APLawrence.com -  Resources for Unix and Linux Systems, Bloggers and the self-employed

Init scripts, ELF part of SCO lawsuit

Apparently SCO has added Sys V style init scripts and the ELF executable format to its list of misused code in the IBM lawsuit. If that were to hold up, it's pretty nasty. Even the most non-techy judge or jury would agree that Linux certainly does use ELF and Sys V init.

But it always did.

This can't be part of the IBM suit, because these were part of Linux from day one, long before IBM realized that the little penguin was useful against Microsoft. In How did we get here, I asked why SCO hadn't jumped on Linux earlier. That was back when we still thought the case was about esoteric memory managment schemes and SMP code. If ELF and Sys V init are what SCO claims they own, they should have had Linux in court a long time ago.

But they didn't.

I am not a lawyer. Every time I think something is plain and obvious, my lawyer tells me there's more to it, and vice versa. So I am definitely talking through my hat here when I say that it seems to me that it is a little late to be yelling theft now. If Linux "stole" Sys V init and ELF, SCO certainly knew that way back when. This stuff isn't buried in a gazillion lines of code, it doesn't take MIT mathematicians to ferret this out: it's right there, in plain sight, and anyone can see it. Again, SCO certainly cannot blame IBM for passing this to Linux, because this predates IBM's involvement entirely!

Sometimes I feel like I must be living in some alternate universe where nothing makes any sense. I don't know whether it's possible to "own" Sys V style init or ELF. I don't know enough about their respective heritage to make any guess as to whether SCO can make this stick. I do have the opinion that it will be a sad day for all of us if they do, because stifling innovation is never good, and killing or crippling Linux would do tremendous damage to Unixish OSes. If SCO gets this plum, it will be Microsoft who ends up eating it for dessert.

I searched around a bit trying to find if Groklaw, Slashdot or anyone else had more to say about this part of the case, but couldn't find much in the general noise. If you do know of any links that reference this, please do add 'em here.



Got something to add? Send me email.





(OLDER)    <- More Stuff -> (NEWER)    (NEWEST)   

Printer Friendly Version

-> -> Init scripts, ELF part of SCO lawsuit


10 comments



Increase ad revenue 50-250% with Ezoic


More Articles by

Find me on Google+

© Tony Lawrence




---July 18, 2004

I don't know what to think about that... Not sure exactly what "ELF" is, but init scripts I know. It's easy enough to replace them, several Linux distros don't touch them, and the ones that still use them probably don't resemble much that is used by anybody. Gentoo has it's own design. Slackware uses BSD-style init scripts. I am sure that Debian's is pretty different, but I don't know how that compares vs RPM-using distros like Fedora or Suse vs real System V init scripts.

That's a whole can of worms nobody wants to open up, imagine every innovation in design created by anywere for anything that gets incorporated into another product is now liable for lawsuits. Think Mac OS's user interface vs Window's Explorer. It would mean the end to all software innovation as we know it. Nobody could ever use any design from any other peice of software without legal precedings...


--Drag

---July 18, 2004


ELF is Executable Linking Format - pretty important. And init is more than the scripts - it's inittab and init itself, which isn't all that easy to just rip out - it of course can be done, but a lot of other stuff depends on it.


--TonyLawrence





---July 19, 2004


And ELF that's a AT&T Unix innovation, that didn't exist in any other OS? And Init, didn't BSD use stuff like that, was it present in Multics? What a bunch of crap. I doubt that SCO bought every Unix innovation ever put out by mankind. Grrr...

--Drag

---July 19, 2004

Well, they think they bought AT&T Unix from Novell. That's what that part of the suit is all about - Novell says they only bought marketing rights. They aren't claiming they developed all of it, just that they bought it.

BSD didn't use SysV sytle init, and doesn't still. Either did Microsoft/SCO Xenix. But SCO OSR5 and Unixware do. So does RedHat, Caldera - I dunno if I've seen a Linux distro that doesn't - you say some don't?


--TonyLawrence




---July 19, 2004

No that was my mistake. I was thinking of the style of scripts... You know with the /etc/init.d/scriptname, and then the various rc directories with symbolic links to the scripts in init.d that get executed.

Slackware uses bash scripts located in /etc/rc.d/ that individually get executed per runlevel.
see http://www.slackware.com/config/init.php for details.

To me RPM-based (Fedora, Caldera, Suse..) uses "system V" init scripts, and Slackware uses BSD-style init scripts.

But all of them still have init, /etc/inittab and all that. I thought it was a standard *nix way of doing stuff.

I am going to have to install FreeBSD now, to see the different way of doing things.

--Drag

---July 21, 2004

> This can't be part of the IBM suit,
> because these were part of Linux from day one, long before IBM
> realized that the little penguin was useful against Microsoft


It *IS* part of the IBM lawsuit

IBM filed a partial summary judgement motion for IBM's 10th counterclaim (that IBM's Linux activities don't infringe any SCO copyright). IBM's motion is based on the fact that SCO had twice certified (with affidavits and in response to specific orders from the court) that they had full responded to IBM interrogatories asking for *all* SCO's alleged rights in Linux, and nothing in their responses indicated a copyright infringements.

SCO now says, ELF in Linux, and IBM's distribution of Linux versions including ELF, is a copyright violation. (SCO also ask to look at everything IBM ever did with AIX and Dynix, since 1984, because some how they think this is going to produce evidence of copyright violations in Linux.

Regarding ELF, the obvious questions are:

(1) Why wasn't it raised before in SCO's two certified responses

(2) ELF is a public standard created by the Tool Interface Standards Committee c.1995. Members of this group include Novell, IBM, and Santa Cruz Operation Inc. (not the current SCO Group, Inc.). See for example: http://www.x86.org/ftp/manuals/tools/elf.pdf . So how can SCO Group have any legitimate claim to it, when their alleged predecessors (Novell and Santa Cruz Operation, Inc.) voluntarily released the standard as a royalty-free public standard many years ago (see for example page 2 of the PDF, which is copyrighted May 1995).

An interested observer


---July 21, 2004

I know it *IS* part of the suit; it just seems outrageous that it is. And yes, why didn't they bring this up earlier in the suit and as I ask, why not years ago?

And whether or not ELF were public, how can IBM be to blame for ELF or Sys V init being in Linux when both of these predate IBM's involvement?

I feel like I am deep in the rabbit hole..

--TonyLawrence



---July 21, 2004


>I am going to have to install FreeBSD now, to see the different way of doing things.

What you'll find is a process named init, with a very different man page, and no /etc/inittab.

BSD init takes arguments that look like run levels, but they actually represent "security modes". The /etc/ttys file is somewhat like inittab, but a bit less powerful.

Having been weaned on SunOS, I'm nostalgically fond of BSDism's. When I first bumped into inittab, I didn't like it :-)

--TonyLawrence






---July 21, 2004

> And whether or not ELF were public, how can IBM be to blame for ELF or Sys
V init being in Linux when both of these predate IBM's involvement?

That's not SCO's theory AFAIK

SCO's theory appears to be, not that IBM is to blame for putting these in Linux

But rather, SCO's theory appears to be, that IBM is to blame for using and/or distributing Linux versions that include these features (allegedly in violation of SCO's copyrights)

It seems to be essentially the same theory that SCO is using against Linux end-users. i.e. it's not their fault the stuff (whatever it is) is in Linux, but they are still liable (unless they pay SCO a license fee), for using and copying Linux versions that include the stuff.

My guess (not too much of a guess, SCO have said it in writing) is that SCO have raised this because they hope to avoid a summary judgement that IBM has not infringed any of SCO's copyrights, by IBM's Linux activities

(Whether SCO actually has any legitimate claim that ELF is in Linux in violation of one of SCO's copyrights, is another matter, one for another day, but one that I am personally skeptical of)

- An interested observer

---July 21, 2004

Thanks, that makes a *little* more sense. It's still BS though: if they think they own ELF and Sys V init, they should have been after Linux in the 90's.

Personally, I need a stronger word than "skeptical" :-)

--TonyLawrence







Mon Mar 30 14:12:53 2015: 12635   TonyLawrence

gravatar


And now they are after IBM again: (link)

I assume it's for the same things, but who knows?

As init has been replaced in many Linux distros, that part at least isn't as annoying as it used to be. ELF would still be hard to replace quickly.

------------------------
Kerio Samepage


Have you tried Searching this site?

Unix/Linux/Mac OS X support by phone, email or on-site: Support Rates

This is a Unix/Linux resource website. It contains technical articles about Unix, Linux and general computing related subjects, opinion, news, help files, how-to's, tutorials and more.

Contact us





When someone says: "I want a programming language in which I need only say what I wish done", give him a lollipop. (Alan J. Perlis)

When Life Gives You Questions, Google has Answers (AJ Carpio)







This post tagged: