decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books
Your contributions keep Groklaw going.
To donate to Groklaw 2.0:

Groklaw Gear

Click here to send an email to the editor of this weblog.


Contact PJ

Click here to email PJ. You won't find me on Facebook Donate Paypal


User Functions

Username:

Password:

Don't have an account yet? Sign up as a New User

No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.


What's New

STORIES
No new stories

COMMENTS last 48 hrs
No new comments


Sponsors

Hosting:
hosted by ibiblio

On servers donated to ibiblio by AMD.

Webmaster
Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have - Updated
Wednesday, May 11 2011 @ 04:17 AM EDT

Remember how SCO told the court in SCO v. IBM that Linux wasn't ready for the enterprise until IBM got involved in the year 2000 and allegedly worked to make it "hardened" for the enterprise by donating code? It said that it wasn't until 2001, with version 2.4 of Linux, that Linux was ready for enterprise use. Linux, SCO said, was just a bicycle compared to UNIX, the luxury car, until IBM did all that.

Not only is that chronology not true, I think I can show you evidence that SCO knew it was not true or could have and should have known. Just in case the case ramps up again in some form, I thought it would be good to add the evidence to our collection.

Here's what SCO wrote in its original complaint:

84. Prior to IBM’s involvement, Linux was the software equivalent of a bicycle. UNIX was the software equivalent of a luxury car. To make Linux of necessary quality for use by enterprise customers, it must be re-designed so that Linux also becomes the software equivalent of a luxury car. This re-design is not technologically feasible or even possible at the enterprise level without (1) a high degree of design coordination, (2) access to expensive and sophisticated design and testing equipment; (3) access to UNIX code, methods and concepts; (4) UNIX architectural experience; and (5) a very significant financial investment.

85. For example, Linux is currently capable of coordinating the simultaneous performance of 4 computer processors. UNIX, on the other hand, commonly links 16 processors and can successfully link up to 32 processors for simultaneous operation. This difference in memory management performance is very significant to enterprise customers who need extremely high computing capabilities for complex tasks. The ability to accomplish this task successfully has taken AT&T, Novell and SCO at least 20 years, with access to expensive equipment for design and testing, well-trained UNIX engineers and a wealth of experience in UNIX methods and concepts.

86. It is not possible for Linux to rapidly reach UNIX performance standards for complete enterprise functionality without the misappropriation of UNIX code, methods or concepts to achieve such performance, and coordination by a larger developer, such as IBM.

SCO removed the bicycle analogy in its second amended complaint, but the theme remained the same:
82. To make Linux of necessary quality for use by enterprise customers, it needed to be re-designed and upgraded to accommodate complex multi-processor functionality that had taken UNIX nearly 20 years to achieve. This rapid re-design was not feasible or even possible at the enterprise level without (a) a high degree of design coordination, (b) access to expensive and sophisticated design and testing equipment; (c) access to UNIX code and development methods; (d) UNIX architectural experience; and (e) a very significant financial investment. The contributions of IBM, which had access to UNIX System V Protected Materials and years of enterprise level experience, made possible this rapid redesign of Linux for enterprise use.
And at the SCO v. Novell trial, SCO's expert Dr. Gary Pisano sang essentially that same song:
Q. Was Linux always able to compete with UNIX?

A. No.

Q. Can you explain?

A. Yes. Initially, Linux was actually developed by a graduate student in Finland, Linus Torvalds, and initially it was really a bit of a hobbyist's toy. A few computer aficionados played with it and made contributions. Over time, however, it began to get used in businesses. Initially, just for very simple business applications. So in -- for a computer server, web serving, just serving up web pages, it is a very simple function. But over time, as it got more capability, it could -- it could take on more difficult tasks and more complex tasks for the businesses such as transaction processing.

Q. Is there something called enterprise hardening?

A. Yes.

Q. Can you explain what that is?

A. Yes. That again refers to building in capability and functionality into the operating system to make it reliable, more available, and more scalable. That is you can use it at -- for lots of -- lots of users and use many of them.

Q. Did there come a time when Linux became enterprise hardened?

A. The first point at which that really happens is with the introduction of Linux version 2.4 which is introduced to the market in January of 2001 or February of 2000 -- announced in January 2001, available in February of 2001. That really began to include some elements that made it enterprise hardened.

Q. And how did that affect Santa Cruz?

A. This had a very significant impact on them because, again, as I explained before, you have this operating system, Linux, which could do a lot of the same things that UNIX could do and it just ate right into the market share.

Darl McBride echoed the same sentiment when it was his turn on the witness stand:
Q. Can you explain how Linux became a competing operating system?

A. Yes. Linux, as we talked a little bit about yesterday, in the initial phases was somewhat of a hobbyist tool, a hobbyist operating system. It was like a garage band operating system with college students and free programmers and people getting together and coming up with this fun little operating system. Somewhere along the way, though, there was a transition where it went from being a fun little garage band toy to being a major operating system that big companies used. What happened along the way is we had a joint venture partnership with the IBM Corporation that went sour, and later we found that important materials of ours had found their way via IBM into Linux, which made it a very viable operating system for businesses to buy.

Let's see if this is true.

Here's a Joe Wilcox article from 1999, about Compaq offering Linux for the enterprise:

Compaq is refocusing its corporate strategy along two fronts, selling volume Windows NT and Linux server computers and targeting niche markets for its version of Unix.

The move means major changes for the Tru64 Unix operating system and the Alpha processor technologies inherited from Digital Equipment as well as Compaq's fledgling Linux effort. The company will focus the Tru64 OS on a handful of specialized markets while expanding Linux's presence on its Intel-based ProLiant servers, as it maintains its high-volume market strategy for Windows NT.

"What we are doing, and what we believe all will do long term, is specializing our Unix in select markets," said Steve Kirchoff, vice president of strategic marketing for Compaq's Enterprise and Services group. "We are seeing this need...so we are moving early."

So, in 1999, Compaq had decided to sell Linux. It decided to sell UNIX for niche markets, but Linux and Windows for the rest of its customers. Does that make Linux sound like a hobbyist's toy?

Santa Cruz, of course, wasn't crazy about the trend, but it viewed it as inevitable by 1999, as this article from 1999, SCO to boost revenue by offering Linux services", shows:

Santa Cruz Operation, a veteran of the Unix market, is about join the growing crowd of companies designing and building Linux systems.

SCO will announce Monday that it will help companies set up Linux computer systems, competing with IBM and a several other companies. It's an important--if not necessary--move for the long-time Unix purveyor, whose revenues are threatened by the spread of Linux.

Professional services--essentially high-priced hand-holding for customers who need someone experienced to set up or run complex computer systems--are a growing business for computer companies. Although Linux can be obtained for free, Linux companies are hoping to make money by selling services. SCO is a new arrival in this area, though, because it sells Unix but not its offspring, Linux.

SCO will help customers decide whether Linux is appropriate and which version is best for their circumstances, said David Taylor, vice president of SCO's 40-person professional services group. It's the first time SCO has offered such a service for an operating system other than its own versions of Unix.

SCO sees Linux as an inevitable feature in the business computing landscape and believes its Unix expertise will give it an edge over other consultants, Taylor said.

So, we see that there was by 1999 "a growing crowd" of companies focused on Linux, and SCO itself -- in that very year -- knew that Linux was "an inevitable feature in the business computing landsape". So, what do you think? Did SCO know that Linux was in use and appropriate for the enterprise in 1999, a year before the IBM activity that SCO objected to in its complaint and two years before Linux kernel version 2.4 was released? The title alone tells you that Santa Cruz expected to "boost revenue", meaning they knew or at least hoped that there was real demand in the marketplace already. Obviously, companies were using Linux by then, and rather than lose that business, Santa Cruz decided to bend to the inevitable, while still trying to direct them to UNIX as it could.

Dell confirmed that Linux was already in use, in fact that it was popular by then on servers, and indeed Dell was offering it in 1999, even on business desktops and workstations, not just on servers:

Dell: Linux could sell in volume

Speaking in an interview yesterday, Dell praised Linux and said the newly popular operating system matches the PC maker's high-volume sales model.

His remarks are significant in light of the close relationship Dell has historically maintained with Microsoft. Dell pre-installs Linux on workstations, servers, and business desktops--the only major PC maker to do so--but the company acknowledges that Linux sales so far are only a small fraction of its overall revenue.

Linux is popular in servers, where its Unix roots lie, but the fact that its source code is open and it has so much developer interest has helped it to spread to more ordinary desktops and even small gadgets. Several companies, including Corel and Caldera Systems, are working on making Linux easier to install and use for average computers.

Some analysts have said that adding Linux to a product line gives computer makers more leverage with Microsoft when negotiating how much they'll have to pay for Windows.

So did SCO know Linux wasn't a bicycle totally unsuitable for the enterprise and that it wasn't IBM that made it possible for Linux to be used for business? So much so that in this 1999 article, we learn that SCO developed a way so that users of UnixWare could run Linux applications on UnixWare. Please notice how this Santa Cruz executive, Greg Schwarzer, viewed Linux back then:
Santa Cruz Operation added the ability to run Linux programs to its UnixWare operating system, declaring that the upstart operating system has helped UnixWare more than it has harmed it.

While Microsoft executives say Linux competes chiefly with Unix systems such as SCO's instead of Windows NT, SCO believes the opposite.

"So far as we've seen it's actually helped us," said Greg Schwarzer, director of small and medium business marketing at SCO. "Linux has got the word out that Unix on Intel is a viable alternative to Microsoft."...

The new Business Edition of UnixWare will gradually supplant SCO's OpenServer in the small to medium business market, Schwarzer said, although SCO will continue to support the company's older Unix product by writing drivers, for example. UnixWare will henceforth get most of SCO's development efforts, and OpenServer won't get any new features.

OpenServer is a Unix product ultimately derived from a Microsoft version of Unix called Xenix. UnixWare, on the other hand, is the version of Unix SCO obtained from AT&T, via Novell, in 1995.

In other words, by 1999 OpenServer was thought of as a dying product. Linux was seen as inevitable. Santa Cruz was offering a way to run Linux applications on UnixWare, hoping to hold on to their customers. What should that tell you? It tells me that there were a lot of applications for Linux, more than UnixWare had or ones that were so popular that there was a danger of losing customers to Linux altogether. Remember, this is still 1999.

Let's look at an article from early August, 2000 to see if it was just students and hobbyist programmers involved in Linux:

Linux is receiving significant attention from programmers, students and software companies such as Oracle. Unix, meanwhile, has been relegated to a mere feature of a server bought from Sun Microsystems, IBM, Compaq Computer, Hewlett-Packard or SGI.

Selling operating system software as a standalone product that can work on a variety of computers is a difficult task, said Illuminata analyst Jonathan Eunice.

The business of "hardware-independent operating systems...seems to work only for Microsoft these days," he said.

D.H. Brown Associates analyst Tony Iams said the Caldera acquisition signals the end of an independent Unix.

"Independent Unix is effectively becoming Linux," and over time, Linux will acquire the high-end features of Unix, Iams said.

Eunice agreed that Linux has tremendous momentum. "The attention to Linux is enormous compared to the attention of any of the individual Unixes," Eunice said. "Solaris (from Sun Microsystems) maybe has more momentum than the average Unix, but Linux is clearly the place where people are paying attention."

SCO initially scoffed at Linux as an immature operating system, an adolescent seeking to play in the adult realm of Unix. But in the past two years, Linux has attracted the attention of major companies such as IBM, Dell Computer, Compaq, HP, Intel, Oracle, SAP and countless others. In the end, SCO had to recognize Linux as a serious competitor.

"It's been a gradual evolution," Giga Information Group analyst Stacey Quandt said about SCO's attitude shift. SCO simply couldn't ignore Linux, despite the operating system's relative immaturity. "Linux was eroding their market share on the low end."

Companies were involved already by 1998, large companies, like Oracle. That's three years prior to the release of the 2.4 kernel. Did you see that Oracle was mentioned as the one really interested in Linux back in 1998? Others on the list of those paying attention to Linux included IBM, HP, Intel, Compaq, etc.

When the deal between SCO and Caldera was first set up, Caldera wasn't even going to buy OpenServer, because it viewed it as a dead product. It had decided not to do any more upgrading of OpenServer. Later Caldera added it to the deal, and here is why:

In addition, the acquisition of OpenServer allows Caldera Systems to control the software's future, making it easier for the company to proceed with its plan to encourage customers to switch either to Linux or to UnixWare, Tamang said.

The new terms are among a number of changes in the Linux business landscape, including layoffs at SuSE's North American operation and Turbolinux's planned acquisition of Linuxcare.

SCO stopped development of OpenServer in 1999, and Caldera Systems plans to continue with SCO's practice of only minor updates while encouraging customers to migrate to other operating systems, Tamang said....

Caldera Systems' strategy is to sell operating-system products that span from low-end Linux machines to high-end 32-processor UnixWare servers with advanced features such as clustering. All these products would be able to run Linux software without it having to be rewritten for the non-Linux operating systems....

The original deal didn't include OpenServer because Caldera Systems wasn't interested in a product that essentially was being phased out and because it wanted to minimize how much cash it had to pay, Tamang said. In addition, SCO was interested in the ongoing revenue stream to fund its Tarentella software business.

Not only did Santa Cruz know Linux was heavily used in the enterprise prior to Linux 2.4, so did Caldera. It wanted to control OpenServer, which it also planned not to upgrade much, so that it could direct customers to Linux and UnixWare. There were a lot of companies involved in Linux, not just IBM, as you can see. Both Caldera and Santa Cruz were themselves involved in Linux. It was Caldera's exclusive business for years prior to this purchase. Yet in opening statements in the SCO v. Novell litigation, now being appealed by SCO, Boies Schiller's lawyer Stuart Singer told the jury this:
SINGER: Now what is Linux? It started as a hobbyist tool. It was invented by a man in Finland, a student named Linus Torvalds. In fact, the very name Linux is a combination of Linus and UNIX.

Now Linux is open-source software, which means that thousands of people around the world make co-contributions to it, and the problem is no one can be completely sure where that code is coming from. And Linux was distributed for free. It started as a hobbyist tool. But IBM decided that this would be a great platform around which it could build other businesses and sell products. And so they decided to enhance Linux, make in stronger for a business tool by putting in UNIX technology. The only problem is that UNIX technology belonged to SCO. It was protected by contracts.

And IBM began doing this in early 2000 and 2001. It had a tremendous effect on SCO's business. In 2003, SCO decided they had to do something about it. They decided to bring a lawsuit against IBM, and to ask users of Linux to buy a license, to say, all right, if you want to use Linux, pay us a license so that we're being compensated for the use of what is our intellectual property.

Does this match any of the older articles I just showed you? Incidentally, Dr. Pisano, SCO's expert, knew that Linux was used in the enterprise in 2000, at least, because in his damages calculations in his expert report, he included information from that year and from 1999, as you can see for yourself if you look at the footnotes. The money was calculated from later, because SCO targeted Linux kernel 2.4 starting in 2001, but look at the footnotes, and you'll see data from 1999 onward. I think Dr. Pisano overlooked something else. It was Caldera itself that contributed at least one of the enterprise-hardening features to Linux 2.4 -- Christoph Hellwig, as Groklaw demonstrated way back in 2003. So there's really no excuse.

The jury decided that it wasn't SCO's "intellectual property" anyhow, so in a way it's moot, but just in case the appeals court does the unthinkable and sends this mess back for another trial, it's important to set forth all the details that prove that SCO was blowing smoke. At best.

By the way, it isn't true what Singer said to the jury about not being able to know who writes Linux. You can know who contributes to Linux. There is in every Linux kernel a file, a credits file, that lists everyone who contributed to that version, so that was a boldfaced assertion, considering that you can find all that on the Internet too, since Linux is developed in public. Here's the credits list for Linux 2.4.22, for example. Please notice that information is provided on how to contact them, and precisely what they contributed? Like I say, boldfaced. But then SCO was nothing if not boldfaced.

I just wanted to get this in our collection before I go, because no one has followed the SCO litigations as closely as I have for eight years, and before I go, just in case the SCO dream didn't quite die and someone tries to resurrect the litigation follies, I want all the evidence I've collected to be here.

Update, 2013: They knew, for sure. And here is the evidence, a 1998 SCO press release about Santa Cruz teaming up with Intel to port UDI to Linux:

"Standardization in this industry is what drives up the performance and innovation curves," said Ray Anderson, SCO's senior vice president, Marketing. "Intel's support of UDI as a standard means that all UNIX OS vendors can use a common device driver on all Intel platforms. SCO and Intel will strongly support the movement to standardize the use of UDI for all UNIX platforms on Intel, which we believe will generate even more momentum for the already exploding UNIX on Intel market."...

UDI for Open Source Community

SCO strongly supports the growth of standards-based computing and encouragement of open systems development. SCO, with Intel and Project UDI, will support the open source community by working to ensure that UDI works on the Linux operating system.

Anderson continued, "The Linux and open source movements are powerful forces in the industry that are creating a huge resurgence in the interest in the UNIX System. It helps to bring the community together again and with UDI available on the Linux system, their developers can use the latest UNIX devices and peripherals on the market."

I know. Hilarious. And here's a press release from 2000, just because I found it at the same time, and it's also funny. It's a press release about Santa Cruz and Caldera joining together to merge Unix and Linux. And Caldera refers to itself as a "Linux for Business" leader:
Caldera Systems, Inc., (Nasdaq: CALD), a "Linux for Business" leader and the Server Software Division of SCO (NASDAQ:SCOC) today announced a joint initiative to combine and offer Linux and UNIX server solutions and services worldwide.

The new offerings, announced today at Forum2000, include the Linux Kernel Personality for UnixWare 7 - the first technology developed to bring UnixWare 7 into the Linux business market. Also announced today is Enterprise Support Services, SCO's worldwide support program designed to meet the business and time-critical requirements of enterprise customers.

These offerings fall under the umbrella of the Open Internet Platform (OIP) which combines Linux and UNIX server solutions and services. The OIP provides commercial customers and developers with a single platform that can scale from the thinnest of clients to the clustering needs of the largest data center. The Open Internet Platform combines the robust scalability of the UNIX system with the low-cost, developer-accepted Linux operating system.

"With Caldera's pending acquisition of the SCO Server Software and Professional Services divisions, we already see some progress in the integration of our technologies," said Ransom Love, CEO, Caldera Systems, Inc. "We believe that the combined strengths of Caldera and SCO are key in offering powerful Linux and UNIX solutions to the enterprise."

So they knew Linux was no bicycle. The point of all the work was to benefit Unix, as this 2000 press release shows:
The SCO Server Software Division, an independent business unit of the Santa Cruz Operation, Inc. (Nasdaq:SCOC), announced that it has developed the Linux Kernel Personality (LKP) for UnixWare 7. The new LKP technology is a key step toward the vision of an Open Internet Platform (OIP), discussed in the recent announcement that Caldera plans to acquire the SCO Server Software and Professional Services Divisions.

The new LKP technology unifies UnixWare 7 with Linux at the kernel level and also includes a complete Linux environment. This adds full Linux capabilities to UnixWare 7 while keeping all the major UNIX attributes such as scalability, reliability and clustering intact. Linux applications can install and run on UnixWare 7 without modification, resulting in the addition of a host of new solutions to the wealth of applications already available on UnixWare 7.

"The Linux Personality is great news for both the UnixWare 7 and enterprise Linux markets," said John Palmer, vice president of Marketing, Server Software Division, SCO. "LKP will make the entire Linux application portfolio available for UnixWare 7 customers, including enterprise server applications. At the same time LKP allows Linux customers to take advantage of a highly scalable and cluster-ready operating system based on the proven UnixWare 7 platform. Because the LKP runs at the kernel level, customers will not know it's there - it will just work."

UnixWare 7 customers can now take advantage of an even wider selection of applications in a scalable, cluster-ready computing environment. Linux channel partners and ISVs will benefit from LKP as it will increase the number of operating platforms they can offer to their customers.

"LKP technology is an integral part of Caldera's vision of the Open Internet Platform," said Ransom Love, CEO of Caldera Systems, Inc. "Caldera's pending acquisition of the SCO Server Software Division speeds the movement toward Linux/UNIX integration and the creation of a single platform for Internet computing. SCO's demonstration of this key technology to their partners at Forum2000 will be particularly beneficial to the industry."

Why, a logical person might ask, would UnixWare 7 need Linux apps, instead of the other way around? Because the developers were flocking to Linux. Remember Steve Ballmer's speech about "developers, developers, developers" ? Linux was hot and UnixWare was not.

  


Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have - Updated | 160 comments | Create New Account
Comments belong to whoever posts them. Please notify us of inappropriate comments.
Corrections here please
Authored by: NigelWhitley on Wednesday, May 11 2011 @ 04:23 AM EDT
Please place any corrections to the article here, preferable with the title in
the form
erorr-->error

Thanks
----------------
Nigel Whitley

[ Reply to This | # ]

Newspicks here
Authored by: NigelWhitley on Wednesday, May 11 2011 @ 04:25 AM EDT
Please place comments on and suggestiions for the Newspicks feature under here.
Kindly include a link to the article so it persists long after the article has
scrolled out of the Newspicks pane.
---------------------
Nigel Whitley

[ Reply to This | # ]

Off topic [OT]
Authored by: NigelWhitley on Wednesday, May 11 2011 @ 04:28 AM EDT
Kindly include off-topic comments under here ie those subjects not directly
related to the article
--------------
Nigel Whitley

[ Reply to This | # ]

[COMES] Comes v. MS docs
Authored by: Aladdin Sane on Wednesday, May 11 2011 @ 04:31 AM EDT
For docs in Comes v. MS.

---
Well, it's either Captain Kirk or Doctor Strangelove. —me, ca. 1984-1985

[ Reply to This | # ]

Linux Heavily Used by 1999
Authored by: maroberts on Wednesday, May 11 2011 @ 04:55 AM EDT
I'm never sure what the threshold is to be considered "Enterprise"
level, but I was happily using Redhat 4 on a number of servers in 1997 for a
large wireless communications company, for which Nortel was developing the
equipment in conjunction with the two companies I was working with. The servers
were used to perform testing of the network and fault analysis, and whilst HP-UX
and Solaris systems were also used, they were in the main fairly
interchangeable.

I also worked for Alcatel circa 1998 and it was heavily used for network
management systems. Linux servers were used for testing and the system finally
ran on Solaris IIRC, but there was no real reason why the systems could not have
remained on Linux. In 2000, just after the date, I was again using Linux servers
for nationwide cellphone network management.

[ Reply to This | # ]

Red Hat and Linux 1.2.13
Authored by: Anonymous on Wednesday, May 11 2011 @ 04:56 AM EDT
There was a huge amount of Linux hype in the media around the time Red Hat went
public in August 1999. No one in the industry could have failed to notice.

When I started work in 1997 I had version 1.2.13 of the Linux kernel on my
desktop machine, and so did most of the engineers at my place of work. That
kernel version was released in 1995, apparently.

[ Reply to This | # ]

Dell was offering it in 1999, even on business desktops and workstations, not just on servers
Authored by: Aladdin Sane on Wednesday, May 11 2011 @ 05:12 AM EDT
It was this announcement of the Linux addition by Dell in 1999 that made me decide I was allowed to work there. Prior to that the company was way too Microsoft for me to set foot at without barfing. (They had campuses all over Austin, I couldn't look at them without getting ill.)

By March 2000 I was a tech there, by 2003 a Linux Server tech, and by 2005 a Linux Test Engineer at Dell.

At the time of the start of the change, 1999, it was 100% Red Hat (RHEL and its predecessor) at Dell, later SUSE (SLES) was added to servers, and Ubuntu to desktops and portables.

---
Well, it's either Captain Kirk or Doctor Strangelove. —me, ca. 1984-1985

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: jjs on Wednesday, May 11 2011 @ 05:29 AM EDT
I worked for a company doing enterprise support for Linux back in 2000. One of
our main competitors was a company called Red Hat. At several technology shows
I attended, many SMB and large companies were looking at linux support - so
clearly many felt it was good enough. BTW - IBM was a big supporter of Linux as
well. I don't remember SCO (the original SCO), but that may be my memory.

---
(Note IANAL, I don't play one on TV, etc, consult a practicing attorney, etc,
etc)

[ Reply to This | # ]

Specious At Best
Authored by: sproggit on Wednesday, May 11 2011 @ 05:41 AM EDT
The SCO claims and assertions excerpted in this article are misleading at the
very least, for a variety of reasons.

One not yet covered here relates to the evolution of hardware. We see Darl
McBride waxing on about how Unix supported 16 or 32 processors at this time. I
won't dispute that here, but I would be interested to know (if anyone can help
answer) whether that performance claim was for true 'vanilla' SVRX Unix, or if
it refers to one of the derivatives being developed by licensees such as Sun,
IBM, HP, or even companies like Cray (a lot of Cray work went into Sun's E10k,
E12k and E15k machines).

Instead, I'd like to focus our thoughts on the hardware of the day. In the time
period that Darl and SCO mention, exactly how many vendors produced
significantly parallel CPU hardware architecture? It certainly wasn't Intel -
that's for sure. Oh there were systems, for sure... and the moment that Intel
announced SMP capabilities companies like Supermicro and Tyan and Compaq started
to produce boards or systems based on this technology.

But I think that if we look at this timeline with a critical eye we might notice
that this evolution of software closely parallels hardware capability. In other
words - and this is critical - I think that the growth in the capabilities of
Linux were more closely linked to hardware evolution than we might realise, as
opposed to purely taking advantage of alleged contributions from IBM.

I think we can go even further. There are numerous examples - a good one being
USB (Universal Serial Bus) technology where GNU/Linux actually had good,
effective and reliable support *before* other mainstream OS like Windows, and
was a country mile ahead of Unix. If you look at the history of Unix development
over the last few years, one of the things that they are slowest at [ and seem
to spend the most time on ] is writing device drivers for new hardware.

So for Darl to make claims that the Linux kernel scaled from 'small' and
'hobbyist' systems to being enterprise ready with the help of IBM is possibly as
close to an outright fabrication as it is possible to get...

Having said that - a footnote. Way back I read a fascinating article about a
Technology Manager and Linux hobbyist at a major IBM mainframe site who produced
the world's first port of Linux to the zSeries platform. This was achieved in
the traditional way - taking source code (from a SuSE distribution, IIRC, and
uploading file-by-file and then compiling. The finished Linux
"instance" ran as a single thread on the IBM hardware, under zOS,
which meant that the instance could be cloned, stopped, started, etc, all from
the mainframe OS itself. The article stuck in my memory for the fact that in the
weekend window that the manager had available for testing, he managed to clone
something absolutely remarkable like 40,000 [yes, you read that correctly]
discrete instances of Linux on one single IBM mainframe.

That was all done by one of IBM's enterprise clients, although Big Blue soon got
involved and interested.

I mention this because I don't ever recall reading SCO or anyone reporting that
they could get SCO's flavour of Unix to scale like that...

[ Reply to This | # ]

The difference betwen Geeks and Management
Authored by: ThrPilgrim on Wednesday, May 11 2011 @ 09:10 AM EDT
The first point at which that really happens is with the introduction of Linux version 2.4 which is introduced to the market in January of 2001 or February of 200

<cynicism>

I don't think that Linux has ever been introduced to the market. No Geek of my knowledge would ever use such a phrase. It may well have been released around then but only management thinks everything is a market place where everyone can be bought and sold.

</cynicism>

---
Beware of him who would deny you access to information for in his heart he considers himself your master.

[ Reply to This | # ]

Oracle shipped in 1999
Authored by: Anonymous on Wednesday, May 11 2011 @ 10:25 AM EDT
According to
http://rcpmag.com/articles/2000/08/09/oracle-uses-linux-to-compete-with-ibm.aspx
Oracle was shipping its flagship product, the Oracle 8i database, as early as
July 1999. In other words, Oracle considered Linux to be "ready for
Enterprise" in 1999. Earlier, because the decision was taken much
earlier.

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: Anonymous on Wednesday, May 11 2011 @ 11:06 AM EDT
I'd like to point out, too, that after the Novell Trial, most of Caldera's
bogo-claims are dead, dead, dead, anyway.

As ESR pointed out waaaaaay back in the beginning, SCO's own Open Server
supported NONE of the services that it claimed IBM had somehow misappropriated
from Linux. Linux SMP processing was, and remains, far more advanced than
anything SCO Open Server offered; Open Server has no Journalling File System,
(neither IBM's JFS nor any other); and, Open Server has no idea what RCU or NUMA
are. If System V belongs to Novell then the fact that Linux has capabilities
that Open Server never had, (and is never going to have), is neither IBM's nor
Linus Torvalds's fault.

(I won't mention, again, as ESR did, that Caldera, themselves helped to improve
Linux's SMP capabilities by donating what was at the time expensive
multiprocessor hardware to Alan Cox to use to develop and test on.)

( http://www.catb.org/~esr/hackerlore/sco-vs-ibm.html )

The Flatlander

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: Anonymous on Wednesday, May 11 2011 @ 02:52 PM EDT
Hm, the Compaq equivalent is a bad idea. I ran production on those machines
(HPC) so I know. The Linux strategy was corporate hype after a completely
misguided takeover of a company (DEC) that Compaq did not have any use for.

Tru64 was atrociously outdated at the time as far as management tools, available
software, etc went. Very hard to administer compared to Linux at the time.

Its strength though was that if you wanted to use those machines for what they
were purchased for - high performance computing - the DEC Fortran Compiler, the
extremely stable kernel, excellent error logging facilities in the kernel when
some hardware broke, made those machines unbeatable.

Linux on alphas was absolutely no contest, much less stable, much slower
compilers (you had to import static binaries built on Tru64 by DECFortran by
hand to run at anywhere near adequate speed on Linux)

I am sorry, as I am certainly not a supporter of SCO's arguments about 1999 (as
a broad platform, Linux was much better). However, in your criticism of Tru64,
you are unwittingly endorsing the ramblings of a completely ill-informed
management that had no idea of how people used their machines in practice. That
was their worst failure, they did not know what their own machines were even
good at ... and hence they sank the whole ship (alpha), as history shows pretty
compellingly. Don't go down the route of endorsing a false argument simply
because it's a cheap opportunity for a jab at SCO's also false (but for
different reasons) propaganda.

... and yes, it would be interesting to hear what jon maddog hall would have to
say on the issue. Linux on Alpha certainly showed great promise on those
systems, again, much simpler to administer - but they never got to where they
needed to go with it (specialized perfection) because the whole platform died
first.

[ Reply to This | # ]

Hobbyist's toy?
Authored by: Anonymous on Wednesday, May 11 2011 @ 04:26 PM EDT
So, in 1999, Compaq had decided to sell Linux. It decided to sell UNIX for niche markets, but Linux and Windows for the rest of its customers. Does that make Linux sound like a hobbyist's toy?
If Linux is a hobbyist's toy, then Windows must also be a hobbyist's toy...

[ Reply to This | # ]

Linux SMP only 4-cpu? Not at all.
Authored by: Anonymous on Wednesday, May 11 2011 @ 04:45 PM EDT
I generalized the Linux x86 SMP startup/management sequence in the 1996/97 time
frame.

The result was capable of 15- or 255-cpu SMP, depending on which kind of x86
hardware was used. I regularly tested it on a 4-cpu Pentium Pro test machine
Intel let me use from when I worked there, and sent patches to Alan Cox/others,
who would then massage it into what they liked for kernel submission.

Heck, I still have that old 4-cpu Pentium Pro test machine (long obsolete, of
course), and gave credit to Intel in comments in the Linux source code for
letting me test it on there in my spare time.

Erich Boleyn

[ Reply to This | # ]

So, who did those Unix folks misappropriate their code from?
Authored by: Anonymous on Wednesday, May 11 2011 @ 05:15 PM EDT

I really enjoyed SCO's original complaint:

"86. It is not possible for Linux to rapidly reach UNIX performance standards ... without the misappropriation of UNIX code, methods or concepts..."

If it's impossible to design code to reach arbitrary performance standard X, that such code must have been stolen, then it begs the question:

Who did SCO steal Unix code from?

Oh wait, that's right. Everybody.

[ Reply to This | # ]

Enterprise Linux?
Authored by: sk43 on Wednesday, May 11 2011 @ 05:51 PM EDT
Let's be honest - the term "enterprise" as applied to software is purely a marketing term. The only way to counter SCO's allegations about Linux in the enterprise is with ... marketing statements.

Caldera eServer 2.3 was based on the 2.2 kernel and thus not "enterprise-ready" according to Gary Pisano. Right?

April 17, 2000 Press Release:

"Caldera Ships Powerful Linux Internet eBusiness Solution; OpenLinux eBuilder -- ENTERPRISE and ASP Editions."

"Included in OpenLinux eBuilder is Evergreen's ECential 3.0, which delivers robust, open standards-based (Java, XML) distributed eCommerce services and components that enable high performance solutions for ENTERPRISE or Application Service Provider (ASP) eBusiness sites."

Oct. 25, 2000 Press Release:

"Caldera eServer 2.3 Wins Network World Blue Ribbon Award; OpenLinux Tops Other Linux Distributions for ENTERPRISE Server Use."

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: dmarker on Wednesday, May 11 2011 @ 08:48 PM EDT

Well summarized and also supports the notion that Darl McBride was put into his
role at Caldera to play up the SCO Unix acquisition & distance Caldera from
its Linux heritage.

If we consider the fact that Caldera was chosen by Ray Noorda (as founder of
Novell & the head of the Canopy Group) to handle the DR-DOS suit against
Microsoft (at which Caldera succeeded), it is reasonable to assume that Ralph
Yarro wanted to emulate his Boss Ray Noorda, when he (Ralph) was attempting to
seize control of the Canopy Group.

Ralph lost his battle to control the Canopy Group but extracted control of
Caldera as part of his walk-away package. Yarro has always appeared to be the
force behind the sue IBM & Linux scam. McBride was his loyal lieutenant.

Microsoft paid a large settlement to Caldera over the DR-DOS suit & had no
love for Novell, Noorda or Caldera. But, did see a golden opportunity to help
Caldera in its litigation against both IBM and Linux.

Remember that Microsoft at this time was in trouble with NT (Longhorn) and
Windows & desperate to slow the competition while they cleaned up Windows
(later as Vista). Also the failure of the Itanium 64-bit chip (Intel & HP),
was a dent in Microsoft's ambitions to push the other players aside in the
enterprise.

Caldera's acquisition of some of Santa Cruz Op's assets (remember Santa Cruz Op
still existed but renamed to Tarantella) was at best part of a plan to shakedown
the Linux market and target IBM as the biggest backer of Linux.

The tactic of changing company names to confuse (if not to amuse) seems to be a
practiced art form by Caldera and its various subsequent incarnations. It seems
highly probable that the purchase of the selected UNIX assets from Santa Cruz Op
(Tarantella) included a requirement for Santa Cruz Op to change its name to
something else as Caldera needed the SCO name as part of its shakedown scam.
McBride & BE&F played that name change like the best virtuoso might play
a good fiddle :)

DSM







[ Reply to This | # ]

CREDITS file is incomplete at best
Authored by: mschmitz on Wednesday, May 11 2011 @ 09:19 PM EDT
PJ,

your reference to the Linux kernel's CREDITS file does not help your argument
that every contribution to the Linux kernel can be traced to a specific
individual. Let me elaborate.

The CREDITS file is horribly incomplete and out of date. No one is automagically
added to this file, nor are additional contributions added later. I have checked
for a few core contributors to one of the earliest ports and they don't appear
at all. I'm listed for work that was really quite peripheral only, and the
majority of my contributions are not listed. Very few contributors care enough
about claiming credit in this way.

In the current age of GIT, the commit messages contain the sign-off of any patch
author. BitKeeper might have had something similar, I've not used it for long
enough to remember the particulars.

For the time frame before Linus started to use BitKeeper and wrote GIT, the
mailing lists for each port or subsystem would be the only source of information
on who submitted what patch. Not all patches were sent to a mailing list, some
just went to the relevant maintainer by private mail. The maintainer would then
send them to the subsystem/port list for discussion, and to linux-kernel for
inclusion into mainline, with the proper attribution made.

In a nutshell - SCOs claim that you cannot know who contributed to the kernel is
still bogus, but it takes a lot more than a look at CREDITS. For someone not
involved with the community anything pre-GIT might be difficult to figure out
but with a bit of effort and goodwill I'd say it's doable. The SCO attitude 'all
your code is belong to us' might hurt in the goodwill department, though :-)
Still, there's copyright notices in the files, and author initials in quite a
few comments in the code. Still doable.

-- Michael

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: Anonymous on Wednesday, May 11 2011 @ 11:04 PM EDT
Thanks! and thanks also for your years of reporting on the trial(s)
coverage.

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: Anonymous on Thursday, May 12 2011 @ 12:09 AM EDT
Do I remember correctly? Pretty sure I do. Perhaps someone else can come up with
"chapter and verse" and provide cogent, direct quotes.

What I think I remember is that Microsoft in its anti-trust defense in DOJ vs.
Microsoft brought up Linux as a competitor. This would be something which SCO
knew or could/should have known.

If Microsoft says in court that Linux is a competitive threat and later on SCO
says in court that Linux is a "hobbyist" operating system put together
by amateurs and is not "enterprise" ready unless one takes into
account an allegedly massive and illicit recent dump of code and ideas by IBM,
it seems to me that the two statements cannot jibe.

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: iraskygazer on Thursday, May 12 2011 @ 12:53 PM EDT
PJ,

I'd like to predict that the underhand dealing that you've exposed will
return with a vengance once Groklaw is no longer active.
Do you know of another forum we, the somewhat active Groklaw members, should
join to continue the observance of the less than ethical activities of the large
corporate software companies?
Yes, I know, we can't replace the energy of Groklaw. But it would be nice to
continue the conversation that helps keep the unethical practices in check. And
at the same time, continue to protect Linux so it remains available to the
world.

[ Reply to This | # ]

HP Enterprise Linux support in 1998
Authored by: artp on Thursday, May 12 2011 @ 11:28 PM EDT
I am defining enterprise as being able to buy product under a corporate purchase
order, as well as software support with guaranteed 1 hour, 4 hour or 24 hour
support response time.

HP was offering their engineering workstations with Linux and support back in
1998. I don't remember if they sold servers with Linux then or shortly after. At
least some of the workstations were multi-processor. These were the XW and XU
series and others. They were used for software development, engineering
analysis, graphics intensive work and CAD.

A few months later, they started offering Linux as an option on all of their
business desktops.

Of course, this was all before the catastrophic merger with Compaq/Digital. And
the Linux offerings didn't last, despite the fact that I bought a business
desktop with Linux. It was the last new computer I bought, because I require
Linux on new machines. So I use old ones. The hardware has far outstripped the
software.

---
Userfriendly on WGA server outage:
When you're chained to an oar you don't think you should go down when the galley
sinks ?

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: coffeelover on Friday, May 13 2011 @ 12:17 AM EDT
I also downloaded and installed my first Linux with Slackware and it's Linux
kernel of 1.0.9 in 1994. I recall moving up to Slackware 2.2 (I think) and the
improved 1.2.xx kernel which eventualy was upgraded to 1.2.13. This was used as
the first web server for the local community college here in Bay county
(NorthWest Florida).
Note: I actually tried using a windows port of the NCSA web server on windows
3.1 ... Talk about unstable!!! The linux/ncsa web server proved uncrashable in
heavy testing.
The campus's first firewall was built using the 1.3.35 kernel on slackware
2.2. Natting was brand new back then and there were no cable/DSL routers yet!
:-)

-Patrick-

[ Reply to This | # ]

Linux Heavily Used in the Enterprise by 1999 - And SCO Knew It or Could/Should Have
Authored by: Anonymous on Friday, May 13 2011 @ 08:45 AM EDT
I was a developer at SCO Canada in 1997, many of us used SCO
when working from home, since the gcc we used was practically
the same. Compile and tests could be done on Linux and then
recompiled at work.

I was certainly not the only one working like that.

[ Reply to This | # ]

You almost proved tSCOg right!
Authored by: Anonymous on Friday, May 13 2011 @ 12:31 PM EDT
Your quotes from SCOfolk and the stuff you and commenters wrote about Linux
enterprise readiness almost proves tSCOg's claims about the 1999 state of UNIX
versus Linux right. Which we don't want to do.

tSCOg apparently defines enterprise class performance as suitability for those
tasks where a Fortune 500 business is willing to spend a million $ on high end
server hardware and wants an OS to match. All the examples of 1999 and earlier
enterprise use of Linux are for secondary systems where management would prefer
not to spend that much, such as sysadmin desktops and test systems. Especially
the latter category implies a need to run the same software (such as Oracle
DBMS) as on the expensive production server.

And all their bragging about UNIX running on 32 CPU systems must be about the
SVRX code licensed to Sun et al, not the SCO x86 products.

Not that I like SCO, but we need better examples to kill those claims.

[ Reply to This | # ]

Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )