Slashdot Mirror
The Economist on The Rise of Linux
nickco3 writes "The Economist is telling the business world that Linux is a worthy adversay to Windows and Unix. It is free, runs on almost any hardware, and generally more secure than Windows As result it is dividing the industry into winners that offer Linux (e.g. IBM and HP), and losers that don't, (e.g. Microsoft). Sun is probably doomed."
Actually, the article, if you read it, did not say that Microsoft was losing. Rather, it said that M$ would come out less of a winner.
The poster of the root article needs to read the article. Doh.
You aren't going to get an answer here on /. that will be sufficient for you (and if you expect one, you are just a troller/flamer)... the best way to know is to use it a bit.
A good example of why I love it so much is that *every* object in the Kernel is synchronizable. You can wait on a file just like you can wait on a thread, process, or mutex. Once you get used to the architecture, it's very sweet.
Solaris is indeed an important differentiator, but the hardware's also extremely important. How many Intel systems scale as well? None. SPARC's never been the leader in cpu benchmarks, that's true, but Sun have consistently produced balanced systems that perform well when running applications rather than cpu benchmarks. If just fast cpus were important, we'd all be using Alphas.
Sun's market share in Unix systems grew last year, where HP and IBM sank. HP and IBM certainly do not 'wipe' Sun at this level, either technically or financially. The low end Sun systems are competitively priced and the storage is excellent, so what's the problem exactly?
I'm confused as to why everyone on Slashdot concentrates on irrelevant cpu benchmarks and is so down on Sun at the moment.
They said, in a fine example of shades of gray, that rather than make M$ a loser, Linux will make them less of a winner.
The fact that a slashdot poster can make such a bold (and stupid) statement is not surprising at all.
Infuriate left and right
Threads do not have message queues foo. Windows have message queues... And the queue is associated with the thread that created it, but that's about it.
PS. A window is neither a process, nor a thread. It's a window.
PPS. I wouldn't trust your declaration of inferiority based on the fact that you didn't know the above to be true...
"they wanted to keep one founder (Andy Bechtolstein), who wanted to design a RISC chip (became SPARC). So Sun sold out SunOS in favor of Solaris/SVr4, so they could switch to non-commodity hardware"
.
Everything in this is deadly accurate except for Bechtolsheim's name, the part about him designing SPARC, SunOS not being a part of Solaris, the need to change operating system flavors to change chips, why Sun went from Moto to SPARC, and the comment about commodity vs. non-commodity.
Without going too deeply into the other problems with the above, the reasons that Sun went from SunOS 4.x, which was based on BSD 4.x, to SunOS 5.x, which was based on SVR4, were manyfold:
1. There was a perceived need to redesign the kernel in any event. The 4.x kernel was single-threaded and showing its age; Sun wanted to deliver a more rationally-designed, modular, multithreaded kernel.
2. Many Sun customers were asking for a standards-blessed UNIX. At the time, *the* standard for UNIX was System V (I'm not saying it was a wonderful standard, just that it was the only de jure UNIX standard). AT&T still owned the UNIX brand, and the BSD fork was seen as less "official." For example, KMart had just thrown AIX out of a major deal because it wasn't System V compliant.
3. There was an opportunity for Sun to make significant improvements to System V, and boy did it need it. SVR3 was missing a lot of the goodness present in BSD, and AT&T was inviting Sun to colloborate on SVR4 and merge the two diverging paths back together, to the benefit of both.
By the way, Sun made a significant gaffe at this point that had a major effect on the future of UNIX. When the Sun/colloboration came to pass, Sun promptly alerted the world that this would become the only UNIX that would ever matter, and that they would be one of the major drivers. This so annoyed/scared the other commercial UNIX vendors that they promptly formed OSF [aka "Oppose Sun Forever"] to create Yet Another UNIX. So much for unification. In Sun's defense, this is pretty much exactly what IBM had just said about AIX a couple of years earlier, and what HP is now saying about Itanium. Doesn't make it right, it just doesn't make it rare...
Bottom line: the development of SunOS 5.x, aka Solaris 2.x, had nothing to do with SPARC at all. When the first Sun SPARC systems came out in 1989, they ran only SunOS 4.x, the BSD based derivative; in fact, IIRC SunOS 4.0 was delivered specifically to support SPARC. SunOS 5.0/Solaris 2.0 didn't come out until three years later, in 1992.
I don't know much about Solaris, so I'd like to ask you guys out there. What makes Linux less capable? What does Solaris do that Linux can't do (at least well enough)? Just wondering.
Well, scalability is one major thing. Linux struggles with more than 4 CPUs and more than 4G of memory; Solaris handles hundreds of CPUs and a terabyte of RAM. Linux lacks the manageability of system resources offered by Solaris 9, which allows system resources to be prioritized for different tasks, with a guaranteed minimum available. And even if Linux could do this, it doesn't run on hardware than can be dynamically partitioned, unless it runs as a guest on z/OS, and in that case it's z/OS doing the work. Tight integration with the underlying hardware is another advantage for Sun; they know precisely every component in every system that Solaris runs on, because they designed and built it, so there are never compatibility issues. Solaris' high-performance, high-reliability filesystems are proven, not just betas (yes XFS is also proven, but in IRIX not Linux). Speaking of filesystems, Solaris has ACLs, whereas Linux just has the relatively crude user-group model. Linux doesn't have remote shared memory or IP multipathing (IIRC).
... wider bus, lights out management, proper console connection, Solaris stack and future plans for dual core processors and N1. Performance wise, at the low end Sun systems might struggle against x86 but they come with a whole lot more. This make them extreamly nice from a sys admin point of view that makes them excellent systems.
Have you ever programmed under Win32? This is simply false. Read the MSDN docs for PostThreadMessage() function:
I give up. What is Apache a clone of?
IIRC, Apache started off as a clone of NCSA httpd. I can't find the reference (I thought it was in the config file or man page, but no), but I think they wanted identical functionality and config files at first.
Of course it has grown into so much more now.
Funny you should mention that: I've been noticing quite a few Linux/Mandrake/Red Hat serial number/crack files on Overnet recently. I downloaded one out of curiousity and found that it was just an executable that installed your typical spyware etc.