Oracle America, Inc. v. Google Inc.

Filing 1156

TRIAL BRIEF re Oracle's Failure of Proof on Causation by Google Inc.. (Van Nest, Robert) (Filed on 5/15/2012)

1
2
3
4
5
6
7
8
9
10
11
12

KEKER & VAN NEST LLP
ROBERT A. VAN NEST - # 84065
[e-mail]
CHRISTA M. ANDERSON - # 184325
[e-mail]
DANIEL PURCELL - # 191424
[e-mail]
633 Battery Street
San Francisco, CA 94111-1809
Telephone:
415 391 5400
Facsimile:
415 397 7188

KING & SPALDING LLP
DONALD F. ZIMMER, JR. - #112279
[e-mail]
CHERYL A. SABNIS - #224323
[e-mail]
101 Second Street, Suite 2300
San Francisco, CA 94105
Tel: [phone]
Fax: [phone]

KING & SPALDING LLP
SCOTT T. WEINGAERTNER
(Pro Hac Vice)
[e-mail]
ROBERT F. PERRY
[e-mail]
BRUCE W. BABER (Pro Hac Vice)
1185 Avenue of the Americas
New York, NY 10036
Tel: [phone]
Fax: [phone]

IAN C. BALLON - #141819
[e-mail]
HEATHER MEEKER - #172148
[e-mail]
GREENBERG TRAURIG, LLP
1900 University Avenue
East Palo Alto, CA 94303
Tel: [phone]
Fax: [phone]

13
14

Attorneys for Defendant
GOOGLE INC.

15

UNITED STATES DISTRICT COURT

16

NORTHERN DISTRICT OF CALIFORNIA

17

SAN FRANCISCO DIVISION

18

ORACLE AMERICA, INC.,

19
20
21
22

Plaintiff,

Case No. 3:10-cv-03561 WHA
GOOGLE’S BRIEF RE ORACLE’S
FAILURE OF PROOF ON CAUSATION

v.
GOOGLE INC.,

Dept.:
Judge:

Courtroom 8, 19th Floor
Hon. William Alsup

Defendant.

23
24
25
26
27
28
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA
665801.01

1

Contrary to Oracle’s counsel’s statements in court today, Google has never suggested that

2

Oracle deserves no remedy for the minimal acts of copyright infringement found at trial. Google

3

has made only the point that, because the infringing works—nine lines of rangeCheck code in the

4

two TimSort files and eight decompiled test files that were never used on any Android phone—

5

are so insignificant, Oracle cannot prove either that it suffered any actual damages or that Google

6

earned any revenues attributable to the infringement.

7

This is exactly why the Copyright Act offers the option of statutory damages—to provide

8

a remedy even where a copyright holder cannot prove actual harm. “Statutory damages are

9

intended as a substitute for profits or actual damage. When injury is proved but neither the

10

infringer’s profits nor the copyright holder’s actual damages can be ascertained, an award of

11

statutory ‘in lieu’ damages is mandatory.” Frank Music Corp. v. Metro-Goldwyn-Mayer, Inc.,

12

772 F.2d 505, 520 (9th Cir. 1985) (citing Russell v. Price, 612 F.2d 1123, 1131-32 (9th Cir.

13

1979)). The infringement verdicts here, as to rangeCheck and the test files, are tailor-made for a

14

statutory damages award, because Oracle hasn’t met and can’t meet the standard for entitlement

15

to an award of infringer’s profits.

16

A.

17

To recover any of Google’s profits, Oracle first must prove a causal nexus between
the infringing works—TimSort, ComparableTimSort and the eight test files—and
some identifiable amount of Google revenue.

18

Google has addressed the basic legal framework regarding an award of infringer’s profits

19

before, but provides the following, more targeted analysis in response to the discussion at today’s

20

hearing. The governing legal standard has been consistently applied by the Ninth Circuit for over

21

a decade: to establish any entitlement to profits, a copyright plaintiff first must offer evidence that

22

the infringer made some amount of revenue attributable to the copyrighted work that was found

23

infringed. See 17 U.S.C. § 504(b). Once the plaintiff makes that showing, the burden shifts to

24

the defendant to prove deductible costs incurred in earning that revenue and, if there are any

25

profits left over, apportioning those profits to deduct amounts attributable to factors other than the

26

infringement. See id.

27
28

The Ninth Circuit explained this rule most recently in Polar Bear Products, Inc. v. Timex
Corp., 384 F.3d 700, 710-12 (9th Cir. 2004), explaining that, “[a]lthough the statute references
1
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

only the broad term ‘gross revenue,’ to conclude that a copyright plaintiff need only provide the

2

company’s overall gross revenue, without regard to the infringement, would make little practical

3

or legal sense.” Id. at 711. Instead, “the causation element of the statute serves as a logical

4

parameter to the range of gross profits a copyright plaintiff may seek.” Id. “The standard is

5

straightforward: a copyright plaintiff is bound to no more and no less than its statutory obligation

6

to demonstrate a causal nexus between the infringement and the profits sought.” Id. at 712. Even

7

where an infringer “derived some quantum of profits because its infringement was part of” a

8

larger part of its business, “it nevertheless remains the duty of the copyright plaintiff to establish a

9

causal connection between the infringement and the gross revenue reasonably associated with

10

the infringement.” Id. at 715 (citing On Davis v. The Gap, Inc., 246 F.3d 152, 160 (2d Cir.

11

2001)) (emphasis added); see also Mackie v. Rieser, 296 F.3d 909, 914-15 (9th Cir. 2002)

12

(plaintiff first must demonstrate “that the infringing acts had an effect on profits before the parties

13

can wrangle about apportionment”).

14

Oracle is wrong to suggest, as it did to the Court today, that an award of infringer’s profits

15

is mandatory because (1) Google’s Android software contains infringing material, even if only a

16

tiny amount; and (2) Android is a “revenue-generating product.” Oracle misunderstands how

17

both copyright damages and Google’s Android business work. Google does not generate any

18

revenue directly from the Android software. Google gives the software away for free, as an open-

19

source product. It is only once the Android software is adopted by handset makers and

20

incorporated onto mobile devices that Google can earn money, when users of Android handsets

21

perform Google searches, click on Google-hosted ads, or buy applications from GooglePlay

22

(formerly Android Market). There is no dispute that nearly all of Google’s Android revenues

23

come from these indirect sales of downstream Google-hosted services on mobile phones, not

24

direct sales of software. And there is no allegation, let alone a verdict, of copyright infringement

25

as to any of the ads or applications from which Google makes money directly. Accordingly,

26

under Ninth Circuit law, all these categories of profits from Android are indirect profits. See

27

Mackie, 296 F.3d at 914 (defining “direct profits” as those “that are generated by selling an

28

infringing product”). The Ninth Circuit has held over and over again that a copyright plaintiff
2
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

may recover indirect profits only if it can establish that those profits have some causal link to the

2

infringing work. See id. at 914-15; Polar Bear, 384 F.3d at 710-16.

3

As Google has argued previously, there is no causal nexus here. The jury did not find that

4

Google’s Android platform infringed Oracle’s Java platform. It found only that two Android

5

files, TimSort and ComparableTimSort, infringed a single Java file called Arrays.java. See Final

6

Charge to the Jury (Phase One) [Dkt. 1018] ¶ 29, at 15 (“For purposes of Question No. 3, the

7

‘work as a whole’ is the compilable code for the individual file except for the last two files listed

8

in Question No. 3, in which case the ‘work as a whole’ is the compilable code and all the English-

9

language comments in the same file.”); Special Verdict Form [Dkt. 1018] at Question 3(a)

10

(calling for verdict on infringement by TimSort files). The Court later granted judgment as a

11

matter of law as to eight decompiled test files, ruling that those files infringed eight additional

12

Java files. May 11, 2012 Order [Dkt. 1123]. These are the infringing works for purposes of any

13

indirect profits analysis. To even get started making out an infringer’s profits case, Oracle must

14

prove a causal connection between those works and an identifiable amount of Google revenue.

1

15
16

Mackie v. Rieser is the case that controls the outcome here and makes it clear that Oracle
cannot establish any nexus. Mackie is perfectly analogous to this case:
•

In both cases, the infringement was a barely noticeable component of a larger work.
In Mackie, it was a single photograph of a copyrighted work incorporated into a
collage on a single page of promotional brochure for the Seattle Symphony. See 296
F.3d at 912-13. Here, it is nine lines of rangeCheck code out of 15 million lines on
the Android platform and eight decompiled test files that never made it onto a phone.

•

17

In both cases, the larger whole that incorporated the small amount of infringing
material was given away for free in hopes of generating downstream revenue. In
Mackie, the promotional brochure was given away by mail to promote Symphony
subscriptions. See id. at 912-13. Here, Google gives away the Android software,
which eventually enables it to host advertising on Android handsets and sell
applications to users of such handsets.

18
19
20
21
22
23
24
25
26
27
28

1

As Google has noted before, it was Oracle that sought, over Google’s objection, the instruction
deeming individual Android and Java files to be stand-alone works for purposes of the jury’s
infringement analysis. RT 2414:20-2418:13 (Charging Conference). The Court overruled
Google’s objection and gave Oracle the instruction it requested. Id. Oracle undoubtedly wanted
a charge that defined the “work as a whole” narrowly—as individual Java files, not the entire Java
platform—in order to increase its chances of an infringement verdict on its literal copying claims
and decrease the odds that the jury would excuse a minute amount of infringement as de minimis.
It got the infringement verdict that it wanted, but it must play by the same rules with respect to its
damages claims.
3
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

•

In both cases, the plaintiff sought indirect profits. In Mackie, the plaintiff wanted a
cut of the Symphony’s revenue purportedly tied to the promotional brochure that
contained the infringing photo. See id. at 913. Here, Oracle is seeking some of the
downstream revenue Google has made from applications and advertising.

•

In both cases, the plaintiff failed to offer any testimony or evidence establishing a
causal link between the minute infringing material and the downstream revenues.

2
3
4
5
6
7

Ultimately, the court held that Mackie had failed to satisfy his burden to establish that the
Symphony generated any gross revenues causally connected to the infringement:

11

Intuitively, we can surmise virtually endless permutations to account for an
individual’s decision to subscribe to the Pops series, reasons that have nothing to
do with the artwork in question. For example, was it because of the Symphony’s
reputation, or the conductor, or a specific musician, or the dates of the concerts, or
the new symphony hall, or the program, or the featured composers, or community
boosterism, or simply a love of music, or . . . ? In the absence of concrete
evidence, Mackie’s theory is no less speculative than our effort in this paragraph to
enumerate even a relatively short list of the myriad factors that could influence an
individual’s purchasing decisions.

12

Id. at 916. Even had Mackie been able to offer evidence tending to show that a percentage of the

13

Symphony’s sales were tied to the infringing brochure, “such a rudimentary analysis cannot

14

determine how many of those individuals subscribed because of Rieser’s work.” Id. at 916

15

(emphasis in original). Because the collage that incorporated a photograph of Mackie’s artwork

16

was “but one page in a multi-page brochure that advertised a series of concerts that were

17

unrelated to the artwork itself,” the Ninth Circuit flatly rejected Mackie’s theory as “[r]ank

18

speculation” that was legally insufficient to support a claim for indirect profits. Id. Given the 15

19

million lines of code in the Android software, and all the features enabled by that software, there

20

are “virtually endless permutations” in this case “to account for an individual’s decision” to click

21

on a Google-hosted ad or buy an application through GooglePlay. There is certainly no reason to

22

think the nine-line rangeCheck function (out of over 15 million lines of code in Android, RT

23

2179:19-23 (Astrachan)) had anything to do with it—and it is certain the eight test files didn’t,

24

since they never appeared on any Android handset. RT 1319:15-1320:6 (Mitchell) If the Mackie

25

photo was a needle in the brochure’s haystack, Android contains many thousands of haystacks.

8
9
10

26

Oracle has tried to run away from the causal-nexus requirement by citing mostly to cases

27

involving advertising and promotional materials, see, e.g., Polar Bear, 384 F.3d at 712-13;

28

Andreas v. Volkswagen of Am., Inc., 336 F.3d 789, 796-98 (8th Cir. 2003); On Davis, 246 F.3d at
4
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

160, but none of those cases dispute that a nexus is required for recovery of infringer’s profits. In

2

fact, those cases explicitly affirm that requirement. See Polar Bear, 384 F.3d at 711 (before

3

apportionment takes place, the copyright plaintiff “must first show a causal nexus between the

4

infringement and the gross revenue.”); Andreas, 336 F.3d at 796 (“The plaintiff has the burden to

5

demonstrate a nexus between the infringement and the indirect profits before apportionment can

6

occur”). Instead, because those cases all involved infringing materials that were the centerpiece

7

of an advertising campaign directed to promotion of a particular product, the courts reasonably

8

concluded that the nexus requirement was satisfied. But Mackie was an advertising case as well,

9

and the lesson of Mackie is that, when infringing material is so minute and barely identifiable,

10

even in a promotional piece, a court will not simply assume a link between the infringement and

11

indirect profits. Such a link has to proved with actual, meaningful evidence.

12

This settled legal framework makes clear that Oracle has not done nearly enough to shift

13

the burden to Google. Oracle cannot rely on Google’s gross revenue generally, or even Google’s

14

gross revenue from the Android platform as a whole; it must identify some amount of gross

15

revenue that is causally linked to the infringement. “Only then would [the infringer] bear the

16

responsibility for apportioning profits.” Polar Bear, 394 F.3d at 715. As Judge Posner explained

17

almost thirty years ago,

18

It was not enough to show [defendant’s] gross revenues from the sale of
everything he sold, which is all, really, that [plaintiff] did. If General Motors
were to steal your copyright and put it in a sales brochure, you could not just put a
copy of General Motors’ corporate income tax return in the record and rest your
case for an award of infringer’s profits.

19
20
21

Taylor v. Meirick, 712 F.2d 1112, 1122 (7th Cir. 1983). This rule bars Oracle’s strategy here.

22

B.

Oracle has no witness who could offer any testimony to establish the required nexus
between Google’s revenues and the infringement.

23
Here, Oracle has never even attempted to establish, at any stage of the case, that Google’s
24
revenues are causally linked to rangeCheck or the test files. Its technical expert Dr. Mitchell
25
offered opinions that rangeCheck and the test files were identical to, and likely copied from, the
26
Java source files. Mitchell Opening Report ¶¶ 233-239 (rangeCheck); id. ¶¶ 241-248 (test files).
27
He opined that rangeCheck, at least, appeared on handsets, id. ¶ 240, though he never expressed
28
5
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA
665801.01

1

any opinion that it was important to handset functionality. The closest he came to expressing any

2

opinion that rangeCheck has any value were his opinions that (1) rangeCheck was “qualitatively

3

significant to” the Arrays.java file, because “it is called nine times by other methods in the class,”

4

id. ¶ 235; and (2) “it is certainly possible that some amount of trial and error went into figuring

5

out how to arrange the tests in this code so that the most informative error condition is reported.”

6

Mitchell Opposition Report ¶ 86. Even then, he conceded that “I have not studied the use of this

7

code in any detail.” Id. This is hardly a ringing endorsement. With respect to the test files, his

8

opinion about their value was limited to anodyne statements that software testing is important and

9

“clearly someone or some group of engineers found it useful to produce them.” Id. ¶ 95. Nothing

10

in his testimony vouches for any contribution by either rangeCheck or the test files to Android as

11

a whole, much less provides any basis for quantifying that value or separating out the contribution

12

of these files from the millions of other lines of code on the Android platform.

13

That’s it for Oracle’s expert testimony on these files. Oracle’s damages expert, Dr. Iain

14

Cockburn, could have tried to calculate a gross revenue figure linked to rangeCheck or the test

15

files, but he never made the attempt. His first report effectively ignored copyright damages

16

entirely, stating only that “because Android infringes the expression embodied in Oracle’s

17

copyrighted class libraries or APIs, Oracle may also be entitled to receive Google’s profits

18

attributable to the infringement.” First Cockburn Report (May 20, 2011) ¶¶ 349-50 (emphasis

19

added). In other words, to the extent Dr. Cockburn drew any causal link between any alleged

20

infringement and Google revenues, it was a link between only the allegedly infringed API

21

packages, not rangeCheck or the test files. Dr. Cockburn’s second report offered an opinion only

22

as to alleged gross revenue from all Android ad and applications sales, and again purported to link

23

that revenue only to Google’s use of the Java API packages. Second Cockburn Report (Sept. 15,

24

2011) at 186-88, ¶¶ 463-68, & Ex. 22 (Android gross revenue calculation through 2011).

25

In response to Dr. Cockburn’s repeated failure to offer any opinion at all assigning a value

26

to any of the allegedly literally copied code—including rangeCheck and the test files—Google

27

asked the Court to preclude Dr. Cockburn from offering such testimony at trial. Google Supp. Br.

28

(Oct. 20, 2011) [Dkt. 549] at 14-15. The Court agreed, barring Oracle from offering any expert
6
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

opinion on damages for any of Oracle’s literal copying claims, including rangeCheck or the eight

2

decompiled test files:

3

5

Dr. Cockburn has not adequately valued that [allegedly copied] code in his report
and cannot do so at trial. This order holds that the jury will be instructed that if
Google is found not liable for infringing the selection, arrangement, and structure
of the API packages, then Dr. Cockburn’s copyright damages analysis is
inapplicable.

6

Jan. 9, 2012 Order [Dkt. 685] at 10. In sum, with respect to Oracle’s experts, Dr. Mitchell didn’t

7

offer any opinions about a causal link between the copied files and Google revenues, and Dr.

8

Cockburn is barred from even trying.

4

9

In the end, the only thing a review of Oracle’s expert reports reveals is that, before failing

10

to get an infringement verdict on its main copyright claim, neither Oracle nor its experts ever

11

contemplated the idea that Google’s use of rangeCheck or the test files could have had any

12

material effect on Google’s revenues. Instead, Oracle made the legitimate tactical choice to focus

13

both its liability and damages cases on its claim that Google infringed the structure, sequence, and

14

organization of 37 Java API packages. But that decision has consequences now. Moreover, not

15

only did Oracle prepare no expert and disclose no fact witness on this subject, it never disclosed

16

the underlying damages theory at any time during discovery.

17

Oracle first disclosed its damages theory, as Rule 26(a) requires, in its December 2, 2010

18

initial disclosures. There, Oracle did not refer to the rangeCheck method or the decompiled files

19

at all. Indeed, the only fact Oracle disclosed relating to damages at all was that Eric Schmidt had

20

said Android’s revenues were “large enough to pay for all of the Android activities and a whole

21

bunch more.” See Declaration of David Zimmer (“Zimmer Decl.”) in Support of Google’s

22

Motion for Summary Judgment [Dkt. 1125] Decl. Ex. A (Oracle’s 12/2/10 Disclosures) at 7:6-7.

23

Oracle also noted that it had not completed its damages calculation, because “it will require

24

expert evaluation of information in Google’s possession.” Id. at 6:10-11. Oracle twice

25

supplemented this response, but neither supplemental response referred to the rangeCheck

26

method or the decompiled files, much less stated any facts suggesting a causal link between that

27

material and any Android revenue. See id. Exs. B (Oracle’s 6/3/11 Disclosures) & C (Oracle’s

28

8/10/11 Disclosures). Both supplemental responses referred to Oracle’s damages expert’s reports,
7
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

which, as noted above, omitted any discussion of literal copying damages. See id.

2

Oracle’s failure to disclose any evidence supporting this theory continued throughout the

3

discovery period. On January 6, 2011, Oracle answered Google’s first set of interrogatories,

4

including Google’s Interrogatory No. 1, which asked for a detailed statement of “Oracle’s factual

5

bases for each allegation of damage or harm that Oracle claims to have suffered as a result of any

6

act or omission of Google.” Zimmer Decl. Ex. D (Oracle’s 1/6/11 Interrogatory Responses) at

7

1:7-8. With regard to its “claim for recovery of Google’s profits attributable to the infringement,”

8

Oracle disclosed various facts about Google’s revenues and business model, but stated no facts

9

tying those revenues to rangeCheck or the eight decompiled files. Id. at 3:5-4:8. Oracle also

10

noted that it would provide further information about its claimed damages, including about

11

“disgorgement of Google’s profits from the infringement,” when it served its damages expert’s

12

report. Id. at 5:24-6:4. Oracle twice supplemented this response, but neither supplemental

13

response disclosed any facts related to the rangeCheck method or the decompiled files. Zimmer

14

Decl. Ex. E (Oracle’s 4/25/11 Supp. Interrogatory Responses) at 5:11-6:16; id. Ex. F (Oracle’s

15

7/29/11 Supp. Interrogatory Responses) at 6:1-18.

16

In short, Oracle has never disclosed any intent to rely on testimony from any fact witness

17

to establish a nexus between Google’s infringement and its Android-related revenues. Instead,

18

Oracle has always taken the position that it would rely on expert testimony to connect the dots

19

between any alleged infringement and Google’s revenues, but, as already noted several times, Dr.

20

Cockburn never connected those dots and eventually the Court ruled he was forbidden from

21

trying. To the contrary, Dr. Cockburn only ever opined that a causal link exists between Google’s

22

Android-related revenues and Google’s use of the structure, sequence, and organization of 37 API

23

packages in the J2SE platform—material Google was not found liable for infringing. Third

24

Cockburn Report (Feb. 9, 2012) at 227-37, ¶¶ 624-649.

25

C.

26

All the evidence in the record shows there is no causal link between Google’s use of
rangeCheck or the eight test files and any Android revenues.
Although the Court invited Oracle to make an offer of proof as to what evidence it could

27
offer to prove a causal link, Google is confident Oracle will come up empty. In addition to
28
8
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA
665801.01

1

having no expert to testify on the subject and failing to disclose any facts or witness in discovery

2

who would offer such testimony, the evidence already in the trial record makes clear that

3

rangeCheck and the eight test files were always completely insignificant to Android.

4

1.

5

To begin with, as the Court is well aware by now, rangeCheck makes up a tiny fraction of

6

the Android platform in a quantitative sense. It is nine lines of source code out of a platform that

7

contains more than 15 million lines, RT 2179:19-23 (Astrachan), amounting to .00006% (or six-

8

ten millionths) of Android. That does not qualify even as de minimis.

9

The trial record shows that rangeCheck has no link to Android revenue.

But rangeCheck is indisputably insignificant as a qualitative matter too. The testimony at

10

trial, from both sides of the aisle, has been unequivocal that rangeCheck is a “very short simple

11

method” that checks three parameters of an array: the starting point, the end point, and that the

12

end point is greater than the starting point. RT 813:7-8, 815:5-9 (Bloch). Josh Bloch, who wrote

13

rangeCheck, testified that “[a]ny competent high school programmer could write” that method.

14

RT 815:13-16 (Bloch). Even Oracle’s expert Dr. Mitchell conceded that “a good high school

15

programmer” could write rangeCheck with guidance. RT 1316:24-25 (Mitchell).

16

In trying to conjure the illusion that rangeCheck is important during this afternoon’s oral

17

argument, Oracle’s counsel misstated two established facts about those nine lines. Google

18

assumes that Oracle will correct the record as to both issues in its brief tonight, but the evidence

19

in the record on those points is as follows:

20

First, Oracle’s counsel contended that rangeCheck “was something that was important to

21

TimSort and ComparableTimSort, that did have a significant performance advantage.” RT

22

4228:11-13. This is false. The only testimony in the record, from any witness, shows that, in

23

addition to being trivial to create and easily replicable by a beginner, rangeCheck offers no

24

performance benefit to Android. Josh Bloch, who wrote TimSort and ComparableTimSort,

25

testified that the TimSort files are useful to Android because they make arrays sorts much faster.

26

RT 812:19-813:3 (Bloch). But Bloch also made clear that not one bit of that performance

27

improvement is due to rangeCheck, as opposed to the other 900-plus lines of code in those files.

28

RT 814:1-4 (Bloch).
9
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

Second, Oracle’s counsel suggested that Oracle could link rangeCheck to Google’s profits

2

for Android because those nine lines of code “benefited Google because it accelerated the time

3

frame that Android could be finished” and helped Android get to market more quickly. RT

4

4221:21-22; see also id. 4225:15-20. That is also counterfactual, given the undisputed that

5

rangeCheck was not even in Android when Google announced the platform in November 2007.

6

Neither was rangeCheck in Android when the first Android phones were released in October

7

2008 and Google made its code available to handset partners for inclusion on phones. Bloch did

8

not even join the Android team until December 2008 or January 2009. RT 733:8-11 (Bloch). He

9

finished TimSort at some point in early 2009, at which point he contributed that file both to Sun’s

10
11

OpenJDK project and to Android. RT 822:4-9 (Bloch).
Further, rangeCheck has been out of the current release of Android for about a year. RT

12

825:8-19 (Bloch); RT 1700:25-1701:10 (Rubin). The proven facts that both the platform’s initial

13

adoption by handset makers and carriers and its recent growth happened in the absence of

14

rangeCheck are additional reasons why no reasonable jury could link any of Android’s profits to

15

the temporary inclusion of those nine lines in Android.

16

In its previous briefs, Oracle has cited Dr. Mitchell’s testimony that the rangeCheck code

17

is allegedly called 2,600 times when powering on a smartphone or starting an emulator. Oracle

18

Br. [Dkt. 1106] at 3:13-17 (citing RT 1329:5-21 (Mitchell)). This is vacuous. It establishes

19

nothing to cite an arbitrary number of calls to a given method in the absence of context, and

20

neither Dr. Mitchell nor any other witness testified whether rangeCheck was called any more or

21

less than any other method in the Android software, during the startup sequence or any other

22

time. Given the capabilities of modern computer microprocessors, 2,600 calls to a given function

23

during the startup sequence of a smartphone could be a low number relative to other functions.

24

There is nothing in the record to enable the Court or a jury to tell either way. Equally, just

25

because a software function is called frequently does not mean it is important; it would stand to

26

reason that a trivial nine-line piece of code that accomplishes a Programming 101 parameter test,

27

like rangeCheck does, might be invoked fairly frequently. Dr. Mitchell never opined that there is

28

any correlation between the number of calls to a function and its significance, much less that
10
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA

665801.01

1

rangeCheck itself is significant. He certainly did not say that rangeCheck offered a performance

2

boost to Android—and Bloch, who wrote it, made clear it does not.

3

The record already shows that rangeCheck was an utterly trivial (and temporary) part of

4

Android. Oracle has no evidence to dispute what the Court has already heard and cannot show

5

any nexus between rangeCheck and any Google revenues.

6
7

2.

The trial record shows that the eight decompiled test files have no link to
Android revenue.

Given its comments at today’s hearing, Oracle appears to believe that its claim for profits
8
is stronger as to the eight test files than as to rangeCheck. In fact, its test-files claim is even
9
weaker, for an obvious reason: those eight test files never appeared on a single Android handset.
10
RT 1318:20-1320:6 (Mitchell). In fact, there is no evidence in the record, and Oracle has never
11
disclosed any during discovery, to suggest that those eight test files were ever used by Google or
12
anyone else during the process of developing or testing the Android software or any Android
13
handsets. But even if Oracle could show they were used by some handset maker for some testing
14
function, that would not come close to establishing a nexus. Because the files never appeared on
15
a handset, they could not have formed the basis of any consumer’s decision to buy a handset,
16
much less assisted any consumer in accessing the downstream Google-hosted ad and applications
17
from which Android generates its revenue.
18
Oracle has also argued that Google must have thought the test files were valuable because
19
Google deliberately copied them. RT 4227:14-25. Again, this is false. The record establishes
20
that the files were not decompiled by Google; they were contributed by Noser, a third-party
21
contractor. RT 1698:11-1699:1 (Rubin); id. 1810:11-13 (Bornstein). In its contract with Noser,
22
Google had expressly instructed Noser not to decompile or copy other parties’ protected files, but
23
only to do original work or use open-source code. TX 2765 at 11; RT 1696:21-1698:10, 1701:1824
21 (Rubin); RT 1798:17-1803:6 (Bornstein). There is no evidence that Google intended these
25
materials to end up in Android or knew what Noser had done. Google removed the test files from
26
the current version of Android once Oracle alerted Google that they had been decompiled rather
27
than written from scratch. RT 1807:25-1810:1 (Bornstein).
28
11
GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA
665801.01

1

Neither is there any merit to Oracle’s argument that, even though the eight test files were

2

“not in Android … there’s no doubt it accelerated Android” in getting to market. RT 4228:1-3.

3

This is made up out of whole cloth. Even if Oracle could prove the test files were used by Google

4

or anyone else, which it can’t, it has no evidence or expert testimony proving that the test files

5

helped Android get to market faster and, if so, by how much and what financial impact that might

6

have had. Oracle cannot simply look to approximations of Google’s present-day Android

7

revenue, as it suggested at today’s hearing, and conclude that getting Android to market even two

8

days earlier in 2007 would have netted Google some large amount of money. RT 4222:1-16.

9

In any event, the record disproves that Google obtained any time or cost savings because

10

of Noser’s decompilation of the test files. Two of the three Google-Noser statements of work in

11

the record, TX 30 (Statement of Work for class libraries) at 8 & TX 70 (SOW for Compatibility

12

Test Suite) at 10, show that Google paid Noser flat rates and required set deliverable dates. In

13

other words, whatever Noser may have done in performing these contracts, it would not have

14

saved Google any time or money. The third Google-Noser contract, TX 74 (SOW for code

15

integration), did not have set deliverable dates, but it capped Google’s payment obligation at

16

$60,000. To the extent Noser cut corners in performing that contract, the most Google could have

17

saved would have been $60,000—an amount consistent with a modest statutory damages award.

18

For all the above reasons, Oracle both is required to, and cannot, prove the required causal

19

nexus between Google’s infringement of the nine-line rangeCheck function and/or the eight test

20

files and Google’s revenues from Android. All the evidence in the record points the other way.

21

Oracle still may elect statutory damages at any time prior to judgment, see 17 U.S.C. § 504(c)(1),

22

but it is not entitled to an award of infringer’s profits.

23
24
25

Dated: May 15, 2012

KEKER & VAN NEST LLP
By:

26

/s/ Robert A. Van Nest
ROBERT A. VAN NEST
Attorneys for Defendant
GOOGLE INC.

27
28
12

GOOGLE’S BRIEF RE ORACLE’S FAILURE OF PROOF ON CAUSATION
Case No. 3:10-CV-03561 WHA
665801.01