MS moved to "Shared Source" sometime ago. It was done in hopes that Linux coders would borrow from MS. So far, it has not happened.
IIRC Microsoft released two or three packages under their Shared Source license, stuff like a new Windows installer based on XML files.
But this has potential to do what MS could not. Solaris is at least respected by the development world. This is simply another trap being laid by Sun and MS against Linux.
Solaris is just another operating system, released under an OSS license. Is *BSD a trap, being laid out by the DoD, just because the DARPA sponsored some parts of the development? It's really surprising, how paranoid people can get...
What is funny is how little ppl seem to remember from just 7 years ago. Sun actually ported to X86 once before AND "opened" their source code. Then when they thought that things were going well, they dropped it. If Sun ever feels like things are going in their favor, it is almost certain that they will do it again.
Bullshit. Sun delayed Solaris 9 x86, when they were already facing financial problems, not during the dotcom aera. And they finally noticed, that Solaris x86 could be much more than a nice-to-have for system administrators and the like.
Right now, they earn more money with x86 systems and Solaris x86 than ever before. And even if they deceide to drop Solaris x86, there's still OpenSolaris.
Well, why do you think, viral doesn't describe the paragraph 2.b) of the GPL adequately?
2.b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.
I don't question the way, GPL violations are handled at court today. But the GPL demands rights on your code which is problematic and made many open source projects either dual licensed (Mozilla, OpenOffice.org) or tagged with an exception (MySQL).
This is hardly a reasonable way to look at it. Many licenses are compatible with this, and the "viral" clause does not stop you from making the code available under any other license you like as well.
I think, it is a valid way. Do you remember the stories about the early days of the GNU project? It was basically RMS' anoyance that he was not granted access to a specific source code, so that he deceided, all source should be freely accessible for all hackers.
I can understand that Sun is afraid that dual-licensing would be one-way only: some code would be borrowed in other GPL projects and would not come back to Sun (or at least not to the Solaris kernel) because it could not be dual-licensed again. But if this is the only reason, then I think that it is short-sighted.
Hmm, I don't think so. Sun did exactly this type of dual-licensing with OpenOffice and there's a lot of code, which wasn't contributed back to OOo. So every distribution comes with its own build of OOo with it's own set of patches from the different sources applied.
I don't see the problem of direct code exchange between Linus und the Solaris kernel as a big one. They are technically very different, so code will have to be adopted massively. There's a site with Solaris drivers for several NICs. The author started by porting Linux drivers, but than switched to write the drivers from scratch and just have a look at the Linux source to find out how the hardware works. So he could license his drivers under a less restrictive and problematic BSD-style license.
Yes, the Apache license is also considered incompatible with the GPL, and the Apache webserver is still the dominant webserver. And I never heard RMS rant about that, and demand writing an Apache replacement under the GPL.
PHP is also licensed under an GPL-incompatible license. Still remember the mess with MySQL? When they relicened their client libraries from LGPL to GPL, because they wanted to urge more companies to buy commercial MySQL licenses, they had to issue an "Optional GPL License Exception for PHP". Otherwise they would have been replaced by many sites with competing projects like Postgres available under a more liberal license. Does this mess (and the commercial interest behind it) help free software or the community?
When Sun released OpenOffice under a dual license, some companies like Ximian or Redhat refused to license their changes under both licenses, they released them only under the GPL. So the patches found never their way to openoffice.org and today every distribution patches their own version of OpenOffice. Did this help open source or the community?
These are only a few points to show that the GPL is not the only OSS license. Sun has contributed a lot of code, so they have made probably more experiences with OSS licenses than most other people and companies. Is it really so hard to understand, that they see issues and don't simply hail RMS and the GPL?
How does that differ from RMS, who could be summarized like "GPL rulez! All other licenses suck!"
That's oversimplified? Yes, that's even more simplified than your "summary" of Schwartz speach. Maybe you should read the article inside the slashdot story.
Linux doesn't have a buggy awk, sed or tar. Solaris 8 does.
Linux had a broken tar for years; development stalled for years and just resumed mid 2004. Check the website.
There are much, much more broken things in Linux, my personal favorite is the NFS implemation, which got much better in the mean time, but still has some bugs.
Most x86 hardware doesn't suffer from the transient error bug that the non-ECC cache of the ultrasparcII processor.
Most UltraSPARC systems, either. Du you just want to repeat some FUD about Sun, you heard somewhere?
BTW: There is so much broken with x86 hardware, be it chipsets, graphics, ethernet or USB2 cards. It just works most of the time, maybe you have to use a patched driver.
I've played a little bit with USB2 in the last time. I've never ever seen so much broken hardware (it's not only the VIA VT6202) before, which needs chances in drivers, which make the driver unusable for some other devices.
Did you know, that Sun stopped the launch of a new workstation, because they found a bug in the ethernet chip? A chip, which already got delivered many times in x86 systems? No?
Don't get me wrong, I like sun hardware (Love LOM), but it and it's software are not perfect.
1) Linux is already ahead of Solaris on Intel hardware, not behind as this guy believes from reading Sun press releases.
Is it? My first filesystem tests showed almost par. ext3 is much slower on some tests than UFS, Reiser is a little bit faster. The GUI (JDS3) feels much more responsive than a Gnome desktop on Linux. Maybe the reason for that is the "interactive" process class of solaris, which Linux lacks. BTW: Windows has such a class, too.
2) Solaris is not known to be portable beyond Sparc, Sparc64 and ia32. ia64 and AMD64/x86-64 might happen but as far as I know don't yet exist.
Solaris 10 does include support for AMD64. And there have been ports of Solaris to Power (which got killed by IBM) and Itanium (which suffered from several reasons).
And how much do other platforms that x86 matter today? If there should be really demand in the OpenSolaris community, they can hack support for PowerPC, ARM or whatever they want.
3) Sun has yet to announce a license for Solaris, it is very doubtful it will be actual Open Source and almost certainly not Free Software in the FSF sense of the term.
OpenSolaris is not the same as Solaris. Yes, there are still the gory details left. But there is already a closed "beta test", and the license will be OSI compliant.
4) Sun is almost certain to keep parts totally closed due to licensing terms with third party suppliers.
We'll see. From what I've heard from people which already have access to OpenSolaris, it looks pretty complete.
5) Sun will rig things to retain ALL creative control from the Java experience. This will preclude any sort of community involvement on the scale needed to compete with Linux.
We will see, how much community interest Sun will get. Interest in OpenOffice seems to be quite good. At least much better than other community efforts like Koffice.;-)
Where do you see "more headroom to grow" for IA-64 than x86 in it's current version AMD64/EM64t? x86 has outgrown any other architecture in market volume and most architectures in life time.
I don't know, if Intel and HP really believed to predict parallelism in the compiler. From my point of view this looks naive. No compiler can do that, you have to rewrite the code to be multithreaded. And right now, Sun, Intel, AMD and others are right now working on massive multithreaded chips for that.
The Itanium is a high-end workstation/server chip. ONLY.
If you read older articles from the times when Itanium was still Merced, Intel pretended they wanted to replace the old x86 line with the new IA-64 processors in the long term. The big irons (and workstations) have been only the first step in this plan.
Would be interesting to know, if Intel still hopes to see this coming true some day, or if they have already buried those hopes completely.
1) Linux is pretty darn good. It would take some *unusually* serious needs before you *have* to look outside the Linux camp to find a workable solution. Linux has XFS, JFS, and ReiserFS, really good support for reliable and fast high-end SCSI, SMP, Beowulf'ing, and a huge community to provide free-as-in-beer help.
Solaris has UFS, which also has logging and many decent screws to tweak. Features like Quotas, ACLs work without the slightest problem on Solaris. The same can be a PITA with Linux.
(BTW: Let's see if the Linux crowd has real arguments against ZFS (which comes with Solaris 10) or just whinig that Sun is evil.;-)
Clustering (for HA, not for compute tasks) is much more advanced on any decent Unix like AIX or Solaris than with Linux.
The SCSI subsystem in Linux is a joke, and it seems to be even worse with 2.6.
2) On a per-processor basis, Linux-on-Intel/PowerPC is faster than Solaris-on-anything hands down. (This will probably change after the next generation of Sparc chips comes out.)
Ah, interesting, but some lines forward, speed doesn't matter any more?:-)
3) Solaris tends to be a pain to port code to. Much like AIX, it's got the AT&T-derived libraries and proprietary crud that doesn't function with as much polish as the GNU stuff. So you end up installing a huge set of GNU tools and libraries on Solaris and... geez by this time you've almost got GNU/Linux again on Sun hardware. AIX 5L has at least started to reverse the trend -- you can get most of the GNU tools pre-installed. (Yes, the native compiler on Solaris and AIX produces much faster code than gcc. Most of my apps don't need the speed, they need the portability. I can optimize at the higher layers and get the speed I need.)
Does speed matter or doesn't it?;-)
Besides that, Solaris used to be a much better system to port software to or from other Unix systems than any other. AIX is maybe not the worst one, but many system features work pretty much different than on other Unix branded systems.
Isn't that the guy, who didn't manage to install JDS on four different PCs? Not a great base for a article about JDS.
I also like what he writes about himself: Using subtle clues and hints in his first-person narrative to imply emotion and intention, Jem Matzan's critically acclaimed writing style is truly unique among fiction authors. Jem's extraordinary characters and distinct dialogue decorate his fantasy universe while coaxing readers' imaginations into providing the specifics.
Even if he managed to get his JDS article published on linux.com, there are better articles on this topic there. This one is about JDS1, but because the changes happend under the hood, it's still valid for JDS2.
Furthermore, this creates a OSS project that now directly challeges Outlook
Hmm, I don't think that this decision was made to attack Outlook. IMHO Novell had primarily the open source competition in mind.
First, there's Sun. With their JDS they are the competitor with the most similar product.
From what I hear from our Sunnies, cooperation with Suse on JDS is okay, but with Ximian it's full of problems:
Evolution is bug ridden beyond all bearind, but Ximian seems to make it as painfull to submitt patches as possible.
And - for a company like Sun very important - Ximian is not willing to provide Sun with a Exchange Connector binary compatible with JDS. Sun staff has to apply a hack to get the Exchange Connector running.
Finally Ximian is very eager to propagate Mono, especially to Evolution and to Gnome. Must I say that Sun as the company behind Java doesn't like.NET?
So Sun is alread working to kick Evolution out of their JDS and replace it with something homegrown called Glow. It's written in Java, so it will have to proove that Java has grown to a viable way to write desktops applications right now. But Glow will have it's own Exchange connector. If selling the Evolution Exchange connector to sun was considered a way to earn money at Ximian (maybe like Suse cooperates with Sun on the basic Linux packages), they can forget this right now. And it's their own fault.
Another competitor in this field is IBM. They just announced a framework for cross platform, thin client applications. Even if they are very interested in Linux and Open Source, they are more interested in cross plattform solutions than in replacing their plattforms with a new, open source platform like Linux. Don't forget, that they invested very much in Java and still have a strong business with Windows PCs and their propriety big iron server lines.
They don't have a direct competing product to Evolution and the Exchange connector. But they are probably the only company big enough to force customers to switch to their new Workplace plattform. And I don't think Novell is that naive to believe, the groupware client in this new Workplace environment will be Evolution.
So, that's just the two biggest competitors, there are some more competing companies. But if Novell want's to survice in this business, they will have to keep the pace with these two. And open sourcing their Exchange connector is a very good way to attract more people to their own product.
Okay, the Adobe DPS part should be a real problem, some drivers probably, too.
But a Sunnie told me some time ago, that they are currently investigating how to get rid of Xsun and get back closer to something more popular like XFree86 or the new X.org X11R6.7.
I didn't see another benefit that getting rid of all the legacy stuff from many years, but it would also make perfectly sense, if you want to open source Solaris, or at least the vital parts.
We want a herd of cats, not a flock of sheep, to help prevent the single points of failure that results from widespread use of a single distro.
I agree to that, but I think that the big IT companies like IBM, HP or Sun know of this danger.
If you compare the Linux companies like Redhat or Suse to the IT giants, the latter are still much bigger. They can and will help to keep several competing Linux companies alive. Not only for their customers benefit, but also for their very own. They know, how worse things can get, if you have to deal with a monpolist like Microsoft.:-)
Thanks for that great article, I've just two additions:
Linux "monoculture"
I don't think, that there is a big thread for Linux to become a monoculture or proriety. (This would be the thread, Jonathan Schwartz was bashed here a few days ago, right?). There are several different distributions, and I doubt that Redhat will become too dominant. The bigger corporations like IBM or HP will be aware, that there are different flavors.
Can Sun legally open-source all of Solaris?
I think they can, they've bought very extensive rights about SVR4 from AT&T years ago. And they got based for paying SCO some money some time ago. So I expect they have all the rights to open source Solaris, at least the SVR4 parts.
And it would be a huge slap into the face of SCO and their cruise against Linux.:-)
PS: Don't get me wrong, this is no critics! I think, she spent much time with Groklaw and she did a really good job. so it's much better to get her into another project than to "loose" her, going back to her studies, an office job, or what she did before Groklaw.
This is a new site, not on Groklaw itself, and it is a community project, not just PJ.
But she is starting it, and it doesn't surprise me, that she is starting this right now. If Baystar want's it's money back, SCO could be history pretty soon. And then there is no more big interest in Groklaw.
This is mostly about consolidating control of not only the data center but the desktop within corporations. LINUX is making a lot of headway - probably faster than Sun & Microsoft ever imagined.
Companies are not yet ready to migrate all of their desktops to Linux. But they are thinking about it, and the most promising place are desktops with a narrow set of applications, like call centers, POS, etc.
And this is exactly the target market, Sun tries to catch with their JDS (java desktop system) products, mostly based on open source.
The biggest problem for a migration to Linux away from Windows is integration with the existing Micrsoft ecosystem: Office, Exchange, etc... Sun already has answers to some of these, but a cooperation with Microsoft could provide them the rest of the stack.
So they can provide customers with an alternative to Windows on the desktop, which may be nothing for true Linux advocates, but interesting to companies with less religious ambitions.
Think about this, if Linux does totally marginalize Sun (like SCO is now) that means Linux has moved onto the big iron.
Even if Linux is successfull on big irons, this doesn't make Sun redundant. The interesting part for big irons is not the operating system, but also the hardware, the service and the know-how how to run those boxes. And Sun has excactly this kind of know-how.
Microsofts biggest fear is the Linux on small servers and the desktops. Microsoft was very successfull with Windows NT on small servers, because in combination with cheap x86 hardware it was cheaper than the more expensive Unix/Risc boxes.
But Linux can be cheaper than windows, this is the biggest problem for Microsoft. They can try to argue that it's easier or cheaper to develop for Windows or to administrate it, but I doubt that this works.
A second strategy for Microsoft would be to rely on patents and "interlectual property". And a partner in the Unix/Linux camp like Sun, whose software does integrate nicely with theirs, could help them very much.
Slashdot Mirror
MS moved to "Shared Source" sometime ago. It was done in hopes that Linux coders would borrow from MS. So far, it has not happened.
IIRC Microsoft released two or three packages under their Shared Source license, stuff like a new Windows installer based on XML files.
But this has potential to do what MS could not. Solaris is at least respected by the development world. This is simply another trap being laid by Sun and MS against Linux.
Solaris is just another operating system, released under an OSS license. Is *BSD a trap, being laid out by the DoD, just because the DARPA sponsored some parts of the development? It's really surprising, how paranoid people can get...
What is funny is how little ppl seem to remember from just 7 years ago. Sun actually ported to X86 once before AND "opened" their source code. Then when they thought that things were going well, they dropped it. If Sun ever feels like things are going in their favor, it is almost certain that they will do it again.
Bullshit. Sun delayed Solaris 9 x86, when they were already facing financial problems, not during the dotcom aera. And they finally noticed, that Solaris x86 could be much more than a nice-to-have for system administrators and the like.
Right now, they earn more money with x86 systems and Solaris x86 than ever before. And even if they deceide to drop Solaris x86, there's still OpenSolaris.
Without any details, this sounds much more like Microsoft's "Shared Source" than like being something clearly identified as F/OSS.
I miss so important topics like the used license(s) or which software packages are available.
If I take in mind, that Sun released OpenSolaris today, this interview sounds to IBM want's to draw attention away from that.
The GPL is not viral.
Well, why do you think, viral doesn't describe the paragraph 2.b) of the GPL adequately?
2.b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.
I don't question the way, GPL violations are handled at court today. But the GPL demands rights on your code which is problematic and made many open source projects either dual licensed (Mozilla, OpenOffice.org) or tagged with an exception (MySQL).
This is hardly a reasonable way to look at it. Many licenses are compatible with this, and the "viral" clause does not stop you from making the code available under any other license you like as well.
I think, it is a valid way. Do you remember the stories about the early days of the GNU project? It was basically RMS' anoyance that he was not granted access to a specific source code, so that he deceided, all source should be freely accessible for all hackers.
http://www.gnu.org/gnu/thegnuproject.html
I can understand that Sun is afraid that dual-licensing would be one-way only: some code would be borrowed in other GPL projects and would not come back to Sun (or at least not to the Solaris kernel) because it could not be dual-licensed again. But if this is the only reason, then I think that it is short-sighted.
Hmm, I don't think so. Sun did exactly this type of dual-licensing with OpenOffice and there's a lot of code, which wasn't contributed back to OOo. So every distribution comes with its own build of OOo with it's own set of patches from the different sources applied.
I don't see the problem of direct code exchange between Linus und the Solaris kernel as a big one. They are technically very different, so code will have to be adopted massively. There's a site with Solaris drivers for several NICs. The author started by porting Linux drivers, but than switched to write the drivers from scratch and just have a look at the Linux source to find out how the hardware works. So he could license his drivers under a less restrictive and problematic BSD-style license.
It's not GPL-compatible.
Yes, the Apache license is also considered incompatible with the GPL, and the Apache webserver is still the dominant webserver. And I never heard RMS rant about that, and demand writing an Apache replacement under the GPL.
PHP is also licensed under an GPL-incompatible license. Still remember the mess with MySQL? When they relicened their client libraries from LGPL to GPL, because they wanted to urge more companies to buy commercial MySQL licenses, they had to issue an "Optional GPL License Exception for PHP". Otherwise they would have been replaced by many sites with competing projects like Postgres available under a more liberal license. Does this mess (and the commercial interest behind it) help free software or the community?
When Sun released OpenOffice under a dual license, some companies like Ximian or Redhat refused to license their changes under both licenses, they released them only under the GPL. So the patches found never their way to openoffice.org and today every distribution patches their own version of OpenOffice. Did this help open source or the community?
These are only a few points to show that the GPL is not the only OSS license. Sun has contributed a lot of code, so they have made probably more experiences with OSS licenses than most other people and companies. Is it really so hard to understand, that they see issues and don't simply hail RMS and the GPL?
How does that differ from RMS, who could be summarized like "GPL rulez! All other licenses suck!"
That's oversimplified? Yes, that's even more simplified than your "summary" of Schwartz speach. Maybe you should read the article inside the slashdot story.
Linux doesn't have a buggy awk, sed or tar.
Solaris 8 does.
Linux had a broken tar for years; development stalled for years and just resumed mid 2004. Check the website.
There are much, much more broken things in Linux, my personal favorite is the NFS implemation, which got much better in the mean time, but still has some bugs.
Most x86 hardware doesn't suffer from the transient error bug that the non-ECC cache of the ultrasparcII processor.
Most UltraSPARC systems, either. Du you just want to repeat some FUD about Sun, you heard somewhere?
BTW: There is so much broken with x86 hardware, be it chipsets, graphics, ethernet or USB2 cards. It just works most of the time, maybe you have to use a patched driver.
I've played a little bit with USB2 in the last time. I've never ever seen so much broken hardware (it's not only the VIA VT6202) before, which needs chances in drivers, which make the driver unusable for some other devices.
Did you know, that Sun stopped the launch of a new workstation, because they found a bug in the ethernet chip? A chip, which already got delivered many times in x86 systems? No?
Don't get me wrong, I like sun hardware (Love LOM), but it and it's software are not perfect.
Isn't UltraSPARC II Sun hardware, too? ;-)
When Sun opensourced OpenOffice, people said, this will help to improve Koffice. Where's OpenOffice today and where is Koffice?
And the GPL is not a requirement for a big developer base. Have a look at Free/Net/OpenBSD, XFree/Xorg, Perl, and many others.
1) Linux is already ahead of Solaris on Intel hardware, not behind as this guy believes from reading Sun press releases.
Is it? My first filesystem tests showed almost par. ext3 is much slower on some tests than UFS, Reiser is a little bit faster.
The GUI (JDS3) feels much more responsive than a Gnome desktop on Linux. Maybe the reason for that is the "interactive" process class of solaris, which Linux lacks. BTW: Windows has such a class, too.
2) Solaris is not known to be portable beyond Sparc, Sparc64 and ia32. ia64 and AMD64/x86-64 might happen but as far as I know don't yet exist.
Solaris 10 does include support for AMD64. And there have been ports of Solaris to Power (which got killed by IBM) and Itanium (which suffered from several reasons).
And how much do other platforms that x86 matter today? If there should be really demand in the OpenSolaris community, they can hack support for PowerPC, ARM or whatever they want.
3) Sun has yet to announce a license for Solaris, it is very doubtful it will be actual Open Source and almost certainly not Free Software in the FSF sense of the term.
OpenSolaris is not the same as Solaris. Yes, there are still the gory details left. But there is already a closed "beta test", and the license will be OSI compliant.
4) Sun is almost certain to keep parts totally closed due to licensing terms with third party suppliers.
We'll see. From what I've heard from people which already have access to OpenSolaris, it looks pretty complete.
5) Sun will rig things to retain ALL creative control from the Java experience. This will preclude any sort of community involvement on the scale needed to compete with Linux.
We will see, how much community interest Sun will get. Interest in OpenOffice seems to be quite good. At least much better than other community efforts like Koffice.
Where do you see "more headroom to grow" for IA-64 than x86 in it's current version AMD64/EM64t? x86 has outgrown any other architecture in market volume and most architectures in life time.
I don't know, if Intel and HP really believed to predict parallelism in the compiler. From my point of view this looks naive. No compiler can do that, you have to rewrite the code to be multithreaded. And right now, Sun, Intel, AMD and others are right now working on massive multithreaded chips for that.
The Itanium is a high-end workstation/server chip. ONLY.
If you read older articles from the times when Itanium was still Merced, Intel pretended they wanted to replace the old x86 line with the new IA-64 processors in the long term. The big irons (and workstations) have been only the first step in this plan.
Would be interesting to know, if Intel still hopes to see this coming true some day, or if they have already buried those hopes completely.
Don't worry, the next JDS release will include Mozilla 1.7 (and Gnome 2.6 and many other new versions).
As legend has it, once a customer asked his Sun sales rep. "You know, IBM is very busy with Java. Do you do something with Java, too?"
;-)
Maybe Sun's marketing heard this, too, when they thought about a new brand name for their software products.
1) Linux is pretty darn good. It would take some *unusually* serious needs before you *have* to look outside the Linux camp to find a workable solution. Linux has XFS, JFS, and ReiserFS, really good support for reliable and fast high-end SCSI, SMP, Beowulf'ing, and a huge community to provide free-as-in-beer help.
Solaris has UFS, which also has logging and many decent screws to tweak. Features like Quotas, ACLs work without the slightest problem on Solaris. The same can be a PITA with Linux.
(BTW: Let's see if the Linux crowd has real arguments against ZFS (which comes with Solaris 10) or just whinig that Sun is evil. ;-)
Clustering (for HA, not for compute tasks) is much more advanced on any decent Unix like AIX or Solaris than with Linux.
The SCSI subsystem in Linux is a joke, and it seems to be even worse with 2.6.
2) On a per-processor basis, Linux-on-Intel/PowerPC is faster than Solaris-on-anything hands down. (This will probably change after the next generation of Sparc chips comes out.)
Ah, interesting, but some lines forward, speed doesn't matter any more? :-)
3) Solaris tends to be a pain to port code to. Much like AIX, it's got the AT&T-derived libraries and proprietary crud that doesn't function with as much polish as the GNU stuff. So you end up installing a huge set of GNU tools and libraries on Solaris and ... geez by this time you've almost got GNU/Linux again on Sun hardware. AIX 5L has at least started to reverse the trend -- you can get most of the GNU tools pre-installed. (Yes, the native compiler on Solaris and AIX produces much faster code than gcc. Most of my apps don't need the speed, they need the portability. I can optimize at the higher layers and get the speed I need.)
Does speed matter or doesn't it? ;-)
Besides that, Solaris used to be a much better system to port software to or from other Unix systems than any other. AIX is maybe not the worst one, but many system features work pretty much different than on other Unix branded systems.
Sun bought extensive rights on Unix from USL in 1994.
If any company has the rights to open source something based on/derived from SVR4, it's Sun.
Isn't that the guy, who didn't manage to install JDS on four different PCs? Not a great base for a article about JDS.
I also like what he writes about himself: Using subtle clues and hints in his first-person narrative to imply emotion and intention, Jem Matzan's critically acclaimed writing style is truly unique among fiction authors. Jem's extraordinary characters and distinct dialogue decorate his fantasy universe while coaxing readers' imaginations into providing the specifics.
Even if he managed to get his JDS article published on linux.com, there are better articles on this topic there. This one is about JDS1, but because the changes happend under the hood, it's still valid for JDS2.
Furthermore, this creates a OSS project that now directly challeges Outlook
.NET?
Hmm, I don't think that this decision was made to attack Outlook. IMHO Novell had primarily the open source competition in mind.
First, there's Sun. With their JDS they are the competitor with the most similar product.
From what I hear from our Sunnies, cooperation with Suse on JDS is okay, but with Ximian it's full of problems:
Evolution is bug ridden beyond all bearind, but Ximian seems to make it as painfull to submitt patches as possible.
And - for a company like Sun very important - Ximian is not willing to provide Sun with a Exchange Connector binary compatible with JDS. Sun staff has to apply a hack to get the Exchange Connector running.
Finally Ximian is very eager to propagate Mono, especially to Evolution and to Gnome. Must I say that Sun as the company behind Java doesn't like
So Sun is alread working to kick Evolution out of their JDS and replace it with something homegrown called Glow. It's written in Java, so it will have to proove that Java has grown to a viable way to write desktops applications right now. But Glow will have it's own Exchange connector. If selling the Evolution Exchange connector to sun was considered a way to earn money at Ximian (maybe like Suse cooperates with Sun on the basic Linux packages), they can forget this right now. And it's their own fault.
Another competitor in this field is IBM. They just announced a framework for cross platform, thin client applications. Even if they are very interested in Linux and Open Source, they are more interested in cross plattform solutions than in replacing their plattforms with a new, open source platform like Linux. Don't forget, that they invested very much in Java and still have a strong business with Windows PCs and their propriety big iron server lines.
They don't have a direct competing product to Evolution and the Exchange connector. But they are probably the only company big enough to force customers to switch to their new Workplace plattform. And I don't think Novell is that naive to believe, the groupware client in this new Workplace environment will be Evolution.
So, that's just the two biggest competitors, there are some more competing companies. But if Novell want's to survice in this business, they will have to keep the pace with these two. And open sourcing their Exchange connector is a very good way to attract more people to their own product.
One example would be the PostScript code in xsun.
Okay, the Adobe DPS part should be a real problem, some drivers probably, too.
But a Sunnie told me some time ago, that they are currently investigating how to get rid of Xsun and get back closer to something more popular like XFree86 or the new X.org X11R6.7.
I didn't see another benefit that getting rid of all the legacy stuff from many years, but it would also make perfectly sense, if you want to open source Solaris, or at least the vital parts.
We want a herd of cats, not a flock of sheep, to help prevent the single points of failure that results from widespread use of a single distro.
I agree to that, but I think that the big IT companies like IBM, HP or Sun know of this danger.
If you compare the Linux companies like Redhat or Suse to the IT giants, the latter are still much bigger. They can and will help to keep several competing Linux companies alive. Not only for their customers benefit, but also for their very own. They know, how worse things can get, if you have to deal with a monpolist like Microsoft. :-)
Thanks for that great article, I've just two additions:
:-)
Linux "monoculture"
I don't think, that there is a big thread for Linux to become a monoculture or proriety. (This would be the thread, Jonathan Schwartz was bashed here a few days ago, right?). There are several different distributions, and I doubt that Redhat will become too dominant. The bigger corporations like IBM or HP will be aware, that there are different flavors.
Can Sun legally open-source all of Solaris?
I think they can, they've bought very extensive rights about SVR4 from AT&T years ago. And they got based for paying SCO some money some time ago. So I expect they have all the rights to open source Solaris, at least the SVR4 parts.
And it would be a huge slap into the face of SCO and their cruise against Linux.
PS: Don't get me wrong, this is no critics! I think, she spent much time with Groklaw and she did a really good job. so it's much better to get her into another project than to "loose" her, going back to her studies, an office job, or what she did before Groklaw.
This is a new site, not on Groklaw itself, and it is a community project, not just PJ.
But she is starting it, and it doesn't surprise me, that she is starting this right now. If Baystar want's it's money back, SCO could be history pretty soon. And then there is no more big interest in Groklaw.
This is mostly about consolidating control of not only the data center but the desktop within corporations. LINUX is making a lot of headway - probably faster than Sun & Microsoft ever imagined.
Companies are not yet ready to migrate all of their desktops to Linux. But they are thinking about it, and the most promising place are desktops with a narrow set of applications, like call centers, POS, etc.
And this is exactly the target market, Sun tries to catch with their JDS (java desktop system) products, mostly based on open source.
The biggest problem for a migration to Linux away from Windows is integration with the existing Micrsoft ecosystem: Office, Exchange, etc...
Sun already has answers to some of these, but a cooperation with Microsoft could provide them the rest of the stack.
So they can provide customers with an alternative to Windows on the desktop, which may be nothing for true Linux advocates, but interesting to companies with less religious ambitions.
Think about this, if Linux does totally marginalize Sun (like SCO is now) that means Linux has moved onto the big iron.
Even if Linux is successfull on big irons, this doesn't make Sun redundant. The interesting part for big irons is not the operating system, but also the hardware, the service and the know-how how to run those boxes. And Sun has excactly this kind of know-how.
Microsofts biggest fear is the Linux on small servers and the desktops. Microsoft was very successfull with Windows NT on small servers, because in combination with cheap x86 hardware it was cheaper than the more expensive Unix/Risc boxes.
But Linux can be cheaper than windows, this is the biggest problem for Microsoft. They can try to argue that it's easier or cheaper to develop for Windows or to administrate it, but I doubt that this works.
A second strategy for Microsoft would be to rely on patents and "interlectual property". And a partner in the Unix/Linux camp like Sun, whose software does integrate nicely with theirs, could help them very much.