Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Amiga

Linus on Amiga decision 259

Posted by CmdrTaco from the stuff-to-read dept.
amiga_dude sent us an article that has a ton of information about the confussion and questions surrounding the recent Linux Amiga QNX news that has been sprouting up. This one is a pretty good catch all piece with some Linux words as well.
This discussion has been archived. No new comments can be posted.

Linus on Amiga decision More

Linus on Amiga decision

Comments Filter:
  • ...for Linux Pagestream [softlogik.com] has long been my favourite DTP program, and Lightwave [newtek.com] has long been my favourite 3D animation program. Take the hint, guys!

  • Re:I'm glad that they chose Linux (Score:1)

    by Anonymous Coward
    NG Amiga stands for Newsgroup, because it's beyond vaporware in that it doesn't even leave the newsgroups and make it's way into print media, like the usual vaporware does.

    Maybe somebody should slap the Amiga brandname onto some sort of PDA or Net PC something, just to keep the Amiga fetishists placated.

  • Re:Information on Linux TCP/IP Stack (Score:1)

    by Anonymous Coward
    I don't know why anyone here thinks that BSD'ers such as myself look down at Linux. Analysis of the second Mindcraft tests showed that the threading in the Linux TCP/IP stack wasn't pervasive enought to scale well onto multiple processors. On the BSD development mailing lists, the question came up, "How would we have faired, and does our TCP/IP stack have the same problems?". The response from some of the core development members was that the current BSD TCP/IP stack suffers from the same malady, a giant lock around the TCP/IP stack. It's not much of an issue on a single processor, but throws the same obstacles in BSD's way as it does Linux.
    I wish people here would try not to be so reactionary, how many of us have ever committed to either the BSD or Linux kernel code? That's what I thought.

  • This is what Linux needs to go mainstream (Score:1)

    by Anonymous Coward
    If Amiga does this well they can make the first real userfriendly system that joe computer user canget on and actually use. Redhat seems to go for this with rpms and whatnot but still is way beyond your normal user. If they accomplish that it could succede I think, with gateway providing the money for marketing since they own amiga. Get a portion of your average computer user market using linux and software manufacturers would start producing Linux software.

    I think it has a good chance of working, I don't know why #amiga and random amiga people I have talked to are so against the idea of Linux.

  • Re:Why Amiga switched (Score:1)

    by Anonymous Coward
    it was so easy to attack the message that there was no need to attack the messenger.

    The messenger has a long history of non-altruistic behaviour. His stack ships with back doors and time bombs. His paranoid rants against so called "pirates" are well known. He considers himself the "King" of Amiga tcp/ip. He thinks all Amiga tcp/ip should pass under his control. Take heart: his reign is just about over.

  • Re:Applications? (Score:2)

    by Anonymous Coward
    Of course, that won't happen unless they're using X, the same libraries, etc., etc.--in other words, if it is "just another distribution."

    Tapping into the existing base of X software is the only way they will succeed. They need to have a critical mass of software apps, and Linux/BSD running on X is the only viable alternative to Windows right now. It doesn't matter how good their OS is, they will end up struggling along with Be if they don't realize this.

    I'm all for dumping my apps every two years and starting from scratch, but most consumers are like lemmings.

  • Almost hillarious QNX worshiping (Score:2)

    by Anonymous Coward
    This article is amusing, the author has clearly never even seen any of QNX's software let alone used it! The comparisions he makes with Linux are simply laughable. For instance, Linux's IP stack may not be the best in the world, but QNX's certainly is no better right now - Neutrino was not created soley for Amiga, I imagine that QSSL was asked if they could provide a real time OS by Amiga and simply re-fitted a few things they were already doing for their real customers.

    The concept that Neutrino is an OS without a market is even more amusing, last year our little company easially shipped 1000 Neutrino licenses, and we are not a big customer. I assure you that QSSL and their OS's will be just fine without Amiga.

    As for the clueless people who post things like 'I want this for my next OS!' - Get a clue! The cost alone should be frightening enough to change your mind: 1100$ USD for a runtime license and another 2000$ for the development tools, and even then it is only interesting if you are going to cut code for it.

  • Why was this moderarted down? (Score:2)

    by Anonymous Coward
    Why are all these wussys moderating down any post that serves up an opinion stronger than a Shirley Temple cocktail?

    Due to over-moderation, slashdot has become as unstimulating as a cold day old bowl of Cream of Wheat [nabisco.com].

    The most frightening aspect is that there exists these slashdot femme-men who are so frail as to be offended by Wakko Warner's pithy observation that holding out hope for the Amiga is akin to trying to will a rotting corpse back to life.

    These moderaters are like toothless old women whose only meal can consist of well gummed Wonder Bread which has been further softened by a soaking in tepid tap water. Do colleges graduate anyone with guts anymore? Or are they all sheep?

    And for the record, I don't happen to agree with Wakko Warner on his point; I'm merely shocked that anyone would feel the need to moderate down his comment.

  • This guy is basically full of shit.

    A big problem is that the TCP/IP code in Linux is not BSD-derived

    I can't even dignify that with a response. Arguing with people like this is a waste of effort.

    The only valid point he makes is that there isn't enough parallelism going on in Linux's network stack to scale well beyond 4 CPU's. But, he only touches on that briefly before spewing all sorts of bile about how Linux is holding back new protocols and "curbing performance on the Internet". Note to self: find out what this guy is smoking, and avoid it all costs.

    BTW, comp.sys.amiga.sys isn't even a valid newsgroup on my ISP's server. But, that's just my ISP being stupid, perhaps not a lot of people are arguing with this guy because the people who could argue are over in comp.sys.linux.*, or even avoiding the Usenet wasteland altogether.

  • I know that Linux's TCP/IP stack is not BSD-derived. That's not the point. The point is that it's not BSD derived is not a problem, which is what the article a was quoting from was stating.

  • Well, the alternative is to have the graphics drivers in a privledged user space program (X), which is just as capabable of bringing down the whole OS. Pick your poison.

  • I really feel bad for people that still hold out hope for Amiga. It's like trying to will a rotting corpse back to life.

    Anyway, can we maybe only post Amiga news if it's actual news from now on? As in, an actual product being shipped? (I have a feeling that we won't be seeing any Amiga news anymore if this happens, though.)

    - A.P.
    --


    "One World, One Web, One Program" - Microsoft Promotional Ad

  • This article [news.com] on news.com has some info about where Amiga is actually headed -- basically it's to become another WebTV information appliance.

    - A.P.
    --


    "One World, One Web, One Program" - Microsoft Promotional Ad

  • Just don't tell "scanner" (my 486 - serves as a host for my scanner and X10 firecracker)that. Or at least two of the webservers here at Clemson - they're both P75s. :)
    I'll have to make sure I don't read this thread at home. taniwha (my 386dx33 firewall running 2.2.6) would be most upset to realise that it's not supposed to be able to run linux. oorodina (486dx66, was, until friday, my main box) would be a tad miffed too. However, rusalka (dual celery 300 (haven't o'clocked it (her? a rusalka is a Russian ghost of a drowned girl) yet) would be a little confused as it knows it's not connected directly to the 'net, and yet it gets good service from taniwha.

  • Amiga used to be a bunch of intelligent, smart, and witty people. Where are you now? I used to converse with people of copper chip timings and such. People used to share 3D rendering code, and hal optimizations with me. Now a bunch of crazy people are posting stupid, un-intellegent, pro-ami , bad linux. What gives?

    Where are we now? Using Linux, BSD, stuff like that. I was an Amiga loyalist for years, yeah, hardware-bashing assembly-writing democoder, there was no way I was going to switch to a PC when the hardware was so much worse and the OS was so much worse.

    But times started to change.. with Commodore's slow slide into the mud, the hardware stopped getting better, the OS stopped getting better, but I still stuck with it, even when people were buying much more powerful PCs for half the money, 'cos I used Windows at work and didn't want to go anywhere near it at home.

    Then I discovered Linux. An OS as good as AmigaDOS, running on that stupidly cheap PC hardware, and with a community full of keen hackers just like the old Amiga scene. So I bought a PC, and packed the A500, A1200 and A3000 away. Not without regret, believe me, but their time was just past.

    And after spending literally years in the mid-90's believing in an imminent Amiga resurrection, I can say with as much confidence as I can say anything, that it's just not going to happen.

  • Amiga fans trivialize what was a good hardware/software platform by assuming that the current owners of the trademark have anything to do with past success.

    The first Amiga was a standout because of it's advanced hardware and OS at a bargain price. This opportunity no longer exists; hardware is a commodity item, and advanced operating systems are free (speech and beer).

    What hasn't changed is the desire to make a computer better than anyone has before. Something "insanely great", as Steve Jobs would say.

    Gateway is in a better position to do this than Commadore ever was. At least they've demonstrated their ability to successfully run a business over the long haul, which is something Commadore had problems with. As long as Amiga remains atonamous, they have the chance to do something great.

    I'm dismayed at the number of /. posters (not to single you out in particular) who are predicting certain failure for Amiga, and seem to have no vision for what can be done. Someday someone will create a better computer, but it won't be done by people who are living in the past.

    TedC

  • The Amiga 1000/500/2000 used the MC68000, which didn't even had an MMU. Sort of hard to enforce memory protection without one, wouldn't you say?

    TedC

  • But the best reason NOT to use the new Amiga is that many of us got fscked over when Atari/Commodor/TI killed our favorite computer. The reason I work on Linux is that this cannot happen.

    Excellent point. It will be interesting to see what Amiga plans to do with their software licensing, releasing specs on hardware, etc. Strict proprietary licensing could really dampen my enthusiasm.

    TedC

  • Funny thing, freeBSD's TCP/IP stack suffers from the same malady as Linux's and would have fared no better in the Mindcraft benchmarks. Serveral major, modern commercial Unices also have or until recently had this problem. Reason is that under real world operating conditions the TCP/IP stack is not the bottleneck, ethernet bandwidth is. Only many separate simultaneous requests and the file size requested is vanishingly small can you overwelm the TCP/IP stack and take a performance hit. This is why Microsoft insisted on very small static pages and multiple 100Mbit cards for the test, that was the only way they could get the number of requests high enough to hit the limits of the TCP/IP stack before they hit the limits of the ethernet protocol (or some other weakness in NT itself).

  • Yep. I remember that fact from my Amiga days. The Miami TCP stack for the Amiga is CRIPPLE-WARE folks. You now know the *REAL* reason why H.Kruse has a grudge against the Linux TCP stack. He basically couldn't steal it.
  • Sorry, but quite simply not true.. A GPL program can make calls to a NON GPL program without much of a problem. This is why vendors can make closed source drivers, as long as the code for the driver is 100% their own.. The kernel can call it just fine, and no need to swallow the GPL..
  • The problem is, many have never read the GPL, and probrably won't, becouse then they'd actually KNOW how friggen idiotic they where over the last 2 years..
  • None of the "It's OK to use Linux as the Amiga's new kernel" posts have managed to address one of the key reasons for using QNX as the kernel.

    Real Time

    The only way to succeed in a true multimedia OS is to have deterministic timing. This allows you to create awsome effects hardware like the Video Toaster (lack of deterministic timing is why it's taken so long to see an adequate replacement for the toaster on newer OS's), and powerful media software like Scala. The sort of thing that Amiga Inc were talking about reviving. Unfortunately Linux doesn't address these problems - even with the RT-Linux patches (which are a good start). It seems to me a big waste of Amiga's time and resources to hack in real time code back into the kernel, when they had a great (possibly the best) real time OS waiting to be used.

    I wait with interest to see what they come up with, but I'm not selling my Linux box down the river yet, I can't wait for the WindowMaker theme to come out that emulates the LAF... :)

    Matt.


    perl -e 'print scalar reverse q(\)-: ,hacker Perl another Just)'
  • Ok, some factual errors.

    First, there is plenty of IPv6 software for Linux, over on ftp.inner.net.

    Second, there are IPsec implementations for Linux over on Free S/WAN and NIST.

    Third, Linux had IPv6 support before Microsoft's experimental stack. I started using IPv6 with no problems as of 2.0.28, when the first patches came out.

    Fourth, Linux supports a wide range of multicast protocols, including PIM. You need to patch the BSD kernels for such support.

    Lastly, Linux has support for QoS, including RED, CBQ, RSVP. I've only seen an Intel experimental version of RSVP for Windows, and it's not that impressive.

  • Everybody seems to be gushing how "this means Amiga will run Linux apps out of the box."
    Of course, that won't happen unless they're using X, the same libraries, etc., etc.--in other words, if it is "just another distribution."
    The best info I can garner is that the Amiga will just be running the kernel, although nobody's clearly stated that yet. Does anyone know?
  • Just because he knows about TCP/IP doesn't mean he knows anything about Linux, and also doesn't mean that he's not just spouting off. So he's a well-read crackpot who is... :P
  • Anything _but_ the kernel is an application. An OS is made up of a kernel and a bunch of applications (plus a little glue, maybe.) GNU tools, X window managers, whatever, are all applications.

    --

  • Holger Kruse:
    Linux is not embraced by the academic community in the same way as many of the BSD-derived stacks (e.g. Solaris)

    Well, not according to this article [lbl.gov] which says Solaris is a reimplementation that doesn't use BSD, just like Linux's IP stack.

    It also says Solaris isn't very good at sticking to the standards, rather like Linux 1.0, but unlike Linux 2.0.30 and 2.1.34 which are pretty good (and I bet 2.2 is better).

    I wonder whether Holger Kruse ever mailed any bug reports to the Linux kernel folks. I think that would be simpler than implementing workarounds in his stack.

  • I checked out the RFC refered to by Kruse [sunsite.auc.dk]. It refers to paper by V. Paxon [lbl.gov] that details severe problems with the Linux 1.0 TCP stack. However, on page 14 they describe the tests they made with 2.0.30 and 2.1.34, and most of the problems seem to have been fixed. They found what looks like a few minor issues that they communicated to the Linux people. They thank Eric Schenk, David S. Miller, Craig Metz and Alan Cox for their assistance in the acknowledgements section.

    This article is the major reference for the RFC and is written by the same guy as the RFC. It also has a lot of tough criticism of other systems, including Solaris and several BSD-dervived stacks. Windows gets a fairly clean bill, and they are very critical of Trumpet Winsock.

    I tried to check Dawson's paper [umich.edu], but his server seems to be down.

    For the other problems in the RFC they were either clearly marked as BSD problems, or I couldn't follow up the references. (Either because there weren't any, or because they were paper and not online.) The RFC doesn't name names, so it's impossible to say which of the others Linux has been guilty of, or is guilty of.

    I think Holger Kruse should tell us what his 4 workarounds are, that he has been forced to put in to work around Linux. Linux has plenty of stuff put in to work around other people's mistakes of course. I guess having to put that sort of thing in your code can make you arrogant.

  • Why? Jutify "QNX is a horrible operating system"

    It is *very* lightweight and *very* stable, two things that all desktop operating systems at this time are not (with perhaps the exception of BeOS but lets not argue about that one right now) It is stable to the point that it can run for years, and survive drivers misbehaving and other *core* system components being changed. Linux is pretty good, but not *that* good.

    Unfortunately with QNX they were going the right way, but I feel that this decision to use linux is not the right one (even though I love Linux) - Note that the decision appears to have been made by management and as such is hoping to ride the current Linux [hype/bandwagon/frenzy] that appears to be in evidence.

    ~Pev
  • OK first off: OSS/Linux. The sound drivers that (probably about) 10% of Linux users use are not released under GPL. There are probably hundreds of other kernel modules out there that are proprietary and/or GPL-unfriendly. Linus made a decision a long time ago that proprietary modules can exist.

    Second off:
    . . . all of which are separate and distinct entities that cooperate and work together. I don't see this distinction where the new Amiga OS is concerned
    Neither do I. If you can create "entities" that sit on top of Linux which are neither kernel modules, programs, nor libraries, then I suppose you might have a point in there somewhere. Unfortunately, you can't, and thus, you don't :). X is no different from Netscape which is no different from false(1) -- they are all just programs which make up the operating system, and they can all be released under whatever licence they want. The game you want to play is trying to draw a line between "applications" and "the OS". Is X part of the OS, or is it an application? How about bash? perl?

    Whatever Amiga does to make their own new operating system, they're going to have to do one of the following:
    - make changes to the kernel. They will have to make their changes public according to the GPL.
    - make kernel modules. These can be proprietary.
    - make cool new GUIs (as a replacement for X). This can be proprietary, as it is just a program.
    - make a cool new API to replace libc. This can be proprietary. It is not against the law to port Motif to Linux; proprietary libraries are A-OK.
  • Lightweight and stable?

    How about underpowered and useless! RTOSs don't
    even have compilers; you have to write everything
    in assembler. How sucky is that?
  • I'm not sure how current his information is; for example, he states that there's no IPv6 and IPSEC support for Linux. This is incorrect; IPv6 was added in 2.1, and IPSEC was recently announced (though it's not in the kernel distribution).
  • There's an obvious dividing line: the system call gateway. Anything that implements a system call is part of the kernel; anything else is userspace and hence is unaffected by the kernel's licence. By this definition, glibc is an application just as much as Oracle is. (It's just an application used by practically all other applications. You can envision writing a system call interface for another language, such as Java or M3, and then writing the userspace code completely in that language, with no C code or library at all.) Amiga can therefore build anything on top of the kernel that they want, and can use any licence they want for the components.
  • Bruce is just an old fart -- he's talking about the pre 1.0 days when the Linux kernel was distributed by punch card, and fit in your hand to boot.

    :-^

  • Speaking of FUD...

    Ancipital's legendary for spreading flat-out lies in comp.sys.amiga.misc - I see that he hasn't changed.

  • The Miami TCP stack for the Amiga is CRIPPLE-WARE folks.

    I was unaware of a definition of cripple-ware that included fully functional and working software. The fact that it has a one-hour session timeout (with a working auto-reconnect feature) doesn't jibe with my idea of crippled.

  • Holger Kruse seems to know what he is talking about.

    To put this in context, Mr. Kruse developed Miami (Modem Internet for AMIga) and Miami Deluxe (a full-featured TCP/IP stack with routing, IP-NAT, etc. built in). More info can be found at Nordic Global [nordicglobal.com].

    I only mention this to say that Holger's not some half-read crackpot who's just spouting off - he really does know what he's talking about.

  • All this trash-talking of Linux is not exactly endearing them to the Linux community. And from where I'm sitting, it looks they need us but we're managing perfectly well without them. It's annoying to me that people who don't even have memory protection, security, or built-in networking in their operating system see fit to lecture us on OS design. I guess they've been keeping themselves going on nothing but belief in their own innate superiority for so long they've completely lost touch with reality. I hope they stick with QNX so we don't have to put up with them.

    Enough is enough. These Amiga articles have reached the point where they're not funny any more. I'm gonna configure Slashdot not to show them.

  • Real time OSes are a stupid way to do audio/video stuff on a commodity computer. The problem is that just having "Real Time" doesn't necessarily mean that your computer is fast enough-- and the tradeoffs are that your machine may become useless for anything else.

    Yes, RT doesn't mean your computer is fast enough, but that doesn't make RT useless.

    The right way to do A/V stuff is with special (intelligent) hardware that works in real time itself (under the control of the machine), and takes the load off your CPU (and operating system).

    That makes sense when the A/V hardware is cheeper then a new CPU for the task (like MPEG2 decode use to be, and still is for some people). It is much less fun for things like MP3 decoders, who wants to pay $50 for MP3 decoder hardware rather then kiss off a few percent of their CPU?

    If you have a nice fast CPU would you feel good at paying $100 for a MPEG2 decoder, or would you rather kiss off 40% of your CPU when you happen to watch a MPEG stream?

    Intel, AMD, Sun, and Compaq spend megabucks on research for their nextgen chips. Should Creatave Labs try to compete?

    Besides most A/V stuff needs relitavly little in the way of real-time garentees, for 30fps video it "only" needs to handle 30 "hard" real-time intrrupts (per second), allow dirrect screen I/O from the real-time task (or have a RT window system, which could be hard), to buffer enough data to smooth out the jitter for non-real-time I/O with a network or disk. Most A/V stuff can be converted to "easy" RT (not soft RT), and big buffers.

    Note: "hard" Real Time has real garenttes on how long intrrupts take, and what can intrrupt a RT thread. For example "no more then 200ns from intrrupt to thread dispatch, no interrupts from lower pri until the thread releases the CPU", these limits are never ever violated, except in the event of a hardware failure. Sometimes valuable equiptment depends on it (computer controled milling machines, or chemical processors maybe), sometimes lives (flight control software in a inharently unstable airplane like the Stelth). "Soft" Real Time makes similar garentees, except they can break them once in a while (say when the disk decides to do a thermal recalabrate). These are industry standard terms. "Easy" real time is not a standard term, but i use it to talk about a system that allows only a subset of operations in a "real time thread", like adding things to a buffer, and taking them out of the buffer, and maybe some non-complex I/O (like looking at the current value of a DAC, or causing a pin to go high or low, NOT talking to a filesystem). It's no where as difficult to do an "easy" RT OS (or extention) because you only have to deal with the intrrupt handler and scheduler, and the small number of permited RT operations. You don't have to fix all the filesystems, all the device drivers, and all the anything else that has non-determinstic timing, or deterministic timing, but too long to make useful RT promises.

  • i think you never saw QNX running? it's small, very very fast, developing for Photon is so easy also!
    Message passing is great, it existed in amigaOS in a way also. QNX kernel is developed for 20 years, and it rocks, i use it all the time at work, i don't want to use a monster like linux!
    --
    http://www.beroute.tzo.com
  • Neutrino works on x86, PPC, SH3, SH4, arm/strongarm, and will run on alot of others microprocessor. The kernel is about 40Kb, and only 10% of this kernel is microprocessor dependant, so rewriting 4Kb of code is easy for QSSL.
    I agree QNX/NTO is not made for the end user, but for embedded appliance and realtime critical stuff, it rocks!
    The guy who wrote the article knows nothing on QNX, i suggest him to contact QSSL so they can show him a demo. All the QNX stuff in this article is bullshit.
    --
    http://www.beroute.tzo.com
  • note that the next TCPIP package for QNX (TCPIP4.4 in alpha right now) is based on BSD4.4 and will have all the funny stuff like NAT or whatever.
    note also that in BeOS the TCPIP stack has been written by only one person!!! ok it lacks some raw socket interface (you cannot use traceroute e.g.) and the performance on a 100Mb are poor, but wrote a stack alone in one week is great :o)
    --
    http://www.beroute.tzo.com
  • i said the NEXT tcpip package, actually it's v4.25, the v4.4 is actually in alpha stage, and will have NAT and whatever.
    --
    http://www.beroute.tzo.com
  • "Having no (or limited) memory protection is very popular among microkernels. AmigaOS isn't the only such microkernel, and it probably won't be the last. Because message passing is a lot easier without memory protection.

    However, that kind of "OS" is a toy. You might as well run DOS or Windows."

    -Linus Torvalds October 1996 on comp.os.linux.development.system

  • | Today on comp.sys.amiga.misc, we've had
    | regulars complaining that linux was so
    | ludicrously bloated that it couldn't run on a
    | p100

    And this was on c.s.a.*misc*? Things have indeed gotten bad down Amiga way. Such silly prattle used to rarely make it out of c.s.a.advocacy in the old days.

    Just don't tell "scanner" (my 486 - serves as a host for my scanner and X10 firecracker)that. Or at least two of the webservers here at Clemson - they're both P75s. :)

    I'll have to come out and agree with the sentiment that some others are posting here: Amiga *announcing* this and that just isn't news anymore. All that's come out of anything officially connected with the Amiga for *years* has been vapor. Vapor!

    I've owned various Amigas since 1987, though the A3000 and A500 I still own are boxed and in the attic at the moment. I quit the roller coaster in 1996 and started using Linux and Windows to get my work done. Now I just use Linux at home. I do still like to read about Amiga news occasionally, though. But who really cares which kernel AmigaVaporOS will be based on?

    Though hearing an AmigaOS user say that *Linux* is unstable really does crack me up ... :)
  • You young whipper-snappers have no regard for history :-)

    There was a time when the entire kernel source in .tar.gz form fit on one floppy and Linux didn't network. Then there was a time when it didn't network well. Then it networked well, but slowly. Then it was the fastest monoprocessor TCP/IP implementation. Then it networked well on multiprocessors, but slowly...

    Get the point?

    Thanks

    Bruce

  • I know how long QNX has been around, and I know it's a stable embedded-systems kernel. No argument there.

    I don't think that free software precludes a tighly focused group of engineers, though. There is lots of evidence for the contrary, especially where the Linux kernel team is concerned.

    I agree that X is moribund, but of course there are free software efforts like Berlin to replace it on Linux. I don't care much that the Amiga won't run X, as long as you can port GDK and CORBA you can run the GNOME tools.

    Thanks

    Bruce

  • I'd heard something like that, but it was so easy to attack the message that there was no need to attack the messenger.

    Thanks

    Bruce

  • Why Amiga switched (Score:3)

    by Bruce Perens ( 3872 ) <bruce@perens.com> on Monday July 12, 1999 @08:21AM (#1807236) Homepage Journal
    The Amiga folks switched operating systems not because of any present technical superiority, but because of the future. If there's something we don't like about Linux, it gets fixed. If you don't like how QNX works, you're stuck. Even if we get the QNX source code, who will want to put lots of free work into this proprietary product? That's why they chose Linux - because it's evolving so much faster than proprietary systems.

    I found the criticism of Linux' TCP/IP rather laughable, because I remember not very long ago one of the primary criticisms of Linux was that it didn't have TCP/IP. Whatever you don't like about Linux' TCP/IP implementation will be fixed soon enough. Meanwhile, it runs fast enough to saturate my 768/768 DSL while the Pentium 120 CPU is loafing along. I can wait a year for it to be fast enough to saturate a DS3. The criticism regarding other OS having work-arounds to interoperate with Linux applies to Linux 1.x (or other old versions), and of course those problems are long gone (along with the FIN_WAIT problem he mentioned). The criticism about lack of IPV6, etc., is bogus, he's not been keeping up with Linux' development.

    Someday, Linux development will taper off and we'll switch to another system - maybe even a message-passing system like QNX. But the system we switch to will be free software, because of its fast evolution, not proprietary like QNX.

    Thanks

    Bruce

  • You noted the same thing I did ... the person who wrote the article was rather clueless about the whole issue.

    My impression was that this article said a lot without saying anything.

    /dev
  • Read your /usr/src/linux/COPYING.
    --
  • Andrew Korn is clueless about almost every technical issue. He used to edit CU Amiga, and would treat PR and manufacturer benchmarks as fact.
  • They could make it a module, and Linus made specific exemption for binary-only modules.
  • I think there were Agnes and Veronica, among others, that handled special tasks.

    The A500, A1000 and A2000 had these:

    • Agnus (for Address GeNerator) - contains blitter, graphics coprocessor (copper) and carries out all memory-register transfers for custom chips
    • Denise (for Display ENabler?) - handles video output (including colour lookup, HAM (Hold-And-Modify) mode, and sprites), collision detection, and game control interfaces (including ADC)
    • Paula (for Ports and AUdio) - handles serial, disk and audio I/O
    • Gary (for Gate ARraY) - contains various board-specific logic and chip selection

    Later Amigas had various extra or replacement chips with names including Ramsey, Buster, Alice, Akiko, Lisa, and Gayle.

  • First of all you're quite ignorant as to how different systems work. No kernel that I know of has a GUI in it. That includes NT's. NT has a component of the kernel called GDI which provides very limited graphics in the kernel. That makes things a lot faster, and really isn't the reason for instability. The GGI project in the linux world has similar goals, and IMHO, they are on the right track. Now the problem with NT is the windowing system. I don't have my roomate's copy of "Inside Windows NT" with me to quote the dlls and exes, but basically they way that it works is that the system boots, the GUI starts. However, if the GUI dies, the startup processes tell the kernel to die, because the assumption is that you're pretty fucked without that GUI. It would be like init under linux monitoring X after X starts and shutting down the machine if it finds X isn't running. This is a very poor design, but it has nothing to do with the GDI layer in the NT kernel.
  • This is true, but I don't see the problem. Let's take my system for example. I have 4 PCI cards in the system. The network card's driver is in the kernel. The SCSI card's driver is in the kernel. The Sound Card's driver is in the kernel. The Graphics Card's drive is in user space. Why? Beats the hell out of me. Sure the graphics card driver could take down the kernel, but so could any of the other ones.
  • a good debunking of the whole thing. Thanks.
  • So, I'm assuming that the components that make up the Amiga Operating Environment won't be GNU, but will they be open-source in some way? Or are these guys just interested in what Linux can do for them, and won't be contributing any back to the community?

    Anyone have any news on this?
  • That's okay, because Slashdot moderators, just like Linux users, wouldn't know their asses from holes in ground.
  • Funny my ass. It is truth. Slashdot moderators and Linux users are dumber than goat shit. I've said that many times on here.

    I mean really, look at what Linux users say on here and compare that to how the moderators score things.
  • QNX doesn't need to prove itself. QNX systems have been up and running since 1994 non-stop. QNX is used in nuclear power plants. I don't think I would trust my life to linux, however. QNX doesn't need to evolve faster than linux because it's already light-years ahead of it. The linux development model is ok for fixing minor bugs and features here and there but doesn't seem to work when major overhaul is required. For that you need a tight focused group of engineers-not 5 million people from the internet. That's why linux GUI's (there is no standard GUI for those concerned with consistent look-and-feel) are still based on x-windows. On the other hand, if this Amiga Linux won't be based on x-windows, people will complain that all there other linux apps and graphics card drivers won't work on it. Forget about the LSB that's for sure.
  • I'm not so sure. The GPL specifically forbids opening up only _part_ of your source code -- that's the viral nature part. If you can prove that an OS is not "one big thing" then I guess you can justify doing that, and admittedly an Operating System is a pretty vague concept, but the kernel is _part_ of an OS, not separate from it, so if you use a GPL'd kernel in your OS the OS is required to be free software b/c of the GPL's viral nature.

    I'm going to read that !@#$% license again...
  • Oracle, Informix, Sybase, DB2, have APPLICATIONS that run on top of Linux. Under the GPL, the viral quality of the GPL is exempted if an application is making a call to an operating system.

    I see a distinction between that and the rest of the Amiga OS. What they claim to be doing is using the Linux kernel as a foundation for the rest of the operating system -- since it's the Amiga OS, presumably a graphical operating system -- which is not the same as an application.

    Different parts of a Linux distribution can be placed under different licenses because Linux is not one operating system per se. You have a kernel, a few different shells (bash, etc.) a windowing environment (X) all of which are separate and distinct entities that cooperate and work together. I don't see this distinction where the new Amiga OS is concerned. I can't see how anyone would be able to classify "every part of the OS except the kernel" as a distinct application that runs on top of the Linux kernel.

    If that is really the argument made, I think it would seriously weaken the legitimacy of the GPL. But that's just my opinion, and, as always...

    ... never discount the possibility that I am WRONG.

  • Proprietary software can exist on Linux b/c there is an exemption in the GPL which allows proprietary code to make necessary calls (or whatever the terminology is) on a computer -- rough translation: a proprietary app can run on top of a GPL'd operating system. If the proprietary app tried to modify the GPL'd OS in any way, well, then there might be a problem.

    The reverse is also true: a GPL'd app can run on top of a proprietary OS.

    The problem I'm having is that I don't see the same parallel between "kernel" and "the rest of the OS" that I see with "app" and "OS in general." I just don't see it. I freely grant that it could be b/c I'm not a programmer, but it seems to me that the relationship between an OS kernel and the rest of the operating system proper must by necessity be a lot more intertwined than an application and an operating system.

    But, as always, never discount the possibility that I am WRONG.
  • Well, I guess I can't argue with that. :-)

  • I see what you're saying, and I'm just about ready to concede defeat, but I want to get a few things clear first. :-)

    Let's take the Macintosh OS for a second: while I suppose the UI is just a program sitting on top of the kernel, it certainly doesn't seem to be sitting on top of it in the same manner that X and the X Window Managers are sitting on top of the Linux kernel! My point with the "distinct entities" comment was that X and the Window Managers and Bash and all that other stuff aren't really integrated into the kernel, so they could theoretically be distributed under any licenses they want. That seems to be your point as well, with the additional point that the Amiga OS will be the same way.

    I'm wondering if it will, though. The impression I kept getting about the AmigaOS was that the UI was very, very tightly integratd into the OS proper, similar to the MacOS, so I wonder how possible it would be to meet both conditions: to be both tightly integrated and just a kernel module, or a layer on the kernel, or what have you. I'm not a programmer, this could be the easiest thing in the world to do, it just doesn't sound like it from my perspective.
  • I'm not a recognized expert on the GNU General Public License, but methinks I see a problem here:

    The GPL, which the Linux Kernel is distributed under, is viral. Any code that makes use of the Linux kernel must, therefore, be distributed under the GPL or a license that is compatible with the GPL.

    Also, if the GPL'd software makes calls to other software or libraries, those software or libraries must also be distributed under the GPL or a GPL-compatible license. The only exceptions to this are, basically, operating systems.

    So how will this new Amiga OS be licensed? If they really are going to be using the Linux kernel, then it _must_ be distributed under a Free Software license. I can't see any way around it. I find it difficult to believe that Gateway would be willing to do this, my opinion is that they either a) haven't thought this through, or b) think they've found a way around it.

    I think everyone should watch this very closely, and this question ("what kind of license?") should be asked, very pointedly, over and over again, until they answer.

    This could be a very important precident for free software - if Amiga thinks they can use kernel technology licensed under the GPL without in turn freeing up their source code, they need to either a) be convinced otherwise or b) be taken to court. If option b), then we'll see exactly how legally binding the terms of the GPL are.

    Then again, never discount the possiblity that I am WRONG.

  • I've got to agree with this. The last thing we need to do is bloat the kernel with unnessary code. There will be no way to bail if the GUI crashes and when it does it will most likly take the whole machine with it.

  • I've re-read this comment several times, and still have no idea what it means. Try using correctly formed sentences, with the first word capitalized. Include one idea per sentence, and spell all of the words correctly. Use correct punctuation. Then all will understand what you are trying to get across.

    P.S. Typos are fine (we all make those!)
  • Samovar awards, July [sunysb.edu]

    Amiga announced recently it is dropping one of the best modern OS's QNX to join the crush of Linux supporters. Although Amiga is right that this market is going to be huge, but it will compete directly with other Linux vendors like IBM, Dell, Compaq etc., etc. with the real risk to loose its unique face. The president Collas said: "PLEASE, PLEASE, PLEASE do not to judge the Linux decision until you have a chance to read the technology brief". Ok, no award for them. And welcome on board, sure!

  • Can someone *knowledgeable* (other than the author) please give their take on the article linked to above. If all the implications are true, its no wonder Linux lost the Mindcraft benchmark tests.

    Simon
  • As long as you don't replace ALL employees at the same time, something remains. New people assimilate the values and the spirit of the team and pass it on.

    Let's assume that the last engineer who actually worked for Commodore left 5 years ago - I'm still sure that if you step into their development facility and speak with the people for a while you will feel something about the atmosphere which is not Mac, not Linux/Unix and definitely not Microsoft. And it's this kind of thing that makes great software.

    Even if you ignore the actual code base, there's definitely something left.
  • I found the criticism of Linux' TCP/IP rather laughable, because I remember not very long ago one of the
    primary criticisms of Linux was that it didn't have TCP/IP

    Huh? I've been using Linux for five years, and it's had TCP/IP for at least that long. In the computer world, five years is an eon, not "not very long ago"

  • As the Ball Bounces (Score:3)

    by eponymous cohort ( 8637 ) on Monday July 12, 1999 @07:47AM (#1807261)
    The new soap opera starring an OS which has been down on its luck that decides that if it just gets a makeover, and dumps its fiance for the hot, young penguin chick that everyone wants to be seen with, things will turn out ok.
  • Um, surely they'd just have to open up their source for the things they do to the kernel - but as an OS is much more than just a kernel, they could keep the rest proprietary.

  • Isn't the whole argument here that graphics in kernel space means drivers in kernel space, and that means a bad driver can bring down the whole OS?

    And in that case, whether Linux or NT, it seems like this could be a mistake. The main difference, of course, is that with open source Linux driver bugs are bound to be more efficiently fixed.

    D

    ----
  • It was very informative - I couldn't have said it better.

    Its expected that the author of any software will always defend and promote his/her creation. Holger Kruse is an extreem example of this as he has conveniently disregarded the changes in the Linux networking code from the 1.x days to the 2.2.x kernel in order to promote his implementation. The fact that he is not aware of the work being done on IPv6/Linux shows that he really has no knowledge of the Linux TCP/IP stack and his arguments are thus nullified.


  • Wasn't one of the last few real Amigas a convergance device? I belive it was called the CD32, and was sold as a game/multimedia machine. Anyway, it's a market that makes sense, although certain companies (ahhm, Microsoft) are already positioned there.

    Either Amiga wants to try to infiltrate the entire computer market en-masse, ...

    Well, the general computer market has grown one hundred times over since the Amiga was sold. Most people using computers now days have never heard of "Amiga", and a good portion of those who have wouldn't even bother. (Same goes for standard Linux, by the way.)

    If this thing is marketed towards the general computing market, they're dead even before they started.

    --
  • I like the Transmeta spin. QNX was cool, but Linux is cooler. X86 is cool, but mythical Transmeta products are cooler. Maybe it will run the Linux kernel on top of the magic chip from the Linus company.

    Did this sound a little contrived to anyone else?

    Does NG Amiga stand for "next generation" or "newsgroup" or something else?

    I think newsgroup Amiga is the most fitting. As in, the only place you'll ever see one is in newsgroup discussions.

    alt.vapor.amiga.hardware and alt.vapor.amiga.software

  • what's "Amiga" about it? (Score:3)

    by jetson123 ( 13128 ) on Monday July 12, 1999 @11:12AM (#1807267)
    The original Amiga was a great machine at a great price. But what about the new (planned) one? It's a different OS, it's a different set of programmers, it's different hardware, it's different owners, and it's probably a different user community by now, too.

    That's not necessarily bad. But what all this amounts to is that a division of Gateway is thinking about building a multimedia computer that uses some of the Linux kernel. The fact that it's called "Amiga" seems pretty incidental to me.

  • When did all the berzerkers and amongers come from and why are they all over /. RANTING about UNPROVEN, BIASED, STUPED FUD!! If you're here on /. for some damn crusade, then go fsck yourself. This is a FORUM for THOUGHT, comparison, and intrigue.

    I am a BSD and Linux user.. they WORK, and I also own a few old amigas which are OBSOLETE. (Cause there arn't any new ones to speak of). I have a 600MHz Alpha that rocks the universe.

    Amiga used to be a bunch of intelligent, smart, and witty people. Where are you now? I used to converse with people of copper chip timings and such. People used to share 3D rendering code, and hal optimizations with me. Now a bunch of crazy people are posting stupid, un-intellegent, pro-ami , bad linux. What gives?

    And what's with some of you people arguing over the stack?? Excuse me, but your "personal feelings" don't make stacks. Hard work, protocol understanding, awesome timings, and know-how DO. How many people here have ACTUALLY looked deeper than make config?? I am by no means a kernel developer, but I am an app developer. And I look into the kernel ALL the time, because I feel it's important to know what impact my code really has! (For instance ioctl's are very specific, and arn't always in the man)

    CCCCHHHHHIIIILLLL OOOUUUUTTTT and smell the code, ladies and gentleman. /. is alot of things for alot of people. What is it for YOU?

    Pana
  • Definitely flame bait. It's well known just what a profound attachment Amiga users have to their machines... advocacy to a fault in many cases, to be sure. It's stuff like this that leads to "my platform is better than your platform" silliness. I still hold out hope for Amiga, but I don't let it run my existence. :^)

    As for actual products being shipped, well, there have been many instances where I've had to wade through plenty of digital speculation on all sorts of fronts without seeing actual product shipping. While seeing the name "Amiga" turns a lot of people off immediately, I think it's important to the Linux community to see just where the kernel is going, and what other systems might be taking advantage of this OS's advantages.

    *sigh*

    Of course, you can always just customize your Slashdot to remove all of these Amiga posts, if they annoy you so much (good job, Rob&Crew).

  • Chances are nobody is going to state ANYTHING clearly until Jimmy-Boy at AI releases his highly touted "Technology Brief" which, according to the hype, will cause Amiga users like myself to perform our own kind of release in our briefs...

    There's no way to determine if Amiga will run Linux apps out of the box until then. However, I think using the Linux kernel makes for easy porting of apps between Linux and whatever-the-hell-the-Amiga-becomes. Hell... the current Amiga line has emulation libraries (ixemul) which allow for X-apps to be ported to WorkBench (there are a pile of well known Ghostscript-using PDF utilities which do this quite well).

    Unfortunately, as with the rest of Amiga development, the phrase "wait and see" rears it's ugly head.

  • I can assure you that the only thing they care about is the $. Gateway's stated reasons for buying Amiga were to increase their intellectual property portfolio (ie they wanted Amiga's patents). There were indications that Gateway was unaware of a the active Amiga community until after the purchase.

    Yes, this has been more than slightly obvious to anyone who has been following Amiga's multiple acquisitions/bankruptcies since the Commdore days. Gateway executive types have always stated publically their amazement at the magnitude of the Amiga userbase ... particularily in the absence of a company. The general jist seemed to be that Gateway was originally geared to grab Amiga's chipset and such for their Destination PC-TV multimedia computers.

    But that's history.

    Jim Collas has done everything short of slapping Ted Waitt in front of the cameras. He has said over and over again that he can't understand why Gateway was spending so much time spinning their wheels with the Amiga project. Now he's off and running trying to get the job done, and I don't think he's doing too brutal a job (the current QNX/Linux debacle aside, of course). There appears to be more in his agenda than satisfying Gateway's need to make a quick buck on the back of the Amiga.

    A lot has happened since February when Collas took the throne. More than I've seen in Amiga development in a long time. The fact that changes apepar to be taking place... even in the environment of vague press releases and apparent secrecy... leads me to believe that something is actually getting done.

    ANYTHING GETTING DONE IS A NICE CHANGE. :^)

    Will their direction be a technological epiphany? Even my overbearing Amiga advocacy shudders at the thought of making such predictions. I want them to deliver, dammit.

  • Anges, Denise, Paula... etc... etc... etc...

    Again, that's history.

    Amiga Inc. has made it clear that there's no point reinventing the wheel. Back when the original Amigas were released, anyone could put together reasonable AV chips. But now, there are companies whose entire existence is devoted to the production of high performance chipsets. Even Amiga can't be so egomaniacal to believe they could out-do companies like Creative Labs and 3DFX. Likely their new hardware architecture will be modular enough to accept these third-party chipsets, as well as permitting a higher degree of upgradeability than the Classic Amiga "We Solder All The Chips You'll Ever Need Directly To The Motherboard" mentality.

    The Amiga is more than just a bunch of custom chips. Their OS was waaaaay ahead of its time... so much so that junkies such as myself still swear by it. It's got a much more straightforward filesystem than the somewhat dated offering of *N*X. etc. etc etc. I'm not writing this to pat Amiga on the back (instant flamebait?) but to point out that there's more to Amiga's charm than just the hardware.

  • What happened to not posting Amiga news until a product started to ship, Wakko? :^)

    Okay, that was a cheap shot...

    As for that article, the Amiga isn't just going to become another WebTV information appliance. That's an easy assumption to make based on articles like the one you read, but that is all Gateway has talked about since they got their hands on Amiga ... multimedia/TV/convergence nonsense. While Amiga Inc's design plans include information appliances such as the ones mentioned in the article, there are even more impressive plans in the works for more traditional applications of new technology in personal computers, network workstations, video production/multimedia powerhouses, and embedded systems in appliances, cars, etc.

    (please, no more jokes about getting Guru Meditation failures when trying to make toast or hit the brakes)

    Either Amiga wants to try to infiltrate the entire computer market en-masse, or invade as many niches as possible so they have somewhere to hide should another Commodore-esque apocalypse come about. :^)

  • Please remember that most people that still advocate the Amiga aren't the brightest bulbs in the box. A lot of them are a bit bitter because they can't get any *nix working (despite there being great debian and redhat m68k ports which run on the Ami). A lot of them don't really understand what a kernal IS and what it does, and thus nothing they say should be taken without a hefty pinch of salt.
    Now, that's inflammatory and deliberately provocative if nothing else. While there are most certainly a lot of misguided souls rampaging through the Usenet, it's not confined to the Amiga newsgroups to be sure. A few ridiculous comments made by certain users shouldn't reflect on the whole. Good thing, too, or else I'd be inclined to deduce that Slashdot commentary-writers are overly reactionary, closed minded, and quick to generalize and lay blame.

    I know better than that.

    As for Amiga users not really understanding what a kernel is or what it does, this is the case for a number of users who have never felt inclined to dig into the guts of their machine. The Amiga equivalent of the typical Windows user. Scary, but they're out there. But in my own experience, the vast majority of Amiga users do in fact understand what the hell is going on, and many of them run/use numerous machines on all sorts of platforms... and do so very well. Oddly enough, these are the same users who have vowed to avoid c.s.a.* like the plague because of the open invitation to stupidly mouthing off that Usenet has become.

    Comments by users about Linux bloat (I laughed out loud when I read that), and lethal instability, should be attributed to the individuals making them, and in a perfect world -- completely ignored as undereducated prattle.

    I'm off to go test the other bulbs in the box now...

  • So, I'm an avid Amiga and Linux user. Should I sit in the corner and proceed to trash talk myself?

    Stereotypes are such shit sometimes.

  • With all the stuff they have to do, they won't be releasing anything in a couple of years; so why is everyone even talking about all this

  • The only fault in the story may be in not making it more obvious that it also works as a Transmeta story.

    As one who used to hack around on a C=64 but dream of getting an Amiga with one of those neat hard drives in it, I'm interested in Amiga news.

    But perhaps the most interesting thing about the article is the speculation about Transmeta contained therein.

    -Augie

  • err no mate

    "may also imply development of the Linux TCP/IP code, which is not well regarded as it stands"

    emm

    where did these people come from I thought that the linux stack came from BSD and thats as fault tolerant as it get's

    what are these people on!!

    unless im wrong ???


    a poor student @ bournemouth uni in the UK (a deltic so please dont moan about spelling but the content)
  • well part of what he says is true but some of it I not so sure of


    but it is VERY scarey

    alan cox where are you ????




    a poor student @ bournemouth uni in the UK (a deltic so please dont moan about spelling but the content)
  • yes thats what My isp uses and look Im useing windoze !

    just because you dont like it does it mean that you wont use it work on it ?

    if so you are loseing out of experances that the rest of use are haveing I use win2000 (dev work) and NT as well as HP-UX solaris and IRIX got a problem talk to IT surport

    (-;
    a poor student @ bournemouth uni in the UK (a deltic so please dont moan about spelling but the content)
  • you should do something about your security

    you are NOT anon

    email me if you want advice

    oh well

    johnjones
    a poor student @ bournemouth uni in the UK (a deltic so please dont moan about spelling but the content)
  • yes under my post why did you post it it only sucks up bandwidth

    do something rob
    a poor student @ bournemouth uni in the UK (a deltic so please dont moan about spelling but the content)
  • "Also, I certainly believe and hope that the Amiga people continue to build on the old Amiga foundation - strong technology. I see them as being commercial, but also as a market where people really care about doing the right thing technologically."

    Gateway (the spotted cow people) bought Amiga a while back, and so in my mind, Amiga = Gateway - they are owned and managed by the same billionaire bean-counters. I worked for three years in Gateway's tech support (the filter through which passes all blunders made by the rest of the company) and I can assure you that the only thing they care about is the $. Gateway's stated reasons for buying Amiga were to increase their intellectual property portfolio (ie they wanted Amiga's patents). There were indications that Gateway was unaware of a the active Amiga community until after the purchase. I can remember checking daily to see if there were any new developments with Amiga for a couple of months after the acquisition and being disappointed when I saw nothing.

    Keep in mind also that Gateway is one of those that won't offer Linux on their systems*, won't offer you a system without Windows, and won't honor the EULA. Do you think they will do nice things with Linux?

    *OK, to be totally accurate, while they don't OFFER a Linux system, you CAN buy a system from Gateway with Linux on it. Be prepared to add at least $700 to the list price for the privilege of buying a custom-integrated solution.
  • Once you've read the page linked above you need to read this [flyingmice.com] as well.
  • >QNX is a horrible operating system. I wouldn't
    >use it to run a toaster, much less a desktop
    >environment

    You again, tlewis? Why are you on this anti-QNX crusade, anyway? For a brief moment, I was wondering whether you work for a QNX competitor, but then I remembered from our last encounter that you don't have any OS or embedded-system experience so that seems unlikely. I can only guess that you enjoy trolling.

    As I pointed out last time, different platforms are intended for different environments, in terms of both the "lower environment" of what the hardware provides and the "upper environment" of what the applications/users require. Your phrasing "...much less a desktop environment..." shows your bias; like many people, you seem to consider the desktop to be the whole world. This is just not realistic.

    >Message passing is for wussies

    Speaking of what's "for wussies" how much real programming have you done? Application-level programming is "for wussies" as far as I'm concerned, and that's all you ever seem to have done.

    Message passing is not "for wussies" at all. Look under the covers of any supposedly non-message-passing system - even a processor/memory bus - and what you'll see are...messages! It doesn't make any difference that the messages are processed by hardware rather than software. Now, given that it's all messages in reality, what's wrong with passing messages that represent higher-level semantic concepts than "get" and "put"? Answer: nothing. Message passing is, in short, an essential part of computing.

    Is explicit message passing implemented in software a good basis for an OS? Yeah, it's OK, if one understands its strengths and weaknesses. Message passing is good for supporting heterogeneous systems, long latencies, and robust fault detection/recovery, compared to memory sharing. Message passing can also suck like a black hole wrt performance. Efficient message passing is essential, as is an overall architecture that keeps the number of messages and/or time blocked waiting for messages down. Many message-passing systems (e.g. Mach) have sucked because they had layer after layer of overly-complex interfaces creating more message traffic, and were insufficiently parallel within each subsystem to tolerate message latencies. But it _can_ be done right. Did QNX do it right? I don't know, but I've heard quite a few comments leading me to believe that maybe they did.

    As an aside, I have worked on both hardware and software versions of distributed shared memory. The latencies etc. involved present some very hard problems which to the best of my knowledge have not yet been adequately solved. I'm inclined at this point to believe that - until we understand the solutions better - anything inherently distributed is better off with explicit message passing than with a shared-memory paradigm on top of a translation to the necessary message passing.
  • The GPL, which the Linux Kernel is distributed under, is viral.
    No. A virus infects a host from the outside, and spreads to other hosts arbitrarily. If you want to use genetic metaphors, you might say that the GPL is a "dominant gene" - the "offspring" (derived work) of a GPLed work is always GPLed.
    Any code that makes use of the Linux kernel must, therefore, be distributed under the GPL or a license that is compatible with the GPL.
    No. Commerical Linux apps use the Linux kernel (all Linux apps use the Linux kernel, that's what makes them Linux apps!), and don't have to be GPLed. What is required is that any code that uses source from the Linux kernel must be GPLed.
    Also, if the GPL'd software makes calls to other software or libraries, those software or libraries must also be distributed under the GPL or a GPL-compatible license. The only exceptions to this are, basically, operating systems.
    No. GPLed code on a SPARCstation can make calls to the Solaris libc - that hardly requires that Sun release the source to its libc.

    Folks, the GPL isn't all that arcane of obscure. Read it for yourselves at http://www.gnu.org/copyleft/gpl.html [gnu.org].

Slashdot Top Deals

Programmers do it bit by bit.

Close