Slashdot Mirror
XFree86 4.0 Now Available
YAH00 writes: "The 4.0 release of xfree is now available!!! I'm downloading it from ftp.xfree86.org as I type!!! " I've played around with the preview releases, and 4.0 looks to be a much needed improvement over the 3.3.x tree, with xinerama [?] features and improved performance for many graphic chipsets.
makedepend stalled at some point during make World. Killing the it off with ps was the answer, compiled without any probs.
Running it now. Mozilla actually looks right, the menu fonts are the correct size. Woohoo!!!
"In short: just say NO TO DRUGS, and maybe you won't end up like the Hurd people." --Linus Torvalds
http://www.au.xfree86.org/4.0/RELNOTES. html
Enjoy. Congrats and thanks to the XF86 team
Eric
(here's the introduction/installation instr. for those with mirror probs):
XFree86 4.0 is the first official release of the new XFree86 4. XFree86 4 represents a significant redesign of the XFree86 X server. It is very important to keep in mind that XFree86 4 is still very much in development, and it contains a lot of new work. That means two things: there is a lot of new exciting stuff to try, but being new code, it hasn't had nearly as much of a workout as the stable 3.3.x releases. If you're looking for a well-tested, stable release, and can't afford the inconveniences that new software can sometimes cause, then you are probably better off sticking with the 3.3.x releases for now. If you have the resources to try out the new version and investigate its features, or if you just like being on the bleeding edge, then please try 4.0!
This release isn't quite as complete as we would have liked. The main missing pieces are a nice configuration tool and support for some of the hardware that 3.3.x supports. The first point means that configuring the server might be more painful than usual. The second means that your hardware might not be supported by 4.0, or it might be supported at a lesser level (conversely, some hardware is better supported in 4.0). We've attempted to provide some information about the second point in our Driver Status document. Please check there first before trying 4.0. Unfortunately that document is still fairly basic, but it should at least give you an idea of whether you're likely to be able to use 4.0 at all or not.
On the subject of configuration, we have updated the basic text-based tool "xf86config" to generate config files in the format required by 4.0 (3.3.x config files won't really work with 4.0). We're also working on some other configuration tools, including one that is built-in to the X server. An early version of this is included in the release, and it works well for some hardware. To try it out, just run (as root) "XFree86 -configure". Both of these configuration options will at worst give you a reasonable starting point for a suitable configuration file. We've put some effort into documenting the 4.0 config file format, and you can find that information in the XF86Config manual page. Please check that and the driver manual pages and related documentation for further information about that.
Oh, another thing you might notice is that our documentation is rather patchy. Most of what is present should be in reasonable shape, but there are gaps. We thought it better to leave out docs that were very out of date rather than providing inaccurate and misleading information.
Finally, before you download and install the binary distributions for this release, please have a quick read through the Installation Document. It may save you some time. If those cautionary notes haven't turned you away (and we certainly hope not), please read on... The sections below describe some of the new features and changes between 3.3.x and 4.0. There is a lot of new stuff, and we definitely don't have enough space to cover it all here.
Want to work at Transmeta? Hedgefund.net? Priceline?
Can your IM do this?
ftp://download.sourceforge.net/ pub/mirrors/XFree86/
---
SourceForge Programmer Type - http://sourceforge.net
---
Drew Streib, dtype.org
ftp://soul.apk.net/pub/xf4/
;-)
enjoy until we get slashdotted!
Mike Roberto
- roberto@apk.net
-- AOL IM: MicroBerto
Berto
Does Xfree86 4 have support 2-d transperancy? Is such support necessary for pulling transperancy in X? One of the things I like about Mac OS X is that it can do slick things with tranparent windows and menus and such. I'd like to some wicked cool enlightenment theme with such trasperancies all over. Even more, I would like it to have some hardware support so that it will be fast.
So, does 4.0 do this? Does anyone even care? Or is this really in the domain of toolkits to handle?
Got HTML? Want LaTeX? Try html2latex
This is basically a problem in your browser. The ls command used by ftp servers has a policy of giving a year if and only if the file is more than x months old. I believe x is 6 for most FTP servers.
Your client, trying to be helpful, attempts to reconstruct the year by assigning a year in the last 12 months if the year is not given. This works well. Most of the time.
The analomous case is when the timezones differ so that the file looks like it's in the future to the client.
Fixing this is not hard, and I did forward it to the Mozilla team.
Moral of the story? Getting times and dates right is hard.
LILO boot: linux init=/usr/bin/emacs
If everyone keeps hitting the main FTP site instead of one of the mirrors, the stuff'll never replicate.
Here's a hint: Mirror List
æeee!
Welcome to the begining of the end for Linux.
:)
XFree86 4.0 represents the culmination of thousands of man-hours work. It is the peak of X windows development, and darn nifty software too.
It hosts a wide range of new and powerful features, and world class performance.
XFree86 4.0 means a lot of good things for a lot of people.
But, Xfree86 4.0 also means something bad, it means the end of the Linux system as we know it.
Not the Linux system as a piece of software, but as a powerful piece of Free Software.
Today, we enjoy the freedom of a completely free system. The linux system gives you freedom beyond what any complete server/desktop OS has ever brought you before. Not only the freedom to use, improve, and repair your system, but also the freedom to use it on common hardware.
A critical aspect of the freedom of the Linux System, that differentiated it from others before it is the ability to have freedom with all components of the system. As you can seldom completely diagnose, repair, or improve one part of a system without making changes to other parts. Thats part of what being a system means.
XFree has become a critical part of the Linux system.
Unfortunatly, many video card vendors wish to keep their source very closed. They feel that this gives them some competitive advantage, because often a smart driver can more then make up for dumber hardware. The fear that copyright is insufficent to protect their intrests, a fear well encouraged by the Microsoft enviroment most came from.
Keeping the source closed also helps them decieve the consumer. Closed drivers can help hide the real capabilities, bugs, and performance issues of their hardware. In the windows world, drivers often cheat to improve performance. Usually it doesn't effect quality very much, but people should have the right to know. In the low margin, highly competive video card market, the makers feel that every point counts.
In the past, Xdriver developers were effectivly prevented from producing binary only drivers because of the difficulity of tracking the codebase. This made them double-think the need for closed drivers. In almost every case, they decided that closed wasn't worth a little extra effort and freedom won out.
Now with XFree86 4.0, they no longer need to fear that. There is a VERY well written and modular API that should require little revision. When a revision happens, it will be VERY easy to track.
Several vendors are already working on binary only drivers for XFree86 4.0.
In the race to support every possible card at the highest speed, to compete with windows in certian markets (games, cad, etc), the XFree86 developers have decided to trade freedom and openness for a few more Xmarks or one more card with 3d support.
The XFree86 development has never been very open, it's developers could quite argueably be said to not understand the benifits of truely open development. In the future of binary only drivers, these gurus will probably have (NDAed) access to the code, they won't be affected personally.
Soon, most common hardware will not have open drivers. Linux developers will have to work harder to work around bugs in other peoples software. People on uncommon hardware, (alpha,ppc) or strange needs will usually be left out in the cold. You will no longer be the master of your computer, it's use will be contingent on your agreement to some Draconian UCTA enforced licence, your ability to improve, repair or hire someone else for that purpose will be signifantly reduced. A chain is only as strong as it's weakest link.
This is a terrible loss to ALL of us. In our race to compete with closed systems, we are giving up the one thing that makes Linux *better*.
Unfortunatly, most will never realize the loss. Years later, those same people who today revel in the 500fps brought out by their videoforce 1024's binary drivers will be wondering why Linux became what every closed system is today: Controlled, secretive, inflexible, restricted, and sometimes unstable.
Today, people are regulary ignored in bugreports on linux-kernel when they say they are running VMware. Privliged closed software makes it almost impossible to debug the kernel with any certanty.
Even Microsoft blames most of the windows crashes on buggy third party drivers, and I suspect MS would have much better luck at getting someones driver code then Alan Cox.
Please think twice before accepting a slightly superior closed solution. It's not really superior in the long term. While it may work for you today, someday *you* might be on the wrong end of the bargan. Please support your future needs, and tell the vendors that we wont take any more unnecessarily closed drivers.
Thanks for you time, sorry for the type-os (gotta post before I'm too far down to be read).
-greg@linuxpower.cx
BTW 4.0 finished patching for me right now, 3.9.18 works great, hope 4.0 is even better.
Bad fonts under Netscape is a solved problem };-)
While anti-aliasing isn't involved, I think you'll find these a significant improvement over X11's out-of-the-box look:
X: A Site for Sore Eyes
(go to the bottom of the page)
iSKUNK!
a) The linux kernel isn't any more open a development environment than XFree is. There are a couple of gurus with commit rights. Everyone else sends patches. Most of those are rejected. It's not a formal process of approval like becoming a FreeBSD developer. But you're equally free to fork the code in any of those cases -- the unwashed public always has full access to everyone's latest stuff.
:)
b) Hiding behind cruft is not a way to fight closed source drivers. They've been offered before, many times. And the response has been increasingly negative. I think we'll continue to write our own, open drivers, and write better ones. I want code I can review running with priviledges on my machine. How about y'all?
c) A cleaner driver model will make it easier to write Free drivers too.
d) Let's keep things positive here, huh?
e) "uncommon hardware, (alpha,ppc)" will rise to crush puny pathetic PCs once and for all
Since you brought it up...
/. years. It holds no water. You, as an obvious repeat AC, know of something called the "Slashdot Effect", it's the strange phenomenon of 100,000 info-crazy quadrapeds stressing the laws of physics and the capabilities of silicon based counting machines. Sometimes this "effect" breaks stuff. Chains being only as strong as the weakest link and all that, bottlenecks and such.
/..(nitpick that punctuation)
I've seen this lament a number of times in recent
So to bring it to a point, sometimes it is, in fact, "informative" to cut and paste some electrons, as such action makes this information more available to all, utilizing the vast resources of a billion dollar corporation who's mouthpiece is affectionately known as
I await your inevitable retort.(and my Karma protects me from flames like Goku's many Dragonballz)
--
+&x
Hopefully companies like Creative Labs will set a good example for others to follow. Creative released a closed-source driver for the soundblaster live. It worked (just) but MP3 playback was noticably worse than in Windows. They got a lot of flack from the Linux community over their refusal to open-source the driver and eventually gave in and released the source under the GPL. The soundblaster live driver suddenly improved beyond recognitions (thanks Alan & everyone else) and now plays MP3's better than the Windows driver on my PC.
If companies realise that there are real benefits to open-sourcing drivers, then they might just do so.
HH
Yellow tigers crouched in jungles in her dark eyes.
Yellow tigers crouched in jungles in her dark eyes.
She's just dressing, goodbye windows, tired starlings.
All this means is that we need to use a different means of keeping the pressure on. Personally, I'm *glad* I'll be able to have a little more flexibility in using binary drivers and at the same time I'm *sad* that this means of pressuring harware vendors to open their specs is now going to be weakened.
There are a few factors working in our favor:
The binary video driver api doesn't give hardware vendors cross-platform access - they wind up having to build and distribute drivers for every platform, multiplying their headaches and workload. This is work than can much more efficiently be done by the distributions and platform maintainers, including making necessary adaptations, for example, byte swapping - a much bigger issue than you'd think.
It doesn't give hardware vendors access to the power of open-source development, and the quality improvements resulting therefrom. Oh, and don't even *think* about trying to pass of a binary version of someone's open source driver as your own.
Closed-specs hardware vendors don't get the "coolness" bonus from, for example, us, the Slashdot community. Don't underestimate the effects of this: we've now become the "brand specifiers" for a huge part of the PC market, especially the games market. I'd say that we had a lot to do with 3Dfx's decline (because of closed-api concerns) and NVidia's rise (because they opened *part* of their specs).
The embedded market XFree is just too big and bloated for the embedded market - anybody care to argue this? Or for installing on old 486's and P90's. I know - I've tried it. We absolutely have to have an alternative, and there are already several projects underway on this. Let's not build in a binary driver api in these new video systems for at least 2 or 3 years, ok? That will keep the pressure on: if you want your card used in the embedded market (possibly much bigger than the desktop market) you'd better open the api. Do it sooner and get a bigger piece of the market. Do it later and become a historical footnote.
Life's a bitch but somebody's gotta do it.
I think your extreme attitude helps marginalize Linux rather than support it. A future where open and closed software intertwine (for various values of closed) is not to be feared. Yes, some vendors will gain temporary advantages through concealment, but you show little faith in the power of open source if you believe that its advantages will vanish as soon as such hybrids appear. Other vendors will go the open-source route, and profit from its advantages.
You mention several advantages of open-source, such as better stability and intregration. These are real, solid advantages that potentially give one vendor a leg up over another. In the face of this, it would be unlikely that no vendor would take advantage of those things. Many won't--old habits from the MSWindows domain will die hard--but a few will, and from their success more will be encouraged to do so.
Look at it from another perspective: we're going to get vendors involved in Linux who otherwise wouldn't be involved. Then we'll convert 'em. Not all of them, but the fact remains that opening up a middle ground like this gives us more of a chance of pulling them into the open-source way of doing things than an all-or-nothing attitude will.
Tear down the walls!
Speaking as someone who has been developing with NVidia's GLX Driver, I've been very disatisfied with their implementation. There have been numerous, significant bugs of note:
1. GL_SWAP_BYTES fails when loading textures. This has forced me to either:
a) make a copy of the texture, perform the endian conversion, load the texture and then free the texture-a significant speed hit
b) Load the texture as GL_AGBR_ext, which is not cleanly supported across all implementations of OpenGL
2. Reads from the framebuffer fail. I've tested this several times, under multiple applications including glQuake and my own engine.
3. The driver reports that SGIS_multitexture is supported in the extensions string. All calls to the SGIS functions result in a segfault. This has been tested in multiple applications, including glQuake and my own engine.
In summary, the commitment by NVidia to Linux has been nothing more than lipservice. I specifically purchased a TNT2 because it was supported under Linux. Had I known that the GLX support was so incomplete, I would not have purchased that card. I have emailed NVidia about these issues, but I have neither received a response, nor have the issues been addressed.
Well, Windows doesn't *really* antialias fonts - it just smooths them out a bit. Which isn't the same thing, really.
But that's not the real problem with Windows and font smoothing (even MS don't have the gall to call it "anti-aliasing").
The problem is that Windows doesn't smooth the fonts that really need smoothing - those at 10pt and below. Smoothing 72pt fonts, while it makes your PowerPoint slides look nice, isn't really all that important. To reap the real benefits of anti-aliasing (deceive the eyes that there's more spatial information than there really is) you need to anti-alias the small fonts. But smoothing looks very sucky at small point sizes, but it's relatively quick, which I guess is why most fonts on Windows don't smooth below about 10pt.
There's only one OS that I'm personally aware of that has ever done The Right Thing in anti-aliasing, and that was RISC OS on the Acorn Archimedes series of computers. That knew about proper sub-pixel antialiasing, and you could tune it quite precisely.
ClearType is more like "real" antialiasing, and I can't for the life of me work out why they didn't put it into Windows 2000.
But that's by the by. You asked about Linux. I don't think antialiasing fonts should be too much of an issue, at least at first. I think that having well hinted fonts, so that all three legs of a 12pt lowercase m only have one pixel each with 1 pixel spacing *every time*, for example, makes for better display. Poorly hinted fonts don't anti-alias well, either. (Although it has to be said that if you don't have any well-hinted fonts on the page, then the badly hinted ones tend to look OK when anti-aliased, at least until a decent font turns up:)
An example of badly hinted fonts is the URW-Fonts collection. They print fine, but they render grim.
The problem with the X font rendering code (and I just *know* I'm gonna get corrected here) is that it can't deal with glyphs as anything but bitmaps; that is, black'n'white. I understand that it would be a considerable rewrite to provide AA support in X.
--
Peter