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

 


Forgot your password?
Close
typodupeerror
×
Unix Operating Systems Software

Historic "Free Unix" white paper by Larry McVoy 61

Posted by CmdrTaco from the stuff-to-read dept.
greg writes "This is a white paper written by Larry McVoy at Sun Microsystems discussing free Unix software, *.BSD, Linux, GNU and FSF and competition with Microsoft. In the paper Larry proposes the opensourcing and standardizing of Unix in gerneral and Solaris in particular. Whats truly impressive is that it was written in 1993 and is still quite relevant and its predictions regarding Linux were very accurate. Here is the link: " Currently only available in Postscript... HTML is coming.
This discussion has been archived. No new comments can be posted.

Historic "Free Unix" white paper by Larry McVoy More

Historic "Free Unix" white paper by Larry McVoy

Comments Filter:
  • It's been six years since this document was written and other than the Linux/FreeBSD movement, what has changed ...

    What has changed is that Linux has become the sole overwhelming success in challenging Microsoft. You can't walk into most bookstores today without seeing Linux CDROMs and texts. Linux has become one of the greatest success stories in the history of computing. Each day the mainstream and alternative press bring more news of Linux's meteroic rise to the top.

    Unfortunately, FreeBSD essentially remains in the same rut where it was six years ago. FreeBSD has mostly stayed in its little niche, without much success to show.

    Solaris is holding its own and is widely accepted but it hasn't got the millions of users that Linux has. Still, after Linux, it is probably the strongest challenge to Microsoft.

    SCO remains a question mark. It was widely used by small business but NT and Linux probably have eaten at some of its market.

    Six years from now, Linux and whatever MS is offering will probably be the predominant operating systems in use by developers and the world at large. It will be fun to watch, whatever way it plays out.

  • You can't walk into most bookstores today without seeing Linux CDROMs and texts.

    To that end, I will draw an analogue by saying that 2 years ago you couldn't walk into a book store without seeing Java books all over the place threatening C++/Windows dominance as a programming language and OS platform. Now we all now that Java is dead and it's not because it wasn't great, it was because it couldn't penatrate the Microsoft domain well enough get accepted as a de-facto standard. Linux will be the same, more power to it, though.
  • Both KDE and GNOME are important efforts at improving all Unix desktops, not just GNU/Linux. Have you ever tried to use CDE?
  • At the time Berkeley had sold all commercial rights to BSD Unix to BSDI and thereby gotten embroiled in a lawsuit with USL (Unix System Laboratories), the semi-autonomous AT&T division (sold to Novell sometime in that time period) which owned the rights to Unix. In addition, a Republican administration in California was busily dismantling their educational system with yearly budget cuts for all public universities, to the point where UCB barely had the money to engage in normal instructional and research activities, much less stuff like PostGres or BSD. And finally Sun had dumped BSD-inspired SunOS in favor of the "merged BSD/SysV" that they had spent much money working on with USL, which was released as System V.4 for mere mortals and as Solaris for Sun customers and which looked suspiciously like System V.3 with a BSD compatibility bag slapped on the side (I know, I know, there were considerable kernel improvements between V.3 and V.4, but from an API standpoint V.4 was much closer to V.3 than to BSD). Given the politics, BSD obviously would not have been a good place for Sun to dump source code upon at that time, despite Sun's close connections to UCB (hey, Bill Joy wrote 'vi' while he was a grad student there, for cryin out loud!).

    -E
  • BeOS is nice, but it's not a server platform. It seems to me more like the Amiga done right. For the children out there who know of the Amiga only as some dusty box in their father's closet, it was the first mass-produced consumer microcomputer sold in the United States with a message-passing microkernel and multi-tasking operating system, ten years before Be... except Commodore didn't know they had a graphics workstation rather than a game machine on their hands, couldn't figure out how to market it, never got the OS completed, and went belly up eventually.

    The "In the US" part is important, because the Sinclair ?QL? was being sold in Europe with much the same feature set, except it had some kind of weird tape-ish type drives instead of "real" drives and had chicklet keys (Sir Clive did not believe in having a real keyboard!). Linus attributes much of his interest in computers to hacking on his uncle's QL while a child.

    -E
  • Thing is, while 0.01% is orders of magnitude bigger than 0.0000001%, it still doesn't count for much.

    I absolutely do not think Linux's success on the Desktop is guaranteed. In fact, I do not consider it to have started. When a customer installs 150+ seats with Linux for general office productivity purposes, that's when the battle starts.
  • It's kinda weird how they mention Solaris when I don't think it really existed, as a product, at the time. Solaris, or more precisely, Solaris 2, was a re-write of Sun's OS, with some fundamental changes. Their main OS was Sun OS 4 at the time - Sun OS 5 is Solaris.

    I'm not exactly sure how things played out after that document. I know that Novell sold the Unix stuff to SCO sometime around then - maybe SCO offered more than Sun? I do know that a few years back Sun spent something like $80M to buy permanent (or 20 years I or something) licenses for some Unix/System-V stuff. Kinda funny how things turn out.

    There have been noises from Sun (mostly late last year, early this year) about opening up the source to Solaris. (in that, everyone can get it, not just developers with money, or people in education). They said the biggest problem is licensing issues - Sun don't own everything in Solaris. Looks like things haven't changed much... BTW Java, Jini, designs for the SPARC and Java chips, are not that only thing Sun have put under their "community source license" recently. Actually, it seems for quite a few products/tools, when they announced new versions they also annouced that it would be out under their "community source license" too. If/when Solaris does go this way, it'll probably be first with Solaris 8 (which is expected early 2000), and might not be back-ported.

  • It's hard to tell just how much influence others had, but Scott McNealy and other big-wigs are quoted as contributing. The document comes across as a serious proposal and not just one-man's foresight.

    I don't really know what happened after this, but I guess Novell couldn't be pursuaded to go along - they sold their IP rights to SCO...

  • To my knowledge, Sun have not (yet) announced or commited to releasing the source to Solaris. Last I heard was just thinking, strongly, about it. Do you have any references?
  • "The following references, which were not part of the November '93 version of this paper, show support for this point of view....."

    So exactly which version are we looking at and how much has changed since the original?

    Suddenly, the paper doesn't seem quite so remarkably "visionary" any more.
  • "If it's not source, It's not software".
    I think I'll make that my .sig
    --
  • Hey, didn't they open up GLX a while back?

    Bruce

  • It's already been used on Slashdot here [slashdot.org] by Aeneas (where is he, anyway?).

    But it's still a very interesting read, especially these days. Good to see how a good idea grew, even if this isn't quite what started it all.

  • Okay guys, goofed earlier and posted a half-butchered version of the PS document. Didn't notice the damn thing until AFTER I'd gone back to read a couple more articles.

    This version has been formatted in the attempt at readability. Any errors in readability are a combination of my fault, goofs with Ghostscript's conversion, and original grammatical mistakes in the document.

    P.S.: NO. I did NOT have anything better to do tonight! I'm a geek remember?

    The Sourceware Operating System Proposal
    Revision 1.8
    Larry McVoy
    and a cast of thousands, see acknowledgments
    lm@sun.com
    +14153367627

    Sun Microsystems Computer Corporation 1

    ABSTRACT

    This document describes a proposal to provide a source form, royalty free Unix as an evolution of the COSE ffort, as a means of unifying the Unix desktop market, and as an application deployment platform, with a focus on running all applications, including those from other operating systems, such as DOS and Windows 3.1. This effort is intended to provide substance to the many Unix unification and standard agreements that exist today.Significant effort has been made to address the concerns of the major Unix vendors, the Unix customer base, the DOS customer base, the Windows 3.1 customer base, the educational and research community,and the development community.

    1. Introduction
      This document is an assessment of the condition of Unix, and a proposal to improve the condition of Unix on the desktop. To get a quick reading, the reader may scan for the highlighted bars; they are a summary of the key points of each section.
      The organization of the document is background on the state of Unix, background on the efforts to fix Unix, a digression on why bother with fixing Unix, a suggestion for how to start Unix on the path to healthiness, more details on the health plan, details on managing the resulting system, alternatives to this plan, questions and answers, and finally,acknowledgments.
      Neither Larry McVoy nor the views in this document are necessarily representing the views of Sun Microsystems or Sun'saffiliates. Sun's tolerance in this matter is gratefully acknowledged.

      ©1993 Larry McVoy 9 November 1993

    2. The Sourceware Operating System Proposal
      • Unix in trouble
        Unix needs our help because Unix is dieing. Unix is no longer even close to competitive. .Unix development in the industry is in bad shape. The major Unix vendors currently spend about $1,000,000,000 a year on Unix "development." A great deal of this expenditure is redundant over the set of vendors. Microsoft spends much less than this and produces a better product. The Unix solution is fragmented. The major Unix vendors each provide their own version of Unix, resulting in at least ten major Unix systems all competing for about three percent of the computer market. Microsoft has one implementation each of DOS, Windows, and Windows/NT.Windows/NT system runs on multiple platforms and compatibility is Their product is better in terms of ease of use, installation, and administration.Guaranteed by virtue of one source base.
      • Unix is too expensive.
        Licensing ranges from $20 to $100 per seat, with vendor mark up for their costly "value add" resulting in customer seat costs of $600 $3000. Microsoft sells Windows/NT for about $150.
      • Unix has become stagnant.
        Unix has ceased to be the platform of choice for the development and deployment of innovative technology. A great deal of the early development of Unix was done by researchers because of Unix's ready accessibility. As time has gone on, it has become more and more difficult for research organizations to acquire source. Microsoft is planning on releasing Windows/NT to universities in an attempt to leverage from the same sources of innovation.
      • Unix has become large and complex.
  • Red Hat's had this in their Knowledge Base on their web site for at least a year now.
    ---------------------
  • Didn't this all get somewhere recently when Sun announced they would be distributing Solaris under the Community License by the end of the year?
  • I was curious about this too. What was McVoy's reasoning for this site?

    Why not Berkeley who could have offered both it and the BSD source at that point?
  • Okay, that makes sense. Which is odd considering you have no knowledge of the history.
  • Yes. For the /. article refer to http://slashdot.org/articles/99/02/25/2352205.shtm l and for the ZDNet article refer to http://slashdot.org/articles/99/02/25/2352205.shtm l
  • Be is compatable with Unix/Linux... the big difference with Be is the Added a diffrent Windowing system and multimedia servers... (Be is a server based OS ontop of a micro kernel.... if you don't understand that don't bother flaiming...)
    "There is no spoon" - Neo, The Matrix
    "SPOOOOOOOOON!" - The Tick, The Tick
  • Wanna Read it in PDF form (for those Post Script Impaired) its here [naops.com]
  • Nice document. I would really like to see some archived comments from whoever read it at the time.

    Can anyone go diving into the Sun archive pile and find comments supporting or denegrating the document? Also where are the people who made the comments are today.

    It is always neat to find a prediction like this and examine it in light of what has happened.
  • Spring was a research project, never meant to go into full production. Some ideas from Spring almost certainly infected other products, and the doors API (basically a fast local RPC) came out of Spring into Solaris. There's also an implementation of doors for Linux, no word on how fast/stable/featured it actually is. I know from watching truss output that doors are used extensively in Solaris.
  • > It's too bad Sun seems to have given up on Unix in many respects.

    What on god's green earth are you talking about? Sun is the only major commercial Unix vendor left that does not have a strategic partnership with Microsoft. That alone should tell you something about their faith in their Unix-based products.
  • On page 9 of the ps version (third to last paragraph of section 5.2) is the following quote:


    Many customers (including Sun's largest customer) insist on source access. NASA coined the phrase: "if it isn't source, it isn't software." It is possible that, in the future, a system will not be considered viable unless source is available.


    How close we are! I just like the NASA phrase.

    Okay, back to work....

    ~afniv
    "Man könnte froh sein, wenn die Luft so rein wäre wie das Bier"
  • And Linux is no closer to hitting everyone's desktop than Microsoft is to releasing NT's source code.

    Are you on crack? The Linux user base is orders of magnitude bigger than it was in those days. We're a hell of a long way to World Domination(tm), but the Empire of the Penguin isn't gonna be built in a day.

  • html link (Score:4)

    by logycke ( 17575 ) on Thursday July 08, 1999 @10:04AM (#1812854)
    The html version of this paper can be viewed at Larry's site here [bitmover.com].
  • It was Yet Another Research Operating System that didn't go anywhere. Very OO design, meant to be "distributed". All system interfaces were written in a specialized interface definition language.

    Sun's pages on it seem to have disappeared, although they still turn up in search engines and links from other sites.

    BTW--- I think "royalty-free" in this context meant just from Sun's perspective, in that it was written from scratch internally. Shouldn't be taken as a suggestion that it was free software. :)
  • If McVoy was correctly predicting 1999 in 1993, does that mean that he is currently correctly predicting 2005 in 1999? Because, if so, Open Source is in trouble. Isn't he the creator of BitKeeper, the non-free source management software?
    ---
    Put Hemos through English 101!
  • I can only guess you are talking about NT. What does that have to do with my topic, which was "Larry McVoy apparently doesn't think Open Source will last."?
    ---
    Put Hemos through English 101!
  • It's too bad this paper got filed away instead of getting more publicity. The positive light it puts linux and GNU software could have helped a lot of projects. Imagine if linux had a few more developers at that early age -- maybe we'd be seeing some wicked bad technology in linux now (instead of meerly wicked good :> )

    Then again, maybe linux was not mature enough to handle an influx of developers. Would "too many cooks" have spoiled the source code soup?
    ==rob
  • "Unix has become stagnant. Unix has ceased to be the platform of choice for the development and deployment of innovative technology."

    "Unix has become large and complex. Obsession with the wrong sorts of compatibility (i.e., unused features) has lead to a bloated, hard to maintain, Unix source base."

    Face it folks, both Unix, Linix and NT are bloated. Maybe a new OS like Be would do better.
  • Perhaps. It is, however, time to bring this to the forefront. Sun, IBM, SGI, HP, etc, would greatly benefit from taking Linux, actively sponsoring ports to their respective platforms, and moving as much of their proprietary software onto the platform. I, honestly, don't even care if all of it is open sourced. Ideally, it would be, but in some cases it will just take time to convince them that it's worth it to do so.

    SGI would be my prime example here - great graphics performance, but a fairly bad track record in the last two years security-wise. Why not work towards transitioning the base operating system to something fairly auditable by outside sources (linux, *BSD, whatever.) Then you get real support systems, with huge development teams contributing and moving Unix forward in ways that hasn't happened in years.

    Utopia, I know, and there are hordes of tertiary effects that would cause many people problems, but this is really the direction the computer industry seems to need to head. Historically it's been either "one size fits all" or "look at all the variety you've got - though, none if it works with anything else"

    Argh!

  • 1993 is about right. I remember going to Soalrus 2.3 in mid-late 1994. Solaris 2.1 wasn't that popular - or good for that matter. Same for 2.2 2.3 was "usable"

  • Obviously this guy is a visionary, I wonder what his views are today? Has he published any other documents?

    If Sun and Novell actually listened, I think times would have been different. At the end he mentions "Linux is a win on the political front and a lose on the maturity front". I know linux was very immature back in 1993, and still needs to grow in some areas today. But it is improving with each release. It would have been nice to see Solaris under GNU. But now Sun missed out and we have Linux to work with. I've always said that "software is a service, not a product". This means that you pay to have something done (written). And you get to do what you want with it later. It seems that Larry McVoy could predict the future. This paper was better than ESR's Cathedral and the Bizarre.

  • He mentions a royalty free OS named Spring, wonder what ever happened to it?

    George
  • I found it quite interesting, and amusing, that he considered the Unversity of Colorado, Boulder as one of the places to donate the source to. CU's CS department only became a full department in the College of Engineering just this last year! I remember two years ago they told me that I couldn't take CS classes to fill my engineering electives because CS wasn't technically in the College of Engineering, nor was it in the College of Arts and Sciences, they were in limbo.
    And frankly the CS department at CU isn't exactly well known, decent program but not a powerhouse in CS. Probably thought that it was mostly harmless to give to CU.
  • Unfortunately I use CDE every day at work. I am not a fan of it, Enlightenment must have spoiled me. It also doesn't help that my Ultra 1 only seems to support 256 colors.
    I admit that CDE is a much better design over Windows, but there are some major improvements that could take place.
  • The point of Open Source is not necessarily an attack on Microsoft.
    I hear all the time about Microsoft's FUD attacks on Open Source. This is proven by.. a memo or two by a Microsoft Employee? What about the 4,000,000 messages I see on slashdot attacking MS in a thread that doesn't even mention them.
    Linux may be no closer to hitting everyone's desktops than NT source being released, but who gives a toss? The whole existance of corporations like MS is their reliance on people buying their product. That's why they market so strong. No sales, means no profit. It's all on their share prices, keeping their shareholders happy at the expense of customers, and that's why it is important to them that everyone owns a copy of their software.
    Open Source, however, gives you the opportunity to choose for yourself. Noone is going to go broke if everyone doesn't run Linux, and it may not be everyone's cup of tea. I run it, I like it, thats all that matters.
  • www.bitmover.com - he is doing contract work. As the other replies noted and what was explicitly stated at the end of the PS version is that he didn't develop the whole paper alone.
  • yep, he is. And Linus will look at the tool. And eventually, it will become the standard Linux kernel source control system.

    Having read linux-kernel at that time he came over as just another Unix ego. I don't trust these kind of people.
  • It's too bad Sun seems to have given up on Unix in many respects. It seems Linux or FreeBSD (or better yet, both of them working together) is the way to go.

    It would be nice if everyone working on Linux got together for a yearly conference or something, so that we could have what I percieve to be the biggest advantage micros~1 has over linux (a unified set of goals and tight cooperation between the various fragments of the community)

Slashdot Top Deals

Understanding is always the understanding of a smaller problem in relation to a bigger problem. -- P.D. Ouspensky

Close