Farewell To OpenSolaris. Oracle Just Killed It Off.

[jadrevenge]

SVR4, it doesn't trip off the tongue but would encapsulate all Solaris derivatives ... but it'd also allow BSD discussion, I guess ...

"Solaris" instead of OpenSolaris might do, except that the Older Solaris 10 doesn't have the graphic, or hardware support that Phoronix generally deals with from a gaming perspective.

*BSD is treated as "BSD Operating Systems" on the forums, maybe we should have "Solaris Operating Systems"

[nanonyme]

Speaking of OpenSolaris and its death: This section of the forums is called OpenSolaris. Now that OpenSolaris isn't alive anymore, should Michael do anything about this? My vote on Opensource Solaris given there might still be projects who want to continue distributing the possible future CDDL code produced by the Solaris project but it would still say that this isn't anymore a forum section related to that particular distro that died.

[PGHammer]

My Feelings on Solaris

Umm ... you do realise that this forum is for Solaris-philes and Solaris-haters, and I guess between us we all pretty much hate windows, or developing on it ...

I actually like Solaris (and had been a long-time user of OpenSolaris) as a testing/education platform; Sun's licensing program (especially for Solaris on Intel, and later for OpenSolaris) made self-paced learning on Solaris worthwhile (Solaris in particular has a solid niche in the financial-services industry; particularly banks and credit unions). However, given Larry Ellison, even I could see the writing on the wall after Oracle's acquisition of Sun.

I was speaking in terms of *development* (and development languages); thus my referring to Mono/Moonlight and alternatives to Java (which is also Oracle-owned), as opposed to operating systems/distributions.

[jadrevenge]

Teaching C++ *used* to be dangerous; however, that was only in the period between the death of the old "Learning C++" (Microsoft Academic) package and the birth of Visual Studio Express (which included C++ from the beginning). That period was less than two years in length. Today, Visual Studio Express has expanded to cover *all* of Microsoft's non-SQL-based programming languages (an Express version of SQL Server itself is also available). Further, Microsoft has *lowered* the requirements of their development tool add-ons (including that for Windows Phone) to Visual Studio Express (from Visual Studio Professional). Further, the runtime-distribution problem has been solved - via the use of Windows Update/Microsoft Update, in most cases, developers (especially small or student developers) no longer have to *fatten* their code with truckloads of runtimes - just include the runtimes used in developing the original application (WU/MU on the target platform handles the rest). Can this be extended to non-Windows development? Rather easily (in fact, I know of no Linux distribution, or even BSD distribution, that lacks an updating method), so using Moonlight/Mono .NET is feasible.

Lastly, you don't even have to be a student - Visual Studio Express Editions are available, in over sixty languages, to most of the planet, for only the cost of storage and bandwidth. (In short, it's a download away.) While Oracle (and to a large extent, IBM) are trying to monetize all their IP, Microsoft, which had often been accused of it, is actually loosening the apron strings. Is Microsoft the anti-Oracle?

Umm ... you do realise that this forum is for Solaris-philes and Solaris-haters, and I guess between us we all pretty much hate windows, or developing on it ...

[PGHammer]

Dangerous Teachings

Fool me once, shame on you. Fool me twice, shame on me.

OpenSolaris is once.

If they repeat this pattern anywhere else in their currently-open-source lineup, I'm going to prod my FSF contacts (I'm a card-carrying member) to scrounge up some campaign funds to resume the "Java Trap" campaign and to advocate KOffice or AbiWord to replace OO.o, and Postgres to replace MySQL. I might even donate specifically for that cause.

The real problem out of the above, though, is Java.

If Java turns into an "IP base" that Oracle uses for suing companies over patents that they own on Java, this is going to hurt a lot more people than the sum total of losing OpenOffice, MySQL, and OpenSolaris.

1. Java is the language taught to beginning computer science and computer engineering students in accredited public and private universities across the U.S. Java is often used as the "gold standard" language when professors are trying to teach students a particular programming concept or system concept that is language-agnostic, because they know that the syntax of the language will not get in the way of students' understanding of the salient points.

Java was a safe bet under Sun -- especially in the years after they open sourced their own JDK. Now, is teaching Java in universities going to be as dangerous as teaching Microsoft C++ (using C++ primarily backed by the Windows API) or Delphi, or Cocoa/Obj-C?

The problem with teaching "closed" platforms to students is that they imprint on the first thing they learn, and they always end up going back to that when they are given a choice of what language to implement something in. So it biases students on using a particular platform. This is not really a problem if the platform is cross-platform and patent-free (like standard C++, or ANSI C, or GLib/GNOME/GTK+), but if the platform is closed, this means that the apps of the future will be written for a closed platform, too, which regresses the goals of the Free Software movement significantly.

2. There are an enormous amount of deployed lines of Java application code, ranging from the system level to middleware, to desktop apps, games, and web apps. Java ME is used extensively for all sorts of applications on mobile platforms. It appears where you least expect it. And it runs on umpteen different JVMs, many of them not licensed explicitly with a monetized agreement with Snoracle -- some of them are even Free Software, like GCJ/Classpath and Harmony. Are the users of apps that run on these JVMs, and the developers of these JVMs, vulnerable to Java patent lawsuits?

3. There aren't really any other languages that you can port Java programs to without much effort. The closest language neighbor to Java is probably C#. The ironic part is that Microsoft has been a better open source citizen about the .NET platform (at least as far as patents are concerned), because they made a "promise" that they won't sue Mono, or even more generally, some other unrelated project that just aims to follow in Microsoft's footsteps and re-implement .NET for any reason. But it is valid to doubt Microsoft's intentions on this, and to think that they might not cover some particular implementation detail under their promise if they really want to sue you. Therefore, Java developers who want to work with a very similar language (syntactically and library-semantically) are stuck going with the next nearest neighbor, which is probably Qt4/C++. But then you've got pointers and stack-allocated objects; Java kids have an allergic reaction to both of these.

TL;DR: Java filled an important hole in the educational and commercial segments, providing a useful Virtual Machine + JIT + static typing runtime, cross-platform, royalty-free, while happily allowing others to re-implement any part of the Java technology stack. Even Richard Stallman -- the most wary evil-corporation-doubter I know of -- had begun to have faith in this platform after Sun opened up the JDK. Now, under ownership of a bunch of absolute idiots who think they can foil our techno-cultural movement and make a cool buck while they're at it, we have reason to doubt whether Java is still a safe bet.

This is more than a sad day for (Open)Solaris. This is a sad day for Free Software.

My opinion is that we should give Oracle one more chance to f*** up. If they pull this stunt on any other piece of software, it's time to kick ass and chew bubblegum. Until then, I'm willing to let this slide, attributing it to mismanagement during the Sun transition. This news, by itself, isn't the end of the world for the vast majority of the people, who don't use (Open)Solaris. But it might be the harbinger of a dangerous precedent within Oracle. If that turns out to be true, then it's time for Free Software and Open Source advocates alike to inject renewed energy into the bazaar development model, and replace all software we depend on that is primarily copyright-attributed to Oracle, GPL or otherwise.

Teaching C++ *used* to be dangerous; however, that was only in the period between the death of the old "Learning C++" (Microsoft Academic) package and the birth of Visual Studio Express (which included C++ from the beginning). That period was less than two years in length. Today, Visual Studio Express has expanded to cover *all* of Microsoft's non-SQL-based programming languages (an Express version of SQL Server itself is also available). Further, Microsoft has *lowered* the requirements of their development tool add-ons (including that for Windows Phone) to Visual Studio Express (from Visual Studio Professional). Further, the runtime-distribution problem has been solved - via the use of Windows Update/Microsoft Update, in most cases, developers (especially small or student developers) no longer have to *fatten* their code with truckloads of runtimes - just include the runtimes used in developing the original application (WU/MU on the target platform handles the rest). Can this be extended to non-Windows development? Rather easily (in fact, I know of no Linux distribution, or even BSD distribution, that lacks an updating method), so using Moonlight/Mono .NET is feasible.

Lastly, you don't even have to be a student - Visual Studio Express Editions are available, in over sixty languages, to most of the planet, for only the cost of storage and bandwidth. (In short, it's a download away.) While Oracle (and to a large extent, IBM) are trying to monetize all their IP, Microsoft, which had often been accused of it, is actually loosening the apron strings. Is Microsoft the anti-Oracle?

[kraftman]

Or more graphically http://xkcd.com/386/

You're both are right. Time to rest.

[nanonyme]

which is better, the phrase "You're both right", or the phrase "You're both wrong" ?

More along the lines of "you are both right and you're both going to lose if you continue this argument".

[jadrevenge]

which is better, the phrase "You're both right", or the phrase "You're both wrong" ?

[nanonyme]

Or more graphically http://xkcd.com/386/

[nanonyme]

You both probably have a point! :> However, it was explained many times to Kebabbert he won't prove what he's claiming, but he continues from some obvious reasons.

And you continue fueling the debate for other obvious reasons.
You really can't win an argument on the Internet. Even if you don't end up losing, you've still wasted a lot of time on something completely useless. ^^