Slashdot Mirror
XFree86 4.4 Released
puriots0 writes "XFree86 version 4.4 is finally out! Grab it while it's still hot, if you don't mind the recent licensing changes... And if you don't care about the license, but the maintainers of your distribution do, this might be the only way to get it for the moment." The XFree86 people seem very eager to claim that the new license is nothing bad; see their FAQ. However, people who have reviewed it, such as RMS and Branden Robinson, think differently. It looks as if the XFree86 people have a short timespan to either rethink their license changes or be dropped from every/almost every Linux distribution in favor of a forked codebase.
It prevents you including GPL-licensed code in an X server derived from XFree86; that is enough reason for Debian to avoid the new release, it seems.
Interesting that the FAQ now acknowledges that the 1.1 licence does not permit redistribution under the GPL; before, the XFree86 people were insisting that in their opinion it was allowed. Unless I'm confusing two different licences here.
-- Ed Avis ed@membled.com
No, from what I'm told, the main reason to fork is the attitude taken by some members of the XFree Core Team. As you said, its their code and they can do what they want but the forking has already happened:
Xouvert
Freedesktop
Cygwin X
Personally I don't see myself ever using XFree 4.4 and am looking forward to a complete release of fd.o. When that happens, I'll likely be moving everything I can off XFree but that's just me.
Freedesktop X server
y windows
PicoGUI and Berlin (or whatever they renamed themselves this week) are still in the development stage. GTK and some games can run under direct vga acess too.
Do you even lift?
These aren't the 'roids you're looking for.
there's the freedesktop.org xserver, as well as directfb, Y and other "non-x" things...
What other alternatives are there to Xfree?
There are not suitable alternatives for end-users on Linux and BSD on recent hardware. freedesktop.org is an experimental play-area for developers where exciting new features are currently being developed not mundane things like updated drivers for newer video cards (Radeon 9600, 3rd party 9200LE, newer Intel 845 series, etc.), not robust "production quality" software for end-users, Xouvert doesn't actual have any unique code of their own the last time I looked, and Y Window system is more an idea and a work in progress.
I think that what you're referring to is the Apache 2.0 license that had a clause about patents that some said conflicted with the GNU GPL but that the Apache Group said did not. I'm not sure, but I think the Apache 2.0 license has been revised further to make it more clear that it is GNU GPL compatible.
And the muscular cyborg German dudes dance with sexy French Canadians
Berlin was renamed Fresco quite a while ago, and has not changed names since.
You know where you are? You're in the $PATH, baby. You're gonna get executed!
XServer is an experimental project, based off of KDrive. It is -
1) An experimental driver architecture
2) An expertimental set of X extensions, dependent on the new driver architecture for performance reasons
3) The umbrella for the Keiths more mainstream extensions to X, including XDamage (which is a dependency for the compositing extension).
According to the guys on the XServer list, the XServer is not only not ready for prime time, but it may not ever be a real canidate for an XFree replacement because of it's experimental nature.
The nice thing about X is that it doesn't matter for your app which server is runngin...
Its called a 'standard' and its a nice thing to have.
The only thing you will have to choose is the library you compile your app against. Very vew apps use the xlibs directly. And even then the various forks of xlib are still compatible.
Jeroen
Secure messaging: http://quickmsg.vreeken.net/
All alternatives suffer from a lack of (accelerated) drivers.
DirectFB.
Because as they say over and over, the old license didn't directly address binary only distributions. The new one does.
T Money
World Domination with a plastic spoon since 1984
They will be moving to the alternatives once they are actually ready. For example, the freedesktop.org xserver hasn't had an official release yet, and isn't ready for general use. From what one of the Debian X Maintainers told me Debian will be moving completely over to freedesktop.org packages once they are ready for use.
The "new license" ONLY covers code which is (C) The XFree86 Project, Inc.
A lot of stuff, eg. some video drivers, come (and came) under different licenses.
Have you ever tried Mandrake? I don't use it normally (I use LFS or Gentoo), but its automatic hardware configuration is amazing. Unfortunately, its package management system kinda sucks.
LOAD "SIG",8,1
I think it is worth pointing out that back in 1998 The Open Group (now known as X.org) changed the licensing of X R6.4 to be proprietary, and only backed down when XFree86 and David Dawes explained exactly what they could with their proprietary server.
XFree86 is the reason we have a free software X server today. It is quite ironic that slashdot is now hating XFree86 because of licensing.
* Several stability issues with the support for the Intel 830M, 845G, 852G, 855M and 865G integrated graphics chipsets have been fixed. Some limitation related to the driver's use of the video BIOS remain, especially for some laptops.
...
* The nv driver for NVIDIA cards has been updated as follows:
* Support added to the nv driver for the GeForce FX 5700, which didn't work with XFree86 4.3.
* The driver now does a much better job of auto-detecting which connector of dual output cards the monitor is attached to, and this should reduce or eliminate the need for manual XF86Config overrides.
* The 2D acceleration for TNT and GeForce has been completely rewritten and its performance should be substantially improved.
* TNT and GeForce cards have new XvPutImage adapter which does scales YUV bit blits.
The SiS driver has seen major updates, including:
* Support for 661/741/760 and support for 330 (Xabre).
* Merged Framebuffer mode.
* Support for DVI, and much more.
* DRI for 300 series (300/305, 540, 630, 730) is supported again.
A new driver for several VIA integrated graphics chipsets has been added.
* The mouse driver has some support on Linux and FreeBSD for auto-detecting the device to use. This makes it unnecessary to supply this information in the XF86Config file in most cases.
* XFree86 4.4 supports IPv6, based on the code contributed by Sun Microsystems, Inc. to X.Org.
* The Mesa version used for OpenGL(R) 1.3 and DRI driver support has been updated to 5.0.2.
* FreeType 2 updated to version 2.1.4.
You probably mean the Y server, not X. And we're using Arch, not CVS.
Andy Suffield has been working on the project; he's got some stuff up at http://people.debian.org/~asuffield/.
By the way, the modifications to libiterm required to support Y have already entered Debian Unstable, so you don't have to install that seperately now.
#include "sig.h"
there's a nice article on OSnews which looks back on the whole XFree86 affair of past year ...
This post is displayed with recycled electrons
Uhhh, moving those tools out of the X11R6 subdir would break anything that expects those files to be there. Besides it's part of the Linux Standard Base and therefore not likely to be changed after all of the work that went into LSB.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
I just want a gui that works nice. What features are in the new X?
:).
X is a low-level windowing system, not a desktop environment like CDE, KDE, GNOME, (or twm
It's new features are support for newer video cards, bug fixes and work arounds for broken video cards (and Dell laptop BIOSes with regards to VESA modes and 845 chipsets), IPv6 support, new version of Mesa (OpenGL 3D support), and FreeType (font library).
indeed, it is very neat. the problem is that all the drivers will need to be rewritten/ported. you can't just write wrappers around the old drivers. right now we only have fbdev/vesa stuff and some other stuff you really don't want to use right now. getting the drivers finisched right now is more important then ever.
The average person doesn't need to know about xfs, or font dirs etc. Their distribution vendor takes care of this.
Again, the distribution vendor takes care of this.
Printing is not an X issue, for the same reason that embroidering is not. It's a completely different medium.
Again, the average user does not need to know about this. The distribution vendor chooses a default, and if the user has a reason to prefer a different one (which implies they already know about them), they may select a different environment. KDE software runs in GNOME, and vice-versa, and motif/CDE/XLib, etc software still runs in either, or neither, or twm, fvwm, etc.
XFree86 runs on a lot more platforms than just x86, which is a good thing now that 64 bit commmodity CPUs are coming out. Even AMD64 is likely to break those optimizations.
What does "safe mode" have to do with XFree86? Most of the users are on UNIX-like systems (Linux, FreeBSD, OpenBSD, Solaris, etc). If you want to run an X Server on Windows 95, you're free to try one of the commercial implementations.
Or even better, wait until there's at least one mirror accepting connections.. *sigh* no mirror seems to have been updated.. Here are the release notes.
torrent anyone?
According to Xouvert's home page a release is scheduled for April 1, 2004. Of course that could be some kind of april fool's joke....
Except that Microsoft does not use the BSD IP stack.
There's also Xfree86 4.4rc2 it's still old license and a good start to a fork.
The FSF doesn't like the new license and begins telling people that it is incompatible with the GPL. Does that mean they are right? Does the GPL mean whatever the FSF decides it means that week?
Well, the FSF did a very good job of backing up their claim that it is incompatible with the GPL. They pointed to the exact sections where the incompatibility occurs. Most people, including a number of the major Linux distributions, seem to agree with the FSF on this interpretation.
The "this week" comment is misleading. The FSF has never changed their opinion on the correct interpretation of the GPL and has tried to make the implications of this interpretation as clear as possible from day one.
More importantly, what exactly is the FSF supposed to do about people who don't agree with their current take and make use of Apache/XFree86-4.4 anyway? Are they going to sue them?
If someone alters a GPL program such that the code is linked with code from an incompatible license, the copyright holder of that code is within their rights to order the distributor of the mixed code to stop. If that distributor does not stop, the copyright holder is within their rights to sue the distributor to make them stop.
If the FSF were the copyright holder, they probably would sue if they absolutely had to in order to ensure compliance with the GPL. The FSF does hold the copyright on many open source projects, so this is a possibility.
Do they have the money, let alone the ability convince a court that the suit isn't frivilous?
Actually, yes, the main function of the FSF is to serve as a central copyright repository for open source and trust fund for the legal defense of those copyrights.
More importantly, can they afford the ill will that would result?
Can the open source software community afford to exist in a manner in which licenses are addressed in a slapdash, "oh that's close enough" manner? Since the nature of Open Source is to coordinate input from many contributors, the exact manner in which the rights and licensing to that input is marshalled is of extreme importance. The rights framework for open source needs to be clear and solid.
Moreover, I'm not sure exactly how much ill will the FSF would garner for enforcing the requirement that if you redistribute GPLed code you follow the terms of the GPL, especially since, well, if the GPL isn't going to be complied with then why does it exist in the first place? The only ill will this would generate is among people who don't think anything should be GPLed ever, and they hate the FSF anyway, so what's the loss?
Irritable, left-wing and possibly humorous bumper stickers and t-shirts
The only code that has changed license is the code Copyright Xfree86. Alan Cox, for example, refused to have his contributions under the new scheme. Individual contributors who retained copyright of their code can choose not to allow the new license. It doesn't solve the problem though.
It's not just FSF that has problems with this...
1 07 696705911864&w=2
OpenBSD project isn't including code with the new license either:
http://marc.theaimsgroup.com/?l=openbsd-misc&m=
You, my good man, are a troll.
There are two X servers on freedesktop.org. One is Keiths experimental server the other is the X.org tree which is XFree 4.4 without the license change bits and with other stuff, and most of the people Dave Dawes fired working on it.
The x.org server is very much ready for prime time
4.4rc2 is still old license
> freedesktops server is just build on xfree86. same drivers.
I think you've confused freedesktop with Xouvert. Xouvert is a fork from XFree86, but they don't appear to have any of their own code yet, and people have been questioning if the project is still alive. The freedesktop Xserver is an independent implementation, based on Keith Packard's Kdrive Xserver. From the freedesktop Xserver FAQ:
86ed
Can we get rid of the X11R6 subdir? (once again, stop thinking X is a world to itself)?
Yes. This is being done for the Freedesktop.org X server. See here for more information. And about time! Another thing is that the horrible imake system is being removed (I personally would like to see it taken out behind the barn and shot). Looks like these guys are serious about dragging X into the 21st century.
I think you're forgetting that the Y project was put back into motion. The goals of that project already overcome those of the current X. With enough effort the Y project could take over if needed.
Vic
The reason the two are incompatible is the fault of the XF86 people, as they are the ones making the recent change that caused the incompatability. If all was fine, then GPL suddenly changed and expected others to change to accomodate it, then any incompatability would be the fault of the GPL.
Is it true? I know about Theo being picky, but FreeBSD and NetBSD folks seem to be indifferent..
The underlying framework supports drop-shadows and translucent menus, but can be used for much more:
- OpenGL will be used for drawing, so you can have very rich, yet fast, vector-graphics based applications.
- Back-buffering each window (which is what enables transparency and shadows) allows you to eliminate all sorts of flicker and lag during resizing and window moving.
- New extensions (Composite + XDamage) allow for clever window managers to do useful things like Expose-style features, screen-scrapers (for things like VNC), magnification tools, etc.
A deep unwavering belief is a sure sign you're missing something...
The furor over the license is generated by commercial interests and little else. Look at where the criticism of the license comes from. Is it a vendor that has sold repackaged XFree86 as a product for years and stands to benefit from calling it their own product?
Once you filter out FUD from vendors, what issues are left?
Even RMS claimed that the new "license requirement qualifies as free software".
RMS also claims that the new licensing policy will "eliminate the (GPL) incompatibility with applications".
It seems pretty clear once you filter out the noise....
It appears to be pretty recent, and not yet advertised, but freedesktop.org has forked Xfree86 from 4.4 RC2. Note: this independent from their own experimental X server to which everybody is referring (but which is not really ready for consumption yet). If XFree86 doesn't revert to the old license, distributors are likely going to package the freedesktop fork. It remains to be seen if the major XFree86 developers will follow.
Re-introducing the Microsoft changes to the BSD code-bases would be distribution, and the BSD doesn't require derivative apps to allow free redistribution. Hence the whole difference between BSD and GPL licenses.
The EULA has nothing to do with it.
> There's also Xfree86 4.4rc2 it's still old license and a good start to a fork.
The folks at FDO have just done that.
you might find this useful...n KDE
http://wiki.debian.net/index.cgi?Debia
my blog
It is well known that you cannot freely mix BSD (old 4 clause or new 3 clause) licensed code with GPL code in the same code base.
No, it's not `well known', and in fact, it's not true!
The revised (`3 clause') BSD license is perfectly compatible with the GPL. Since 99.9% of all `BSD licensed' code uses the revised BSD license, there's basically no problem at all.
Of course the resulting aggregate program must have its source distributed (&c) as part of its source uses the GPL, but that's pretty obvious if you're using GPL'd source code.
[The original (`4 clause') BSD license is indeed incompatible with the GPL, but that's largely academic, as no one actually uses that anymore.]
We live, as we dream -- alone....
Right. Like GPL compatibility.
This is like 4-clause BSD, which was also incompatible with the GPL.
The GPL does not require this crediting -- therefore, any licence that does is not compatible.
They added an advertising clause.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Friends don't let friends enable ecmascript.
Eh? Elsewhere it's been said that drivers have to be rewritten for the freedesktop.org software, and I'm sure it has to be rewritten for Y. Graphics card makers are notorious for refusing to release the information required to allow open source drivers--suppose they refuse to write more than one version, and stick with XFree86? PCI Express will take over from AGP, if I read the chipset roadmap articles rightly just this morning...so if the manufacturers stick with providing binary-only drivers for XFree86, how will anyone, much less Granny, be able to use a free X?
Enlightenment (www.enlightenment.org) has been around for quite a few years now, and can look like a lot of different things - but if you give it 64 virtual desktops with a different background image on each and turn on all special effects it will run very slowly. If you use a sane configuration it will run OK on a pentium 90 with 32MB.
There are also many others which have been developed since then - there is more software available than what comes with your distribution.
One final thing - gnome is not X, it is a whole suite of different programs, which is why it takes so long to start up. The whole idea of there being one program that does everything is just an artifact of the Microsoft court case - your web browser is NOT part of your operating system, and the gnome panel is NOT part of X, they are seperate programs.
...that BSD and the new XFree86 licence require you to acknowledge the authorship and GPL forbids you from adding restrictions not in the GPL - including an advertising clause.
I would be happy to see the licence backed out and the major distributors voluntarily add a splash screen giving major credits and referring the viewer to a website with extended acknowledgements.
I would be just as happy to see a GPL "A" variant arise which was GPL plus advertising clause. This would allow you to GPLA-license BSD code which you are modifying, effectivly getting the GPL's sterner protection but without treading on the intentions and rights of the original authors.
Of course, if you're working from a no-advertising BSD licence, these problems evapourate. Personally, I'd like to see XFree86 with a splash screen since it'd give you something to read while KDE (or whatever) starts, and it'd overwrite the splash screen from NVidia (or other manufacturer) who are not at all shy about claiming credit for stuff.
Got time? Spend some of it coding or testing
I'm not going to upload KDE 3.2 to sid until KDE 3.2.1 is released. The KDE 3.2 release had so little testing done to it that they immediately released required interim fixes to kmail for it among other things like kwifi not even being compilable. There are already 10MB+ of diffs from KDE 3.2 -> KDE 3.2 BRANCH (3.2.1), 3.2 BRANCH is supposed to be just bug fixes so that should tell you 3.2 should be avoided. ;) You can always use the experimental (READ BUGGY UPSTREAM SOURCE) debs as mentioned on http://wiki.debian.net/DebianKDE.
Uh... you get to recompile DRM with every kernel upgrade, as well as ALSA, i2c etc and just about any other kernel module. nvidia's driver isn't unique in this.
Did you read the most recent ASF position on this? The matter is due to a misunderstanding how patents work under the ASL versus how they work under the GPL. The matter will probably not be completely resolved until there is a better understanding of software patents and/or a court case involving patents and these open source license.
One important point: GPL-compatibility was not the only "justification for the new license" by a long shot. That was one of many goals, but not the main point of the license.
So, there has been progress on this issue, but it's not as clear cut as you make it out.
Who said Freedom was Fair?
Linux is a kernel, not an OS. Calling an entire
system "Linux" is just incorrect. GNU/Linux is a
specific term which signifies the combination of
the GNU system and the Linux kernel, however, it
is also incorrect, as many of the core components
of a linux-based system (sysvinit, lilo, fsck,
modutils/module-init-tools) are not part of the
GNU system, but neither are they part of "Linux".
A bare-bones linux-based system is
generally comprised of GNU libc, GNU bash, GNU
fileutils, sysvinit, module-init-tools, fsck,
LILO (or alternately, GRUB), some sort of
packaging system (optional), and Linux itself.
in some cases, most of those tools are replaced
by something like busybox, in which case the
system contains no strictly GNU programs at all.
What part of that can be considered the operating
system? If you go by what is required to make
programs work, then it's just the kernel, a shell
(whether it be ash, bash, csh...), and a glibc.
If the program doesn't use the shell, that's not
needed, but it might. If the program is
statically linked, it doesn't need glibc, but
most programs aren't.
If you say "I use linux", you are correct.
If you say "my machine runs linux", that's right.
If you say "linux is a better OS than windows",
you might realize that most of what makes the
system what you would consider an OS is not,
in fact, part of Linux.
Am I missing something that you noticed?
Well, a lot of us have been waiting for that new VIA chipset driver. It's for an integrated video chip which is in a lot of those mini-PCs and also some desktops and laptops. Unfortunately, only a minimal version of the code for XFree86 4.3 exists (basic accelerated 2D, but lacking everything else -- see here).
I am a contributor and I am happy for my portions of the code (the Matrox drivers and documentation) to be relicensed under the new license. I've already told IBM that, and XFree86.
Why?
Because the GPL does not suit many forms of development, and if you wanted *ANY* drivers at all, you have to get the documentation. The competition between chipset manufacturers is extremely fierce, and they don't let just anyone access to their code.
Matrox, for example, provide you with a binary BIOS blob to work with, and you can't get this unless you sign an NDA. If the GPL was out there, the blob would become infected. Viola, no Matrox driver.
The fervour of some to be *all* GPL is misplaced and wrong. The GPL is not the be and end all of licenses, nor is itself perfect. The fact that it is deliberately broken with respect to other software in itself a flaw.
This license change will affect only contributors who wish to submit their code as GPL'd code rather than under the XFree86 license. They can always contribute their work to the driverless freedesktop.org, or they can give up a bit of license purity and actually get some work done.
And people wonder why Microsoft is winning.
Andrew van der Stock
Wrong. Look at this link:
http://wint.decsy.ru/du/DIGITAL/UNIXNT/CH2.HTM
And learn something about Windows NT system level programming and its relationships to BSD (including the differences in networking code).
The old one had four clauses, the fourth being the advertising clause. The new one has only three.
The old/original BSD license had 4 `clauses' of which one required you to mention Berkeley in any advertising/documentation for the end product. This `advertising clause' was, as you might imagine, not only a huge practical problem, but incompatible with the GPL.
The revised BSD license -- which almost all current `BSD licensed' software uses -- deleted the advertising clause, removing the conflict with the GPL.
We live, as we dream -- alone....
The oh-so-lame FTP client was always based on BSD. It's pretty gross code, and no one wanted to touch it. Only the minimum work was done to make it build and (sort of) run under NT.
The telnet client has a strange history. At one point (back in NT 3.1 and 3.5, I think) it was basically just the normal Windows terminal app. Someone had the bright (?) idea to have it talk to a virtual serial port. So, a driver implemented the virtual serial port, which talked to a user-mode helper service, which issued the socket calls to talk to the remote telnet server. Now, just imagine the ring transitions, imagine the context switches, imagine the buffer copies. And, of course, imagine the total lack of performance.
Later (NT 3.51 maybe?) the virtual serial port bullshit was ripped out, and the terminal app was hacked into being a (slightly less lame) telnet client.
For Win2K (maybe) and XP (for sure) a "new" telnet client was shipped. I've never seen the source to this version, so I do not know its origin. However, looking at its structure in a debugger, I would guess that it's not BSD based. It uses threads and async I/O, handles different language locales, etc. In other words, it appears very un-BSD like. But that's just my hunch.