Property, intangible

I have not been shy in expressing my opinion about the copyright registration system. I’m not a fan of the requirement that one must have a registration before a lawsuit can be filed; as a result, the first stages of a copyright infringement lawsuit are collateral attacks on the registration.

Cisco alleged infringement of the “Cisco CLI,” “CLI” being a commonly used initialism for “command line interface,” a non-graphical software interface. Cisco provided copies of 26 copyright registrations and applications for the software.

Arista Networks was started by former Cisco executives. Cisco alleged that Arista copied some of its software, in its brief citing public statements by Arista, “The training is very easy and a Cisco CCIE expert would be able to use Arista right away, because we have similar command-line interfaces and operational look and feel. Where we don’t have to invent, we don’t.” So you can see why Cisco had reason to believe that its software was copied.

I have looked at the Cisco registrations and they are a thing of beauty. A summary of the registrations is below:

As you can see, Cisco, with great thoroughness, provided registrations for the various versions of the IOS software as well as IOS XE and IOS XR. Where appropriate, in the application field called “Previous Registration,” Cisco responded “yes” and stated that “this is a changed version of the work.”

From the list shown it’s clear there are no more than three works that Cisco claims infringed, IOS, IOS SR and IOS XE. The relationship of the three isn’t clear, and it could be that XR and XE contain IOS, as a Cisco white paper (and Reddit) suggest.

Nevertheless, Arista successfully confused the court over the relevance of the copyright registrations, convincing the court that (in Arista’s words) “the Cisco CLI is a lawyer created construct that simply mirrors Cisco’s copyright infringement allegations against Arista and has no life outside this litigation.” The court then proceeded to, of all things, stumble around compilation copyright, deciding that the Cisco CLI was a compiliation composed of pieces of the various registrations:

For purposes of this motion, Cisco asks the Court to determine that Cisco owns a valid copyright in “Cisco CLI.” But Cisco has failed to demonstrate that “Cisco CLI” is a compilation that its author(s) put together rather than a creature of its litigation strategy. In Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 111 S. Ct. 1282, 113 L. Ed. 2d 358 (1991), the Supreme Court addressed the issue of the copyrightability of compilations. The Court explained that to qualify for copyright protection, “a work must be original[.]” Id. at 345. “[O]riginality requires independent creation plus a modicum of creativity.” Id. at 346. Factual compilations can even possess originality when the author makes “choices as to selection and arrangement, so long as they are made independently by the compiler and entail a minimal degree of creativity[.]” Id. at 349. But the Court emphasized that copyright “protection is subject to an important limitation. The mere fact that a work is copyrighted does not mean that every element of the work may be protected. Originality remains the sine qua non of copyright; accordingly, copyright protection may extend only to those components of a work that are original to the author.” Id. at 348 (emphasis added). In this case, Cisco has not presented evidence of where Cisco CLI comes from or how and when it was compiled. Although Cisco has copyrights covering its IOS, Cisco does not have a single copyright registration covering the compilation it calls the Cisco CLI.

Perhaps not, but that’s because, presumably like any software, the IOS software and its CLI component evolved over time and features and functionality were added, updated and removed. But it’s still all one work, just different versions. That’s clear from the applications, which state that the reason for the new registrations is “changed version.”

Rather, the Cisco CLI is composed of pieces drawn from 26 different copyright registrations covering Cisco’s IOS. This may not have been an issue if Cisco had presented evidence about how the Cisco CLI compilation was created but it did not. See, e.g., Transcript 12:5-21 (disagreeing that the Cisco CLI is an attorney construct but not indicating how the Cisco CLI was compiled). In opposition, Arista argues that the Cisco CLI is a lawyer created construct that simply mirrors Cisco’s copyright infringement allegations against Arista and has no life outside this litigation.

Well of course Cisco wouldn’t have come up with evidence about its compilation because the Cisco CLI isn’t a “compilation” in the way the court is thinking of it. A “compilation” is “a work formed by the collection and assembling of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship.” Like an encyclopedia, anthology or database. A software program can indeed be a “compilation” and probably most are; they are complicated and consist of many components. But the Cisco CLI does not appear to be what the court is describing as the “compilation,” a random set of components pulled together piecemeal from 26 different registrations in order to gin up an infringement claim. It is rather simply an evolving part of the larger IOS work, with the changes to the CLI reflected in different registrations because they were introduced over the 12 year period of development that the certificates cover.

Luckily, though, all that happened is Cisco didn’t get summary judgment on copyrightability, but it hasn’t lost either.

To the extent that IOS, IOS XR and IOS XE do not share the same command line interface, then it is a valid infringement question whether one work that combines pieces of three works has copied enough of each to infringe one or all of them. But that’s an infringement question, not a scope of copyright question. Arista has also raised a number of challenges, like whether the copying was de minimis, whether elements of the CLI were scenes a faire, whether the content copied was copyrightable, and whether the copying was a fair use. All of that is well and good, all fair game for litigation over software. But throwing copyright registrations into the mix just further complicates what already is about the most complicated infringement analysis we ask anyone to perform, infringement of the copyright in software.

The irony is that had Cisco provided only one registration, the last one for IOS 15.4, it probably could have relied on the effective registration doctrine for its infringement claim. But courts have screwed up that theory too, so in an abundance of caution, and persuasive value, Cisco provided a slew of registrations. You just can’t win.

And for those of you unhappy with the Federal Circuit opinion in Oracle v. Google, this is going to be the Ninth Circuit’s bite at the apple. We may find out what the Ninth Circuit REALLY thinks about the copyrightability of “SSO” in software.

Update: Sharp reader Engel Nyst has pointed out that the case has a patent infringement claim to, so it too will go to the Federal Circuit.

Cisco Sys. v. Arista Networks, Inc., No. 14-cv-05344-BLF (N.D. Cal. Aug. 23, 2016).

This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.