same time, they provided all the flexibility and power that us techies needed.
They ran on 68040's running at 25 Mhz; 33 Mhz if you were lucky, on a wopping 32 MB of RAM. Yet still, the NeXTSTEP environment was elegant, easy to develop for, powerful, and fast, and with 12-bit video when most everyone, except perhaps high-end Apple and SGI, was runing 8-bit video at most. Window movements, redraws, where all amazingly fast on antiquated hardware, and it still had all the bells and whistles of a modern GUI. You could drag a file from your home directory and place it in an email, and it'd attach automatically. This is common place today of course, but this was before Microsoft could even get Windows 95 out the door.
All of this was on a modified BSD 4.3 underpinning running on a Mach kernel. Everyone claims X is light and fast, but NeXTSTEP puts it to shame on the same hardware in both speed, functionality, and overall usability.
NeXT wasn't the only company with a powerful, easy, and elegant GUI around that time either, BeOS was also a system under the healm of Jean-Louis Gassee.
Both of those systems were doing full motion video, gaming, 3D, and even millions of colors years before X systems even cared or even Microsoft.
Now why can't the open source community come up with something like that? Why wouldn't a brand new windowing system, designed from the ground up, without all the complicated libraries, without the slow network-based plumbing, without the battling environments, without the technical baggage, with capabilities that other even non-MS GUIs came up years before, be good for getting open source on the desktop?
Thus far the arguments pro-X11 come down to "X is awesome you suck" which is hardly an argument, "there's nothing wrong with X" which I agree with if you're a techie, but if you're trying to get Linux and others on the desktop then I respectfully believe you to be wrong, and the argument that there is too much running on X today.
The only argument that I've heard thus far that I think holds merrit is that there are thousands of apps running on X currently. Like I've said, no great leaps in technology has come without a price, without abandoning old code and old ways. There is always resistence, everyone likes the status quo. X is the establishment in the open source world. Not because it's the best, but because it's the only thing out there right now. Remind you of any other situations?
Porting apps to a new environment wouldn't be all that difficult. Even mozilla, which is a daunting project itself, has been ported to Mac OS X's Aqua environment with Chimera. Port Mozilla, an instant messenger app like GAIM, OpenOffice, and a few other widely used X apps and you're halfway there to kicking Microsoft's ass.
Creating a legacy environment within the new environment wouldn't be that difficult, except possibly that the old environment was so complicated (X11 libs, KDE, Gnome, ugly fonts). Still, X can run atop BeOS, Windows, NeXTSTEP, and Mac OS X, making for a functional transition period.
So that issue has been addressed.
This isn't a very popular opinion, but slashdot is a place to discuss various opinions, or so I'm lead to believe.
Personally, I prefer Mac OS X and BeOS as my preferred desktops. They are elegant, clean, and easy to use, so no, Windows isn't my golden rule. It is, however, better than X in many, many, many respects.
But not in a way that consumers like.
2. It's universally supported by Unix desktop applications.
In other words, it's established. That doesn't mean it's good, and that certainly (and hasen't) meant it's an attractive alternative for the desktop.
Take a look at this response: http://slashdot.org/comments.pl?sid=40190&cid=4284 363
Bascially, my main beefs with X are lousy fount control, KDE versus Gnome, lack of a standardized widget set and environment for developers to develop on (part of the Gnome/KDE issue), poor performance, the patchy, hacked nature of X11 (band aid upon band aid), the fact that it was developed with engineers in mind and is now being attempted to sold to consumers, lack of any kind of ease of configuration for things like adding fonts, upgrading packages, utterly poor performance, even with extensive band-aids (it's network based, which is great for sysadmins, but consumers don't care).
I got tired of ugly fonts, poor performance, battling environments, the need for tinkering, and at the same time pretending it was somehow better than Windows, so I stopped using it.
Sorry, check my other posts in this semi-thread. There are numerous reasons and responses listed against reasons for X.
I've done nothing but back up my claim, pointing out the various problems with X and the reasons they are affecting open source desktop proliferation. Now you back up yours.
It's a simliar issue with Mac OS X and Mac OS 9. They made a decision to drop the old system in favor of the new. It was the right thing to do, the old system wasn't allowing Mac OS to evolve anymore and offer features that consumers wanted. But at the same time they knew they couldn't completely abandon the older Mac OS, so they made the Classic environment, which allowed them to run old apps on top of.
That's because your idea is misguided. When you casually suggest throwing away ~200000 LOC (for the server alone) and at least ten times as much code in libraries, clients, desktop gizmos, screensavers, drivers, video rendering and 3D services - you name it, I would expect YOU to come up with a reason why you think that would be a good idea, not the other way around.
All great leaps in innovation in the operating system world required abanding old code and old systems. It wasn't an easy choice, but it's one that needs to be at least discussed.
How many patches, shortcumming, and complicated installs, and leaps in technology on competing platforms. How many more excuses and scapegoats to blame for X's lack of proliferation on the desktop are going to be made before is enough. Before enough is enough, and a clean break is need to be made.
Linux itself has several such milestones, such as LibC, VM code, packet filtering, and more.
Here comes the "X doesn't suck" post again ;)
1) There are no good alternatives. Period. DirectFB doesn't support nearly as many cards, and Berlin isn't even ready.
That's hardly ever been a reason not to make a change. If that was the reasoning years ago, open source and Linux and FreeBSD never would have come to be. There are several good candidates to choose from, personally I'd like to see something done with OpenBeOS. BeOS was a wonderfully full featured and elegant GUI, as well as very easy to use.
2) Network transparency. Some people claim that it's useless today but that's just false. It's still being used in corporate environments and it's becoming more and more important in the embedded market. If you want to create an alternative, it better be network transparent.
While that is useful for the corporate and geek environment, it is absolutely useless for a consumer. There are several alternatives to network transparency while still mainting the capability, such as making a GUI that is capable of being network transparent, but not by default.
3) X is proven. It's more than 15 years old now. Don't think X sucks just because XFree86 isn't the best implementation.
X is proven in the sysadmin and engineering environments. However, it has failed in the desktop realm.
4) X is extensible. Nearly all shortcomings can be worked around using extensions. Take a look at XRender for example. Or DRI. Or DGA. And in the near future: translucent windows, screen resizing and rotation (RandR or something).
Nearly all of X's shortcomings perhaps, but not all. Developers still have to choose whether to develop in KDE, Gnome, or X by itself, or develop their own widget sets (ala Mozilla). They have little idea of what type of environment a consumer would have.
5) X is fast enough. No X isn't slow. Moving windows doesn't seem to be smooth, but that's because of the communication between the window manager and the window, not because X is slow. When I switched to Metacity, moving windows suddenly became *a lot* smoother.
X by itself might be fast, but to get it to anywhere near the usability and even asthetic qualities of other GUIs, it becomes slow. Thus I still say "X is slow", or perhaps more accurately, "the X environment is slow".
Yes, X communicates through sockets. But locally, pixmaps (95% of all traffic) are transferred through shared memory (at least XFree86 does). CPUs are becoming faster and faster, so socket overhead should become smaller and smaller. Of course, assuming that the driver is good and fast.
Just another example of how X was designed with a completely different set of requirements than those that apply to a consumer desktop.
6) XFree86 configuration is currently complicated. But that won't stay that way. Why ditch XFree86 and replace it with something new and incompatible when you can just improve XFree86? The developers are already planning on getting rid of XF86Config completely and go for hardware autodetection.
That's a major step, but how will it handle driver updates? Will grandma have to recompile her kernel?
While not totally X-related, the split between KDE and Gnome is only making things more difficult. Competition generally benefits all, but it's creating a rift between an already niche market.
If the desire is to keep X and open source desktops in the realm of the geek, then these steps are fine. But if there is really a desire to get them onto the desktop and bust the Microsoft monopolies, these flaws I've listed and many others need to be addressed by X or by a new GUI.
There are so many things wrong with X that it would take alot more than a slashdot post to list them.
Please, go ahead and list any benefits to X to the consumer, I've never heard anyone from the X camp list any, so if there are, I'd love to hear them. The only arguments I've heard thus far are either knee-jerk rhetoric damning my anti-X heresy, or benefits of X that regular computer users could care less about.
I saw take a step back, look at what we've got with X and Gnome/KDE, why it isn't working, and look at projects like those you listed and projects like OpenBeOS and see if things are still going in the right direction. As it stands, Microsoft is sweating bricks over open source operating systems on the server side, but they are laughing themselves silly at the failed attempts to get them onto the desktop.
It's necessary only if you think Linux or other open source operating systems (heck, even commercial) should be widely used on the desktop.
Everytime I bring this up and people come up with knee-jerk reactions to defend X, I've yet to see anyone actually come up with a compelling reason why X is as good as or better than the other GUIs out there (Mac OS X, BeOS, MS Windows) as far as a consumer-oriented desktop goes. It's goes some great features if you're a system administrator or an engineer, but the regular user any benefits of X are worthless to them.
Linux and open source needs to just drop X as the GUI, and come up with something new. Something developed with the general consumer in mind. Something that doesn't require KDE or Gnome to patch it's many shortcomings.
Why are we all deluding ourselves into thinking that X11 and it's KDE/Gnome companions are viable desktop environments for consumers when they really aren't. It does the open source movement a disservice to constantly hype up an inferior platform while ignoring it's many shortcomings, simply because it says "open source" or "Linux".
While you can patch it yourself in open source, how many open source users actually have the ability and skill necessary to patch code? Especially if that product is on the more obscure side, do you round up the origional developers with some software gestapo and force them to fix it? That prospect would turn off quite a few potential open source developers as well as commercial.
They shouldn't be legally required to do anything, it might be bad practice, but it's their choice. Open source has no obligation to fix bugs, an open source project could get abandoned, and years later a fatal security flaw is discovered. Should the old developers be compelled to spend their valuable time fixing it if no one else has the time or ability to do so? I don't think so, software should be about freedom (as in free) and not perpetual servitude to the project.
Kahn really hates Kirk because he gets all the girls. I'm sure on Kahn's wife was all "isn't Kirk dreamy?" and "Did you hear? Kirk saved Earth again! That probably drove Kahn mad by itself, and when Ceti Alpha VI exploded, it probably only made him only slightly more irritated at Kirk.
"Kirk! You get all the women!"
"Kahn, I'm laughing at your feeble attempts to get the ladies."
"Full impulse power! Damn you!"
That's pretty standard. I get that on my Linux box with SCSI even, also similar for FreeBSD.
Actually, the phonton isn't the carrier of magnetic force. I think they are L and Z particles, virtual particles, like the graviton, but I'm not sure.
I'm waiting for slightly higher resolution LCD displays to come out, such as 17-inch displays that do 1600x1200, instead of 1280x1024 (which isn't even 4:3).
My laptop has a 14-inch display at 1400x1050, and I love it. I love the extra space, and the fact that it's an LCD makes up for the small type.
I started using LCD displays a year or so ago and I don't want to go back. I do a majority of my time on computers reading and writing, and the crispness and sharp lines of LCD's square pixels as opposed to a CRT's round dots (although Sony's dots are square, making Trinitron the best CRT IMHO). It's just much easier on the eyes, and with anti-aliasing/cleartype on XP, and Mac OS X, it's even more pleasant to read.
I've got my 1280x1024 display at home with a Nvidia GeForce 2 running Quake 3 at 1280x1024 at about 90 FPS, and the pixel latency is so low on my Dell LCD, it's as good or better than CRTs for games, which are typically LCD's weakness.
The speed of gravitational waves and "the speed of propagation" could be two different speeds, although that's just theory. http://216.239.35.100/search?q=cache:uDPNhqhck6kC: research.spinweb.com/_tp/000001fe.htm+gravity+spee d+of+light+%22gravitational+waves%22&hl=en&ie=UTF- 8
No one has been able to test that gravitational waves travel at the speed of light, it's speed is only inferred through Einstein's general relativity.
I don't know how one would go about broadcasting and receiving something over the speed of propagation, but it's interesting to think about.
Run this command from the Mac OS X shell:
dd if=/dev/zero of=testfile bs=1024 count=102400
This should make you a 100 MB file. dd on some systems will give you how long it took to write out the file in seconds (I don't have access to a Mac OS X box, just a FreeBSD box right now). If it doesn't, just put the time command in front of it:
time dd if=/dev/zero of=testfile bs=1024 count=102400
Devide the number of seconds it took to write the file out by 100, and you'll have a rough estimate of how many megabytes per second it wrote to disk.
Again, not the most comprehensive metric of course, but it gives you a rough idea. For an ATA system, it should probably do between 15 and 25 MB/s without too much trouble for a fast drive. The interface can be much faster (ATA/33, ATA/66, ATA/100), but the UFS and HFS+ file systems aren't optimized for sequntial writing, and/or there are other factors involved, so about 20 MB/s is a good mark I think.
We need subspace Ethernet. That would let us play UT with our Martian neighbors. I wonder if communicating via gravity would be possible. Gravitational waves may travel only at the speed of light, but the effect of gravity is instantaneous. If the Sun were to disappear right now, we wouldn't see it disappear for another 8 minutes, but the Earth's orbit would change immediately.
I guess that's why everyone finds Linux GUI's so very easy, and use them instead of Apple or Microsoft.
Or, they could develop for KDE and/or Gnome, but a developer, commercial or open source, wouldn't know if the user would have KDE or Gnome installed.
Or, they could go the Mozilla route, and write their own font and widget sets. That would take a while.
Either way, Linux and other open sources operating systems are a basket case for desktop development currently.
It's one thing to get a Linux box up and running and looking pretty for Grandma, but as soon as you try to deviate from the narrow scope of a Linux install (and other open source operating systems), such as changing resolution, adding a scanner, plugging in printer, Linux just doesn't compare. Not yet anyway.
As I recall, IBM doesn't have a license to run the AltiVec stuff, which is kind of a hinderance to deploy machines the IBM Power processors, but maybe that doesn't affect the Xserve machines. I wonder if they'll finally get a license or resolve that. I could be wrong though.