Slashdot Mirror
Sun Opens OpenSolaris.Org
An anonymous reader writes "Sun has launched the first version of opensolaris.org, featuring a small initial drop of source code. The idea is to make a display of good faith to the Solaris community while the rest of the source code due diligence is completed. The source code for Dynamic Tracing (DTrace) is available for download under the terms of the newly OSI-approved CDDL license."
Sun really seems to like the Open-.org naming convention. They are probably trying to oppose Steve Jobs' iNaming.
Qui ne va pas à la chasse n'a pas de gibier
PHP Queb
I just want the cool features of solaris (such as hot-swappable processors on a multi-processor system) to be ported to Linux. Honestly, bot OS can and should merge into one entity. less fork, more merge.
I'm glad that the source code is starting to be released, but could someone more knowledgable explain what Dynamic Tracing is? Is it something that would be useful to a normal user?
Shocking, I tell you.
Take it easy? I'll take it anyway I can get it . . .
What a lot of Slashdotters might not realize is that Sun has spent literally millions of hours over the last couple of years "unencumbering" Solaris from patented code that was owned by other companies opposed to the open sourcing of their intellectual property. They did this for no reason other than to prove to the open source community that they are serious about open sourcing Solaris, and hopefully to sell some good Sun iron in the process.
It would be nice to see some Slashdotters give Sun their well deserved props for a change, instead of ripping on them.
"What? You gave us OpenOffice? That's not good enough..." I hoping this thread doesn't turn into another Sun bash fest because this time they deserve a little respect for giving away what I see as the crown jewels of their company.
"When the president does it, that means it's not illegal." - Richard M. Nixon
Their press release at sun.com said OpenSolaris via the CDDL will make 1,600 patents available to open source.
-- Microsoft is the most expensive commodity operating system and office suite vendor in the marketplace.
Otherwise it will enjoy a slow death. Obviously maturing and growth of Linux, scared the hell out of Schwartz and his cohorts. Now they are trying to appeal to the OS community to give their precious operating system, which they locked up under layers of safes many many years and expect them to stop or slow down working on linux and make their solaris better instead, which they will be more than happy to incorporate the development and charge the corporations an arm and a leg.
I am not sure about you but I am not buying this half-hearted OpenSolaris movement.
Come, come to my web little fly, said spider...
__________
The more I know people, the more I love animals
Yes, I've already read the wonderful glowing market-speak summary of the CDDL at OpenSolaris.org...
What I really need, and haven't yet found is a nice overall summary of the key licensing points behind the CDDL from someone who isn't Sun.
Anyone?
Thanks in advance!
Didn't Sun buy a SCO license? Oh wait, that isn't a guarantee of not being sued by SCO, I forgot. my bad
http://uptime.netcraft.com/up/graph/?host=www.open solaris.org
OS Server Last changed IP address
Linux Apache/1.3.27 (Unix) mod_jk/1.2.6
Linux Apache/1.3.27 (Unix)
I run SUSE and Redhat ES server here at work but I can't help being excited about DTrace and what it can offer the whole Open Source *Nix world. Sun is definitely helping the Open Source movement by first reelasing OpenOffice and now, DTrace, the most talkeda bout feature of Solaris 10. Wonder how that will effect this between a Solaris developer and a Linux kernel coder?
This guy is way out there
Part of this release is the opening of more than 1,600 patents to the open source community.
:)
link
IBM just got outdone on their 500 patent release. Let's see them come back with 5,000! Come on, it can be a Sun/IBM "who can give away the most patents to open source" war
Finkployd
Solaris x86 and Solaris SPARC are 90+% the same source code, differing only where porting requires. So, the OS programmers on the SPARC side == the OS programmers on the x86 side.
-- Microsoft is the most expensive commodity operating system and office suite vendor in the marketplace.
How is opening their OS codebase, one of the crown jewels of the company, half-hearted? They are trying to remove as many barriers to Solaris adoption as they can. Costs to much? Free downloads have been available for years now. So "Free As In Beer" has been handled. Can't view the source code? Here comes OpenSolaris, under an OSI approved license. The license is not the GPL to be sure..but neither are the BSD licenses, the Mozilla license and the Apache license...you gonna bitch about them?
The fifth board member will be Bruce Perens. I think I tipped it with this. For the link shy:
FTA: However, a source close to Open Source Risk Management (OSRM), which commissioned Ravicher's review, claimed to know what the Jan. 25 announcement was and told NewsForge that it had nothing to do with Ravicher's study.
So, come on Bruce... what's the announcement. We know they meant you! Spill it baby
put the what in the where?
UNIX was open source long before Linux ever came around. Then things changed and all UNIX was closed source for a while. Now it will be open again. Linux may have affected timing, but hackers and open source would have come around eventually, no matter what.
I am not sure about you but I am not buying this half-hearted OpenSolaris movement.
You won't have to buy it--it will be free!
-- Microsoft is the most expensive commodity operating system and office suite vendor in the marketplace.
It doesn't matter nearly as much if they invented them or not, but it sounds like they're making an IBM-style pledge to use their patents as a shield for the open source community rather than a sword against it.
I wish I could write clever and witty sigs.
It's only an expression, you should really just relax :)
Finkployd
Bryan Cantrill, one of the DTrace developers wrote this blog entry as a general introduction to the source code layout and also to DTrace. This post by Adam Leventhal goes into some more detail.
82678 lines of C were made public. No registration, no click through license before download. The OpenSolaris FAQ is pretty good btw, and there's also a roadmap page.
According to this blog (the entry dated 15:43), those in the pilot program (more than 100 developers out side of Sun) have today gotten access to the entire Solaris source base, and have already built their own version - screen shot.
Because after all we know it's a sin against RMS to run a business.
7 November 2006: The day Americans realized corruption and incompetence weren't addressing 11 September 2001
here
I've abandoned my search for truth; now I'm just looking for some useful delusions.
If Solaris is based on SCO's System 5 code, then wouldn't opening it's source (whether Sun has the right or not) potentially pollute other open source projects that borrowed from it?
If Solaris is based on BSD and has no SCO code in it, I guess that's not an issue. But then why did they take out a SCO license? I imagine some conspiracy theorists will say simply to hurt Linux, but that can't be the whole story, can it?
IBM had a SCO license too, but that's because AIX has SysV code in it. That's not the code they gave to Linux, but if they were to open-source all of AIX and pieces of SCO code migrated to Linux, that would be a problem, no?. So why not with Solaris too?
Posted from my Android phone. Oh, I can change this? There, that's better...
You've got no idea what you're talking about.
Maybe for you - the geek in his parent's basement and a nervous caffine twitch preventing from typing properly - things like a $10k price gap between sun hardware and tigerdirect hardware is an issue. But not for anyone that needs reliability and support.
Yes, things like cooperation are a part of why various open-source based companies are doing well. You know why else they're doing well? They've got sound support, sound development, and a good record to back up their word.
Oh, and the first two lines of your post make about as much sense as this string: asdlfkj23ksdlds.
Thank you for your coherrent discertation. The citizens of slashdot salute you.
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
Ok, I read the link to the sun page about D-Trace but that really didn't answer the questions I had. So can any Sun users explain:
:)
1. Why has Sun open sourced this of all things?
2. It seems very similar to gdb in role. Is this assumption correct? Does it compare favorably?
3. Is a Linux/BSD/whatever port of this desirable/attainable? Or does it rely to much on the guts of SunOS? Do we have better tools already on those OS's?
Please be gentle.
I presume (though I don't really know) that Solaris needs to be built with Sun's C compiler. Is this compiler coming forth as an open source release too? If not, is it going to be freely available? If I remember correctly, you currently need to pay in order to get Sun's cc.
If it is coming, this is great news. A compiler highly optimized for Sparc may benefit all operating systems that run on it. Who knows, maybe their x86 compiler has some good features too. Sun's libc (probably highly optimized for Sparc) would be a nice thing to have. Anything else?
And OpenSolaris is the anti-Linux. OpenSolaris is a professional open source operating system centrally controlled and run by a corporation, rather than anti-capitlists and anarchists who's only goal is chaos and hacking.
For those reasons, Sun has released OpenSolaris. It is a direct effort to take the best open source developers off of disorganized projects like Linux and onto a more corporate-friendly project that will in time become the de-facto operating system for all Unix deployments.
Just out of curiosity, who owns CDE? Is Sun able to release this as open source? I've wanted to get my hands on version that works well with Linux or is open source for awhile now, with no luck...
So, I downoad the code, and I take a look at it - the first thing through my mind, is "OMG - look at all the spagetti code!"
Then I realized I opened a C file with Unix returns with notepad.
Oops.
Try to hack my 31337 firewall!
They can be enforced against GPL software including the Linux kernel.
Bruce
Bruce Perens.
over 1,600 patents are being contributed to the open source community from the open solaris website. now the question remains when the refer to the "Open Source Community" are they referring to Open Source as in their CDDL licence - or Open Source Community as it is normally interpreted? Not that I support software patents at all though mind.
Electronic Music Made Using Linux http://soundcloud.com/polyp
Like Mr. Rental!
Note that only projects under the CDDL are allowed to use these patents. Linux developers MUST REALIZE that they are NOT covered by this IP release. Sun has just fired a warning shot across the bow of Linux, and will almost certainly take legal action against Linux developers in the near future as more and more OpenSolaris code makes it way into the "wild". Anyone thinking of implementing DTRace in Linux (or anything REMOELY like it) should realize that they will most likely be the target of a lawsuit by any number of companies including Sun. Additionally, if it can be proven that they visited opensolaris.org, they will be liable for triple damages.
The problem is that this may have been done too late. Linux has the benefit of a huge brain trust. The comp.os.linux.* Usenet hiearchy alone is an extraordinary support mechanism.
I certainly hope this all succeeds. The more competitors in the market, the better their products will become. But Linux's real strength is the community.
The world's burning. Moped Jesus spotted on I50. Details at 11.
Wake me when Java goes Open Source...
Ha! Suck it, SCO!
There is a big world out there, and not one solution is always right for everything.
Outside of the knee-jerk reactions on /. , the whole world should not switch overnight to Linux-Apache-MySQL-PHP. Sometimes, other systems are the right answer, for many complex reasons.
I happen to have a particular fondness for Solaris, having been a fan of their hardware for the last 15 years. It's the Devil I know, and I'm comfortable dancing with him.
I think it's amusingly disingenuous of the slashdot Linux-script-kiddie mentality to ignore that for ten years, SunOS Ruled The Roost in open software, for many good reasons. It is not without its warts (Solaris 2.0 through 2.4 being oozing pustules of lossage), but for an entire generation of sysadmins, Sun was the one system you had to know ... You could add on some of the other big players like Digital, AIX and HP-UX, and maybe one or two of the smaller also-rans like the BSD 4.4 cousins or Linux, but the 800 lb. gorilla was Sun.
Finally: any monoculture is a bad thing, whether it is BSD 4.3 on VAXen, SunOS 5.9 on US-IV , or Linux on Wintel hardware -- and it behooves anyone who wants to be taken seriously to study the differences between systems rather than put all of their energy into denegrating that which isn't their pet.
I think that last part really sums up what I find disheartening with the slashdot collective consciousness. It's that the slashborg will put an infinite amount of energy into defending their point of view, without investing any into analyzing the competition. And, sadly, that more than anything is the sign of ignorant zealotry.
How does the Slashdot Effect happen given that no slashdotters ever RTFA?
They already make better products. Better than Linux, at any rate.
"People who do stupid things with hazardous materials often die." -- Jim Davidson on alt.folklore.urban
I would agree that any monoculture is bad. The thing you and those who share your view are missing is that of all those in your list, linux is the only one which is NOT a monoculture.
Just because it's all labeled linux doesn't mean it's all the same. If there are two opposing camps who disagree on how a component is best designed BOTH will be written and available for compilation. There isn't one linux, there are hundreds of linuxes. It may have the same name, but the linux you run on a wristwatch is NOT the same linux you run on an IBM mainframe.
There's a fox in the barn.
I don't like this one bit. Why this emphasis on Netbeans, OpenThis, OpenThat if these programs run on any hardware? Hardware that is normally cheaper than Sun(although admitedly not as cool)?
If every Sun software became open then Sun, the hardware company, will go the way of VA-Linux.
Have they given up and are now willing to die shooting (at Microsoft)? What is left in their magic hats now that Solaris is free?
Broken Hearts are for Assholes. - Frank Zappa
hah, maybe 5 years ago it was easy to convince management to buy Sun. Not anymore, paying twice the price for half the performance, with dreadful service/maintenance costs? No, mangement wants more Linux, more migrations to Linux, even for proprietary software like Oracle. What with 4 and 8 way Intel boxes whooping the ass of 24-way UltraSparc, it's a no-brainer.
Sun reserves the right to enforce the patents if you use code under a different license.
Although the terms of the license would allow you to fork under the gpl or contribute to a gpl'd project sun could still nail you with the patents.
We used to call Unix Open systems. However, it has not been open since the ATT break up. Sun just wants to capture the label of what they had back in the 70's, when there was real competition.
I prefer the "u" in honour as it seems to be missing these days.
"But not for anyone that needs reliability and support."
Small to mid-sized business needs reliability and support and for them a $10k price gap on every piece of hardware is a VERY significant issue.
It seems like that is the problem sometimes, everyone seems to be marketing to fortune 500's or to home users. They ignore the fact that 90% of those purchasing hardware fall somewhere inbetween.
Linux doesn't have a buggy awk, sed or tar.
Solaris 8 does.
Most x86 hardware doesn't suffer from the transient error bug that the non-ECC cache of the ultrasparcII processor.
Linux works on parking meters.
Solaris doesn't work out of the box with an A1000.
Most quality nics work out of the box with linux.
Most netras and ultras have to either be hardset or vice versa, and won't work the other way.
Don't get me wrong, I like sun hardware (Love LOM), but it and it's software are not perfect.
Dude, it's free...
Either the licenses are compatible or they are not. You cannot mix GPL'ed and CDDL'ed code, therefore they are not compatible. Assigning blame does not solve anything.
They already make better products. Better than Linux, at any rate.
There are various types of better. If I have to deploy an e-commerce site that gets thousands of hits a day then perhaps Sun products are better. For anything that isn't scaled on that level the value proposition favors Linux.
Sun's been using the term "Open" in their stuff forever. Remember, Sun's X environment was called "OpenWindows", and even though they've since discontinued the old OpenWindows window manager, their X server still resides in "/usr/openwin".
Though Sun's definition of "Open" has traditionally been "open standards", as opposed to the F/OSS definition which I believe to be "open implementations".
The patented JRE that people can't legally reverse engineer.
If they charged $100 (on average) for the next version, and they got 10 million people to buy it, that would net them, say after me: 1 billion dollars.
And all those java programs out there would HAVE to buy it, or not be able to run. And java 2, or rather 1.5, is already patent-encumbered. So there. They can fuck java programmers anytime.
Let's not forget that they have a deal with MS on Java. If Sun goes under, where does Java end up? Will anyone buy it and have to honor the agreement with MS? Would IBM do that? Would HP do that? No. Microsoft would buy it. They're just letting Sun bleed itself to a slow death before coming to rescue Java and enforce the patents. (and then its bu-bye independent J2EE implemetations. It'll become Microsoft Java Enterprise Edition. And then MS is where it wants to be, with lots of large corporate customers running its software.
Of course, I wish Sun the best. They, however, do not pay me for advice on where to take their company ($400/hr if they do), so I won't.
"Piter, too, is dead."
The press releases say those patents are only for software under the CDDL license and the OpenSolaris process.
You wouldn't expect them to allow Linux to outright lift Solaris code and put it in their kernel, would you?
"When the president does it, that means it's not illegal." - Richard M. Nixon
Who knows? Maybbe, just maybe, part of the agreement and $$$ between SCO and Sun was to free Sun of any future leagal threats because they were planning on doing this.
Sun has great technology, and has always been available and helpful to the user/developer community whenever they could.
-- When you think of how evil MicroSoft is, remember that they learned it from IBM.
Now I can run it right next to my OpenVMS. -- Scott
its gradient background is not dithered!
Nobody care about 64k-screen users anymore... <sigh>
hmmm... dumb...
Bruce
Bruce Perens.
Incredible. It's actually happening. WAY TO GO SUN!
Now, the great question in open source will be whether the GPL3 and CDDL can be made to be mutually compatible. If this can happen, OpenSolaris and Linux could conceivably combine all their strengths and change the face of computing.
Solaris is Proven and an Industry Tool. Linux has more "street smarts" and some better designed parts (IIRC the scheduler might be an example here?). If I understand matters right now, CDDL code could go into GPL code but not vice versa. Which is a shame in some ways, because I think the Solaris name is a coin that open source could make good use of.
"I object to doing things that computers can do." -- Olin Shivers, lispers.org
True, but what are the odds they would do so? I'm guessing if they wanted to they could slap Linux with patent lawsuits anyway, regardless.
Sun cannot be so stupid. There is no money to be made in the long run with Linux patent lawsuits - all you will do is kill Linux. And probably open source itself. Sun is not Microsoft - they cannot scorn all allies and hope to survive.
Really, I don't think much of this whole patent thing in general - releasing 1000 or 10000 isn't enough to remove the threat - a lawsuit over 1 weak patent is still enough to sink most open source projects. IBM's move was good because they stated openly that Linux was part of the patent standoff, but I think that's more or less implicit for most of the software industry anyway. The Linux kernel happens to have both powerful champions and major enemies, but EVEN THERE the patent demon has yet to seriously rear its head, when it is probably the single effective way to kill Linux. It's MAD in a true sense, because if patents ever become offensive weapons in the software world there's going to be nothing left of the US IT industry but a smoking crater full of radioactive legal slime, and the first bomb blowing up over open source would be unlikely to change that in the long run. I bet the release was just an attempt to keep the trolls from claiming the CDDL was an attempted lawsuit mousetrap.
"I object to doing things that computers can do." -- Olin Shivers, lispers.org
You can't run DTrace on Solaris 9 and earlier; the kernel's incompatable.
-eventhorizon
#Secret Windows Source Code, in MS C% - if (uptime >= "24 hours") then bsod() else print "Windows License Violation!"
Note also that IBM's grant came just in time to drown out news about 61 European Parliament members asking to restart the software patent debate there from zero. IBM is one of the main parties lobbying for European software patents. Their grant is part of a larger strategy to convince European legislators that Open Source and software patenting are compatible, that could indeed kill Open Source, because it would leave us vulnerable to many software patent lawsuits.
Bruce
Bruce Perens.
It possibly means people attempting to claim code witten by others as their own.
-eventhorizon
#Secret Windows Source Code, in MS C% - if (uptime >= "24 hours") then bsod() else print "Windows License Violation!"
A lot of people are interested in OpenSolaris. And the lawsuits are things all businesses have to contend with, and no matter how much you try to live in an idyllic fantasy will change that. The two billion Sun got from Microsoft is justice served for how much Microsoft screws the whole industry over. Microsoft needs a few more billion dollar payouts to set them straight.
And, please, put "slowaris" to rest. It's the OS-equivalent of a baseless racial slur. If it had any substance to it, Sun wouldn't be setting records on Opteron and UltraSPARC IV, right now.
-- Microsoft is the most expensive commodity operating system and office suite vendor in the marketplace.
Ignoring for a moment the question of whether it's buggy, who gives a damn about Solaris 8? That was the 90s, man. OpenSolaris is based on Solaris 10, the release of which is imminent. It's a boatload of new technology plus two full releases' worth of bug fixes removed from Solaris 8. If you had a bad experience, we're sorry, but please don't continue feeding people misinformation based on a badly outdated release. Should I talk about my experience with Yggdrasil Plug and Play Linux (based on kernel 0.99!) in 1993 and pretend my difficulties then are reasons to avoid Debian GNU/Linux 3.0?
Next, pedantry.
Yes, sed and awk have bugs filed against them. Probably everything except maybe /usr/bin/true does. Since you didn't say what kind of bugs, it's hard to assess the legitimacy of your complaint. If you'd care to elaborate, there's a chance your problem can be fixed. Otherwise everyone just assumes you're a crank.
Most of the "bugs" in Solaris userland that draw a lot of complaints from GNU users are not bugs at all but rather artifacts of painstaking maintenance of compatibility with long-extant standards. The GNU tools you're familiar with follow different (or their own) standards, because backwards-compatibility is not a GNU priority. It's a constraint imposed on Solaris development by our customers, who seem pleased with the results. You can in some cases make the GNU tools behave similarly if you wish. Solaris also has several different versions of these tools, each of which matches different incompatible standards. You might just be using the wrong one.
Finally, Linux doesn't have awk, sed, or tar. It's a kernel. I've assumed here that you mean the GNU tools, though of course each distro ships its own version and combination of specific tools, but it's certainly possible to build a system with a Linux kernel and non-GNU userland. In fact, soon it will be possible, at least in theory, to make a system with a Linux kernel and a mostly-Solaris userland, if you really want to.
Ok, on to your other concerns.
Most x86 hardware doesn't suffer from the transient error bug that the non-ECC cache of the ultrasparcII processor.
What does this have to do with Solaris? If you think x86 systems are more reliable, you can run Solaris on those instead.
Linux works on parking meters.
So does Windows CE. Should we all start using that? Standard Linux won't run on less than about 4MB of memory, even if you're talking about Linux 2.0. Newer means bigger. Solaris as shipped today by Sun also requires more resources than would normally be found in a small embedded system. It's a safe bet that, like Linux, the SunOS kernel could be reduced sufficiently for such an application. It's just software and work. The fact that a highly modified version of Linux can run a parking meter is a curiosity and a testament to the effort of the hackers who did it, not evidence of inherent software superiority.
Solaris doesn't work out of the box with an A1000.
I can only assume this is yet another reference to some past bad experience you've had. Since you're not telling us what it is, it's impossible (a) to know if it's still a problem, and (b) to fix it if so. Please, be specific if you want to complain. Hint: referencing bug IDs is a good idea.
Most quality nics work out of the box with linux.
Which nics would you like to use on Solaris? In the 10/100 space there's iprb (Intel), elxl (3Com 3c9xx), dnet (Tulip/21xx0), and about a half dozen others. In the Gb/10Gb space there's e1000g (Intel), bge (Broadcom), sk98sol (SysKonnect), and xge (S2IO). See the reference manual for drivers if you want to participate in an informed discussion. The examples I gave don't include the various vendor-supplied drivers or any of the Sun-specific d
There is a big difference between the IBM and SUN patent pledges.
:) Fun stuff though, and I think pragmatism will win the day if there good stuff is delivered.
IBM listed a broad range of software licenses, importantly including the GPL, which means linux is covered.
Sun's license so far is limited to Solaris, or at least it looks that way, where they have contributed code under the CDDL. This means if you take a method (or read about a method) that they use in Solaris and apply it elsewhere you can still get slammed.
Not a black and white issue though, as the discerning reader will note that the GPL has not patent clause at all, so the CDDL is stronger in one sense there. Not sure if Linux is any worse off.
But it will be interesting to see how Solaris comes out as open source, incredibly it has gotten to this point for those who remember the Sun of the past (and even some of the current ranting). Losing market share is an incredible motivator it seems
CDE?
I mean, what kind of nutcrack would you have to be to crave for such an ugly piece of software as CDE?
I mean, ANYTHING in GNU, from fvwm to blackbox, not to mention gnome or kde/Linux is ages ahead.
NO SIG
What's really hilarous is that the screen shot has telnet in use. Telnet. I can only hope its disabled by default, but even so ... I wouldn't choose that as a demo for any OS (Yes, I know connecting to 127.0.0.1 is safe, it's the principle of the thing).
Separately to that, looks interesting.
You're mistaken, AC. OpenSolaris can co-opt BSD code, because the BSD license specifically allows anyone to use their code for any purpose -- simply put, the BSD license is compatible with any license, as long as credit is given. It's as free, in that respect, as you can get (without going public domain).
The BSDs, however, cannot use any of OpenSolaris' code, because you cannot relicense CDDL code as BSD (if you could, it would be trivial to put it into Linux -- after all, BSD licensed code may be added to the Linux kernel, and in fact has been on several occasions.)
The BSDs will benefit no more from OpenSolaris using their code than they do from NT using their code.
This is essentially how BSD works -- everyone benefits from BSD, but no one is required to give back. The BSD people don't see any issues with this, so more power to them.
Linux can't benefit from OpenSolaris either -- but because of the GPL, this means that OpenSolaris also can't benefit from Linux. So it's a symmetric relationship in that respect.
Anybody know where to look?
I'd not mind if this allowed us to escape the "non-supported" part of Solaris by allowing 32bit SPARC machines and the peripherals they run(Yep, that includes some ZX support) to run a build of this. Throwing out a good chunk of the installed userbase on Solaris 10 and then releasing source for it, is not the best idea out there.
Twitter supports and protects racists - by smearing their critics with the "Hate Speech" label.
They might fall in between, but there isn't really an "in between" support option: you either have it, or you don't. Should RedHat (et al) provide preferential treatment towards small companies?
Just do the math. $10k for a small company, while only having to hire one or two tech employees, is insignificant compared to the cost of $6k, and having to hire even 3 or 4 employees to do the support: those 3 or 4 employees will be more expensive - indvididually - than the 2 or 3 employees with our $10k option, because they'll not only have to know more, but they'll also have a much broader responsibility: essentially building and maintaining distros themselves, with nobody to fall back on except the original package managers.
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
I am interested. I have been for a couple years. I've downloaded Solaris ISO's since version 8 for x86, and I got my hands on some x86 ISOs from a friend of mine for version 7. Please do not talk as though you speak for all open source software users, or some larger, abstract community you think is represented by the lifeless dozens you hang out with in #linuxizdeebest on freenode. I use FreeBSD. I use OpenBSD. I use OS X. I use Windows. I use NetBSD. I use Linux. I used Solaris at my old job. I am not interested in pigeon-holing myself based on mindless idealistic zealotry. Your opinions are indeed your own, so please don't lump others into some all-encompassing "we" based upon your own voluntary closed-mindedness.
I had forgotten about Yggdrasil, but now that you remind me, it was a good bit easier than trying to put together a runnable 0.97 out of dozens of bits and pieces. Especially when one had no live Internet connectivity and had to uucp or email archive fetch the pieces from somewhere or other. I did get a runnable 0.97, and it was a revelation. A 486/25 which was nearly unusable under Windows, made a very speedy compile machine under Linux.
For starters the Mozilla license is not compatible with GPL, yet nobody makes silly comments about Firefox not being open enough.
The license Sun is using is just a variation of the Mozilla license.
There is more to Open Source than just Linux. The whole point is we now have two serious Open Source server OS competitiors to MS.
Linux is free as in speech, but the Enterprise distros are not free as in beer. Of course other non-certified distros are free as in beer Fedora, Mandrake, Debian etc.
Solaris is free as in speech and as in beer.
Many people seem so blinded by anti-UNIX rage that they don't even notice that Sun has contributed more lines of source code to Open Source than any other company. Virtually their entire software line will run on any OS including Linux. They are totally obsessed with crossplatform compatibility - hence there obsession with Java.
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? ;-)
"that will not violate CDDL as long as they make their changes to the CDDL code available to others" AND place it too under the CDDL. CDDL code is exactly as "viral" as GPL, except that (to some) the "viral" nature of the GPL is seen as an inoculation.
I've now read their Common Developement and Distribution License (CDDL), and it seems rather complicated, complex and problematic. It is a free software license, and even some type of copyleft. But it is so wordy and uses such a complex language that it is hard to tell what it really says in some sections, and some of that could even be used to construe a workaround for the copyleftism... In addition, it contains a shitload of of special patent-related rules... All in all, it is NOT GPL compatible, and not SANity compatible to read :( They must have smoked some expensive shit when writing this...
--The knowledge that you are an idiot, is what distinguishes you from one.
Personally, I'm a bit disappointed with the rather meagre commenting and suspect stuff like parenthesizing return statements.
@peetm
I hope someone is looking into porting that DTrace beauty over to linux if it's possible (i.e. if it's messing around in kernel space or is using process/system info only available on Solaris it might not be possible...?)
they apply the same rules to licenses to be considered "open source" (OSI) or "free" (FSF).
the only difference is philosophy
That's reasonable, though it *will* make life easier for someone who's already breached the network perimeter.
I wouldn't choose telnet as an example for anything though, and frankly would hope it's disabled by default in any modern OS.
> CDDL code is exactly as "viral" as GPL
...
No, it isn't -- CDDL is only viral to the file level (any change to a file licensed under CDDL must also be licensed under CDDL). GPL is vial to the *link* level -- any code which links against GPLed code must also be GPLed.
So you can have CDDL files in a source base that is otherwise BSD-licensed, or is otherwise closed-source, or
http://devurandom.blogspot.com/
Outside of the knee-jerk reactions on /. , the whole world should not switch overnight to Linux-Apache-MySQL-PHP. Sometimes, other systems are the right answer, for many complex reasons.
I agree completely.
Linux/Unix-Apache-MySQL-JSP/J2EE..
I'm not going to have a pissing contest because we could probably go back and forth all day...but just one thing.
a re/story/0,10801,66102,00.html
;-)
Du you just want to repeat some FUD about Sun, you heard somewhere?
This was _not_ a bug I heard somewhere. I was one of the first people to ever get this bug. I ran 14 netra t105s, 25 ultraIIs and 4 220rs. All of them had the UltraIIi processor. Most of them were incredibly stable, until a cache error crept in. Then I'd find them at an OK prompt with nothing to say for themselves. I'd go digging and find something like "Transient Error CPU x" (that's not what it said, but those were some of the words) in messages. After explaing to the boss why the oracle/web/nfs server died in the middle of the night, I'd have to make a support call to sun (You know, those guys you pay an extraordinary amount of money to).
Sun would give me the run around on the phone, then after a couple hours and talking to every "engineer" there, I was told it was a transient error, and it would only realistically happen once in a server's lifetime.
Only later did I run across more people having the problems, and none of them were given a straight answer from sun. This, of course, shattered any confidence I had in Sun.
Later, when McNealy actually acknowledged the problem, Sun earned some of my respect back.
More:
http://www.computerworld.com/hardwaretopics/hardw
BTW, regarding your BTW:
BTW: There is so much broken with x86 hardware
I paid for Sun hardware expecting the best out of either class. When I buy cheap x86 hardware, I'm prepared for the worst. However, when I buy a 3com905b, I know I'm getting a good qualitity card. When I buy a Cyrix, I'm prepared for the worst, I'm ready for a bug. When I buy a Sun, I assume the hardware has been tested and some engineers have poured their heart and soul into the designs of the unit. Now, I just purchase AMD and in the future I will probably try to get some IBM P720s to test.
Isn't UltraSPARC II Sun hardware, too?
I don't really know, sounds like you do. I know that the last sun I owned had it, and I loved it.
p.s. Sorry if it's all garbled, I haven't had any coffee yet.
Reading these threads, here are some points that I think need to be stressed:
The real reason we want open source software to be compatible with the GPL is not because we want everybody to have drunk the kool aid (okay, there are some fanatics who do want it for that reason), but for the more practical reason of wanting to be able to incorporate code from one open project not under the GPL into code from the large body of existing GPL code, either to create a new app or enhance an existing one. If Sun's license is not GPL compatible, then code from Solaris cannot be included into the Linux kernel, for example. Yes, there are a lot of kool aid fanatics, but there is a practical reason behind desiring this compatibility. (Of course, that still doesn't obligate anyone.)
This is a recurring problem, and not just with Sun's license. The original BSD and Apache licenses, the Mozilla license and many others are not GPL-compatible. OSI-certified licenses are arguably going to also be "free software" licenses as viewed by the FSF, yet the GPL still won't play nicely with others -- even those on the same side of the philosophical fence!
The design of the GPL is the problem, not the myriad OSI-certified licenses which are incompatible with the GPL. The GPL should be modified to allow GPL code to be combined with ANY free software (or at least copylefted free software) without requiring that ALL the software be distributed under the GPL itself. At one stroke, the "GPL incompatibility" problem could be solved, but only by the FSF.
I asked Richard Stallman about this once. He felt it was unnecessary to adapt the GPL to the reality of license proliferation because those other projects should just adopt the GPL instead. (This is pure hubris, of course.) However, he did admit that at least a few major licenses (like the MPL) were unlikely to go away and that it was unfortunate that GPL and MPL code cannot coexist.
He then argued that it would be inherently dangerous to modify the GPL to allow other licenses, because such a change might inadvertently open a loophole allowing proprietary software to be used with GPL software. Ignoring the fact that loopholes already exist, this is a valid concern. OSI has shown the way; particular licenses can be certified. While not foolproof, it would be much harder to sneak in a loophole in the text of a specific license (where the overall intent is probably obvious) than to craft a license designed to subvert precisely specified criteria.
Obviously, the FSF could certify free software licenses (as OSI does) and then modify the GPL to allow GPL software to be used with ANY software under an FSF-certified free software license. This places trust in the FSF to do the right thing, but so does "GPL version 2 or any later version". In fact, this very trust is the only means by which much GPL code could become more compatible with other OSS licenses.
Unfortunately, RMS dismissed this idea out of hand, mostly due to the burden of maintaining such a list and the risks of possible hidden loopholes from combining licenses.
I don't believe such a list would be very burdensome to the FSF. OSI already does all the necessary work -- the FSF could routinely wait for OSI certification before even considering a license for certification. Then the FSF's lawyers could double-check the license, but with the knowledge that others with the same goals had already examined and approved it.
Plus, there is no inherent need to certify every license -- they could just certify the high-profile ones, like old-style BSD, Apache, MPL and others where a significant codebase exists under the license.
Better yet, the certification process could provide a funding source for the FSF. For a suitably large sum of money upfront plus ongoing costs, the FSF could analyze a proposed license for certification, and work with the license author to resolve any conflicts. The FSF would keep the upfront money whether or not the license is eventually certified, and use it to fund free software development.
Companies (like Sun)
Deven
"Simple things should be simple, and complex things should be possible." - Alan Kay
Small to mid-sized companies may or may not even have ONE inhouse Technician. They often hire 3rd party tech consultants only, or perhaps in combination with one inhouse administrator. This gives them the price of one set of hands onsite and the expertise of a whole team of techs.
ObDisclaimer: IANAL
Where did you get that idea from? Yes you can mix GPL and CDDL code. Sun's License FAQ basically says so.
No, the FAQ says that files under the CDDL may be used with files under other licenses, but that those licenses might prohibit combining the code.
Look at it like this. I take your GPL code and I add some CDDL code. I modify it a bit to make it all work together happily. Then I distribute the source code on my website. What are you going to do? Sue me? You can't, I gave you the source code. GPL obligation fulfilled.
It's not that simple. If you want to mix GPL code and CDDL code in the same file, you cannot distribute the combined code under the terms of the GPL, which means that the GPL does not provide you with a copyright license for this. You would need a copyright license under an alternate license (if one exists), or need to be the copyright holder of the GPL code yourself. (The copyright holder can violate the stated license with impunity since they don't rely on the license for the authority to copy.)
You cannot combine GPL and non-GPL code in the same file with impunity, because the terms of the GPL are very strict and specific. There is no dispute that including GPL and non-GPL code in the same file creates a "derived work" under copyright law, and distributing (copying) such a derived work requires the copyright authority for the GPL code, either directly or via a license.
However, the definition of "derived work" is a legal gray area, and one that would ultimately have to be decided by the courts. Unfortunately, the answer to this question would probably vary on a case-by-case basis depending on the facts of each case -- and on a jurisdiction-by-jurisdiction basis as well, based on the precedents and laws applicable in each jurisdiction.
If you want to be safe, assume everything in the gray area is a derived work. This is the FSF's interpretation, and they will defend it, so unless you're prepared to be a test case (and possibly lose), think twice before venturing into that gray area!
The idea of "user does the link" as a method to evade the GPL was considered and discussed when GPL v1 was released. The FSF has always maintained that code written to work with GPL code is derivative of that code, making the "user does the link" scenario an attempt at evading the requirements of the license. However, this is a question of law, and the FSF's opinion isn't binding. To date, nobody seems willing to test the issue in court.
However, there is at least one precedent which suggests that the FSF's interpretation of copyright law may be too inclusive, at least in some jurisdictions. (Other jurisdictions seem to have different standards, which further confuses the matter.)
If the "Abstraction-Filtration-Comparison" test is used to determine what defines a "derived work", then code which simply uses the API of a library probably wouldn't be considered "derived" from that library, any more than using the Win32 API would give Microsoft copyright powers over Win32 applications or even Wine. Mixing and matching source files might be considered derived; it probably depends on the facts of the case.
Note that this does not apply to statically-linked binary executables -- those are unquestionably creating derived works by translating the sources and then combining them into the same file. Dynamic libraries, on the other hand, are also a gray area.
It's possible that you aren't creating a derived work by simply using the GPL code from other files, but the FSF believes you are. If you aren't, then the GPL and LGPL would be equivalent in effect. The GPL becomes moot with regard to non-derived works, which are even allowed to be aggregated on the same media. The plain text of the GPL makes this clear. But don't expect the FSF to budge on their interpretati
Deven
"Simple things should be simple, and complex things should be possible." - Alan Kay
8 way Xeon came out in 2001
I had to do benchmark for client with 4-way DL 5800g2 with 2 hyperthreaded CPU and 16GB ram against 24 CPU 6800 attached to same SAN - 63,000 TPS for SUn and 148,000 TPS for HP. Linux x86 machines not "big iron" because you can't CPU/IO partition them, and can't hook them up to as many things - the big Sun boxes will kick butt in that regard, but I don't see them typically being used that way at my clients.
Numbskulls like you deserve to be sued the shit ouf of, by someone like SCO. Learn to give credit where it's due. All you ingrates at slashdot will one day realise that calling sun evil & throwing the Linux buzzword takes you nowhere. Can you count upto 10 million? No? That's how many lines of code sun will be giving up. They're opening up their crown jewel.They didn't have to do that. S10 has enough great features to make it scorch the market, beating your low end, kick ass cheap linux servers hollow. They did that to reiterate their commitment for the community. They've ALWAYS given more to the community that IBM or HP. The latter 2 are the evil ones, hoodwinking shortsighted slashdotters by throwing the Linux word. Yes. Linus torvalds did a great thing by giving away the kernerl free. But so will sun now. FREE.OPEN SOURCE. Enough is enough. You now have a SUPERIOR OS open sourced. Go with it or be left behind, using some ancient linux version and dancing to the tunes of IBM & HP. btw, look at the 500 patents IBM has given out & the 1670 that sun has.You'll see that sun's patents pertain to current,hot, bleeding edge technologies. And of course, i shouldn't even mention HP. Who cares even they give out the code for ALL their rickety management software and the legacy HP-UX. They're just an integrator, an imaging & printing company. They never have & never will matter.Time you saw the light, ungrateful, dim witted slashdotters!I'm sick of your brainless anti-sun rhetoric.Use some logic.Get your grey cells working. There's nothing wrong in supporting another open source initiative for a superior OS. And this has nothing to do with M$, for chistsake.
... give us Java, you fucking morons!
What about the patent situation? Does linking to a CDDL file then provide access to the 1600 patents that are only available to CDDL code? Or would the entire application have to be CDDL? 'cause w/o that patent protection the use of those files is dangerous.