Oracle America, Inc. v. Google Inc.

Filing 1149

RESPONSE (re #1146 MOTION in Limine To Exclude Evidence of Google's Android-Related Revenues Until Oracle Proves a Causal Nexus with the Infringement ) ORACLE AMERICA, INC.S OPPOSITION TO GOOGLES MOTION IN LIMINE TO EXCLUDE ANDROID REVENUES filed byOracle America, Inc.. (Holtzman, Steven) (Filed on 5/14/2012)

1
2
3
4
5
6
7

B O I E S ,
S C H I L L E R
&
F L E X N E R
L L P
B O I E S ,
S C H I L L E R
&
F L E X N E R
L L P
O A K L A N D ,
C A L I F O R N I A
O A K L A N D ,
C A L I F O R N I A

8
9
10
11
12
13
14
15
16
17
18
19

MORRISON & FOERSTER LLP
MICHAEL A. JACOBS (Bar No. 111664)
[e-mail]
MARC DAVID PETERS (Bar No. 211725)
[e-mail]
DANIEL P. MUINO (Bar No. 209624)
[e-mail]
755 Page Mill Road, Palo Alto, CA 94304-1018
Telephone: [phone] / Facsimile: [phone]
BOIES, SCHILLER & FLEXNER LLP
DAVID BOIES (Admitted Pro Hac Vice)
[e-mail]
333 Main Street, Armonk, NY 10504
Telephone: [phone] / Facsimile: [phone]
STEVEN C. HOLTZMAN (Bar No. 144177)
[e-mail]
1999 Harrison St., Suite 900, Oakland, CA 94612
Telephone: [phone] / Facsimile: [phone]
ALANNA RUTHERFORD (Admitted Pro Hac Vice)
[e-mail]
575 Lexington Avenue, 7th Floor, New York, NY 10022
Telephone: [phone] / Facsimile: [phone]
ORACLE CORPORATION
DORIAN DALEY (Bar No. 129049)
[e-mail]
DEBORAH K. MILLER (Bar No. 95527)
[e-mail]
MATTHEW M. SARBORARIA (Bar No. 211600)
[e-mail]
500 Oracle Parkway, Redwood City, CA 94065
Telephone: [phone] / Facsimile: [phone]
Attorneys for Plaintiff
ORACLE AMERICA, INC.
UNITED STATES DISTRICT COURT

20

NORTHERN DISTRICT OF CALIFORNIA

21

SAN FRANCISCO DIVISION

22
23

ORACLE AMERICA, INC.

Case No. CV 10-03561 WHA

24
25
26

Plaintiff,
v.

ORACLE AMERICA, INC.’S OPPOSITION
TO GOOGLE’S MOTION IN LIMINE TO
EXCLUDE ANDROID REVENUES

GOOGLE, INC.

27
Defendant.

Dept.: Courtroom 8, 19th Floor
Judge: Honorable William H. Alsup

28
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

In the Copyright Act, Congress decided to require that an infringer account for its ill-gotten

2

profits. Found by a jury and by the Court to be a copyright infringer, Google has filed yet another

3

motion, styled as a “motion in limine,” in an attempt to be relieved of that burden altogether.

4

Google’s motion to exclude Android financial evidence from Phase 3 of the trial merely rehashes the

5

arguments in its unsuccessful motion for summary judgment. Google’s motion should be denied.

6

I.

8

F L E X N E R
&

Heedless of that admonition, Google has filed a motion in limine that is not just disguised as a

11
C A L I F O R N I A

S C H I L L E R

summary judgment motions” as motions in limine. (Order Regarding Pretrial Filings (Dkt. No. 384).)

10

O A K L A N D ,

B O I E S ,

The Court previously warned the parties not to bring “categorical motions and disguised

9
L L P

7

Google’s “Motion in Limine” Is A Procedurally Improper and Legally Inadequate ReHash of Its Failed Motion for Summary Judgment.

summary judgment motion – it is actually cut and pasted from the summary judgment motion Google

12

filed just two days ago. This morning, the Court denied that motion from the bench, stating: “That is

13

more properly brought as a Rule 50 motion after all the evidence is put into the record.” (RT at

14

3912:21-24.) Google’s “motion in limine” is little more than a reply brief on the summary judgment

15

motion Google already lost.

16

In its summary judgment motion, Google argued that Oracle should not be allowed to present

17

to the jury its infringer’s profits case because Oracle supposedly cannot meet its burden of

18

demonstrating a “causal nexus” between infringement and Android revenues. The Court denied that

19

motion. In its in limine motion, Google argues that Oracle should not be allowed to present to the

20

jury the portion of its infringer’s profits case that includes Android revenues, because – again –

21

Oracle supposedly cannot meet its burden of demonstrating a “causal nexus” between infringement

22

and Android revenues.

23

As Google makes the exact same argument as before, its motion fails for the exact same

24

reason as before. “Causal nexus” does not mean what Google thinks it means. (It does not even

25

appear in the Copyright Act, although Google erroneously cites the Act for that proposition. (See

26

Dkt. No. 1146 at 3).) In an effort to show that Oracle cannot establish the supposedly necessary

27

causal nexus, Google repeats its argument that rangeCheck and the decompiled files represent a very

28

small percentage of the total volume of code in Android. But the amount of infringing material in the
1
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

defendant’s product, be it great or small, does not matter to the plaintiff’s burden under 17 U.S.C. §

2

504(b). Once Oracle identifies revenues that are attributable to the product that exploits the

3

infringing code, the Copyright Act puts the burden on Google, not Oracle, to separate out any portion

4

of Android’s “profit attributable to factors other than the copyrighted work.” 17 U.S.C. § 504(b).

5

There is no exception to this burden-shifting provision for small infringements.

6

In its Opposition to Google’s summary judgment motion, Oracle pointed out that numerous

F L E X N E R
&

(2nd Cir. 2001)—establish that the relevant causal nexus in an indirect-profits case is between the

11
C A L I F O R N I A

S C H I L L E R

Volkswagen of Am., Inc., 336 F.3d 789 (8th Cir. 2003); and Davis v. The Gap, Inc., 246 F.3d 152

10

O A K L A N D ,

1985); Cream Records, Inc. v. Joseph Schlitz Brewing Co., 864 F.2d 668 (9th Cir. 1989); Andreas v.

9

B O I E S ,

appellate cases—including Frank Music Corp. v. Metro-Goldwyn-Mayer, Inc., 772 F.2d 505 (9th Cir.

8
L L P

7

infringing product (here, Android) and the revenues, not between the copyrighted code and the

12

revenues. (See Dkt. No. 1135 at 4–12.) Accordingly, every appellate case to consider this issue has

13

shown that, if the infringing product generates revenue, it does not matter how small the infringement

14

is as compared to the infringing product. The burden of apportionment is on the infringer. (See id.)

15

In addition to Frank, Cream, Andreas, and Davis, in Bergt v. McDougal Littell, 661 F.Supp.2d

16

916 (N.D. Ill. 2009), for example, the plaintiff met its burden by establishing defendant’s gross

17

revenues from the sales of an approximately 1,250 page textbook that reproduced his copyrighted

18

painting on one page. Id. at 927–28. The court explained that the plaintiff’s “failure to present a

19

buyer whose decision to buy the textbook was solely made on the inclusion of the painting is

20

immaterial in both direct and indirect profit cases to initially establish [defendant’s] gross revenue.”

21

Id. at 928 n.8. Rather, it was the defendant’s burden to show that none of the revenue it received

22

from the sale of the textbook was due to the use of plaintiff’s painting. Id. at 929. Indeed, both

23

parties’ experts admitted that in neither of their experience had a single painting been the basis of the

24

decision to buy a textbook, yet the court denied defendant’s motion for summary judgment on

25

infringer’s profits, based on nothing more than the testimony of plaintiff’s experts that photographs,

26

images, and illustrations are one of several things that would be a basis for choosing a textbook. Id.

27

at 928-29. See also Ordonez-Dawes v. Turnkey Prop., Inc., (S.D. Fla. Apr. 22, 2008) (sufficient

28

nexus between defendant’s gross revenues from the sale of three homes where the homes sold after
2
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

publication of an advertisement that included three of plaintiff’s copyrighted photographs). The

2

burden then shifts to the infringer to prove … that some or all of that revenue is attributable to factors

3

other than infringement of the copyrighted work”).

4

Google claims that Mackie v. Rieser, a case decided two years before the Ninth Circuit’s

F L E X N E R
&

photograph appeared in page twelve, directly following a page that contained information about the

10

symphony’s “Pops” series. Id. There was no other connection between the street art and the Pops

11
C A L I F O R N I A

symphony series for the Seattle Symphony. 296 F.3d 909, 912 (9th Cir. 2002). The street art

9

S C H I L L E R

plaintiff’s street art, was incorporated as part of a montage in a direct mailing that promoted a

8

O A K L A N D ,

of the Copyright Act. Google is wrong. In Mackie, the infringed work, a photograph of the

7

B O I E S ,

decision in Polar Bear and on which the Polar Bear court expressly relied, supports its interpretation

6

L L P

5

performances, yet Mackie claimed profits for the entire 1996-97 Symphony season, as well as future

12

seasons, not just for Pops performances but for all performances. Id. at 913. The connection between

13

the infringement and the revenues was therefore quite indirect; the infringing component was not

14

incorporated into the revenue-generating work, and Mackie claimed profits not only from the series

15

that may have been promoted by the infringement, but for unrelated series as well. The Ninth

16

Circuit held that before Mackie could collect those indirect profits, he had to demonstrate that “the

17

infringing acts had an effect on profits.” Id. at 915.

18

Mackie is nothing like this case. The profits here are not indirect. Android actually

19

incorporates rangeCheck, and Google’s experts have admitted that the decompiled files have value to

20

Android. Oracle is aware of no case in which profits were considered “indirect” when the infringing

21

work, which generated the revenues, actually contains the copies and actually benefitted from the

22

infringement. There is simply not as attenuated a nexus between the infringement and the revenues

23

as there was in Mackie. In fact, Mackie’s own expert had admitted that he “could not ‘understand’

24

how it would be possible to establish a causal link between the Symphony’s infringing use of ‘The

25

Tango’ and any Pops series revenues generated through the inclusion of the collage in the direct-mail

26

literature.” Id. at 916. There is no such admission here.

27
28

Moreover, the standard Mackie announced, and Polar Bear applied, is not as stringent as
Google argues. There must simply be a showing, even in a case where profits are extraordinarily
3
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

attenuated from the infringement, that the infringing acts “had an effect” on revenues. Id. at 915.

2

There is no requirement that Oracle prove, to a certainty, that Android earned revenues because of

3

rangeCheck and the decompiled files; indeed, Polar Bear made crystal clear that there is no such

4

requirement. Polar Bear, 384 F.3d at 715 (“there is no requirement that Polar Bear put Timex

5

customers on the witness stand to testify that they purchased watches because of Timex’s use of

6

‘PaddleQuest’ images.”).

7
8

&

First, Google claims that the ratio of rangeCheck lines to Android lines is smaller than the
ratio of Kismet minutes to Hallelujah Hollywood minutes. But nothing in either of the Ninth

11
C A L I F O R N I A

S C H I L L E R
B O I E S ,

distinguish Frank Music on its facts simply shows how on point that case is.

10

O A K L A N D ,

F L E X N E R

L L P

9

Google also claims that Frank Music must not apply to this case, but Google’s attempt to

Circuit’s opinions in Frank Music suggests that the relative quantity of infringing to non-infringing

12

material would have any bearing at all on the causal-nexus inquiry. In fact, the Ninth Circuit

13

expressly rejected this type of tallying when it considered the district court’s apportionment of profits

14

(where the burden was on the defendant):

15

If the district court relied exclusively on a quantitative comparison and failed to consider the
relative quality or drawing power of the show’s various components, it erred.

16
Frank Music, 886 F.2d at 1548. Google cannot wave away Frank Music with a comparison that the
17
court never endorsed, and in fact discouraged.
18
Google also attempts to distinguish Frank Music by claiming that the defendant there
19
admitted that the infringement enhanced its revenues. But the very passage that Google quotes shows
20
that, at best, Google stands in the same shoes as MGM. Google points out that the defendants in
21
Frank Music admitted that the hotel and gaming operations of the hotel were “materially enhanced by
22
the popularity of the hotel’s entertainment[, including] ‘Hallelujah Hollywood[.]” Frank Music I,
23
772 F.2d at 517 (emphasis added). The defendant never admitted that Kismet enhanced its revenues,
24
only that its entire entertainment program, including the entire revue that included Kismet, did. Id.
25
Here, of course, Google has made the exact same admission: it admitted that Android materially
26
enhanced its revenues. (Schmidt at RT 1456:10-19, 1458:9-1459:3 (admitting, among other things,
27
28
4
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

that Android covers its costs and a “whole bunch more”).) Google has no grounds on which it can

2

distinguish Frank Music.
Otherwise, Google’s cut-and-paste brief adds nothing new to the arguments the Court has

3
4

already considered and rejected. Oracle has already addressed those arguments and the relevant case

5

law in detail. (See Dkt. No. 1135 at 4–12; see also Dkt. No. 1106 (May 8, 2012, brief outlining these

6

issues). Oracle incorporates those briefs by reference.

7

II.

The Proper Vehicle For Google’s Argument Is A Rule 50 Motion Directed at the
Evidence – Not a Motion in Limine to Exclude Evidence.

8
The Court has now admonished Google twice that the proper vehicle for its arguments is a

&

11
C A L I F O R N I A

S C H I L L E R
B O I E S ,

10

O A K L A N D ,

F L E X N E R

L L P

9

12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

Rule 50 motion after the evidence is in. (RT at 3721:17-19; RT at 3912:21-24.) Instead, Google
slaps a new label on an old argument, and asks the Court not to let the evidence in at all.
III.

Oracle Will Present Evidence At Trial That Meets Its Burden on Causal Nexus
For the reasons stated above and in Oracle’s prior briefs, the plaintiff’s burden to show a

causal nexus is not as great as Google claims, and proof of causal nexus is not foreclosed merely
because the copying is small compared to the financial success of the infringing product. (See Dkt.
No. 1135 at 4–12. See also Cream Records, 864 F.2d 668 (infringer’s profits available for infringed
ten notes), Andreas, 336 F.3d 789 (infringer’s profits available for an infringed thirty three-word
phrase), Frank Music, 864 F.2d 668 (infringer’s profits available for six minutes of “tribute” songs).
It does not matter how many lines of code the files contain, nor does it matter how those numbers of
lines compare to Android. It is sufficient to show that Android itself exploits the infringed code.
Moreover, although Oracle need not prove any particular amount of revenues attributable to
the infringement, Oracle will offer evidence that Google’s infringement of the code files has
contributed meaningful value to Android. TimSort and ComparableTimSort, the files that the jury
found infringed, were “much, much faster” than the existing sort functions in Android; Bloch
bragged that “Seeing a speed up 20 times faster was not uncommon. I’ve seen things even more than
that and, you know, twice as fast is a good sort of overall estimate if you average everything.”
(Bloch at RT 812:19-813:3.) In fact, when he wanted to be added to the Android team, Bloch wrote

28
5
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

an email to Andy Rubin and tried to prove his value specifically by boasting about the performance

2

benefits that TimSort delivers: “I am currently working on a drop-in replacement for Harmony's sort

3

function, which has demonstrated a huge (up to 20X) performance improvement on G1 hardware. . . .

4

This will be my first contribution to Android.” (TX 184.) In his deposition, Bloch admitted that

5

TimSort’s performance enhancements could be expected to affect user experience of an Android

6

device. (Bloch Dep. 163:8-164:19.) Further, the exact copying of rangeCheck itself was important to

7

TimSort because Bloch wanted to be sure that rangeCheck threw the exact same exceptions that users

8

of Java would be familiar with (Bloch Dep. 181:7-182:11.)

&

always been, the proper place to discuss the value of what Google took. The jury is entitled to infer,

11
C A L I F O R N I A

S C H I L L E R
B O I E S ,

Oracle has not yet had an opportunity to prove these facts at trial, as Phase 3 is, and has

10

O A K L A N D ,

F L E X N E R

L L P

9

from this evidence and more, that TimSort contributed value to Android, and that the only way to

12

achieve the performance enhancement as quickly as needed was to copy rangeCheck. Moreover,

13

Google engaged in willful copyright infringement, and even knowingly permitted former Sun

14

employee Bloch to work simultaneously on Java and Android, despite the fact that Google

15

recognized a “conflict” in doing so. (TX 1060). The jury should be able to infer from the fact that

16

Google – a company that claims it prides itself on its ethics, its commitment to the Java community,

17

and its commitment to the protection of intellectual property – engaged in such willful and reckless

18

conduct that it saw value in the copyrighted code it obtained as a result. (See also Oracle’s

19

Opposition (Dkt. No 1135 at 15–16) (describing these, and other, facts from which the jury could

20

infer value).

21

Similarly, the decompiled “test” files—which are not test files in Java—are not valueless to

22

Google, even if, as Google insists, they are not currently present on phones. Even if they are test files

23

in Android, Dr. Mitchell explained why this does not mean they are unimportant: “testing is a very

24

important part of the software development. It’s expensive. And software companies want to do it

25

correctly so that the code that they ship is bug free and usable to their customers.” (Mitchell at RT

26

1330:15–24.) Google developers would have benefitted from these decompiled files because, if

27

indeed they are test files, “if this helped them test other code they were developing, and speed up and

28

lessen the cost of testing and quality assurance, then that would have a big value to them.” (Mitchell
6
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

at RT 1330:25-1331:5.) Even Google’s expert, in his deposition, stated that software testing is “a

2

very important component of developing software” and that “I think most people would agree that

3

testing software and having files specifically for testing is good engineering practice.” (Astrachan

4

Dep. Tr. 160:3–17.)
The jury should also be able to infer value from Google’s willful infringement, including the

5

F L E X N E R
&

10.) (See also Oracle’s Opposition to Google’s Motion for Summary Judgment (Dkt. No. 1135 at

10

16–17) (describing these, and other, facts from which the jury could infer value). Indeed, Noser’s

11
C A L I F O R N I A

S C H I L L E R

available on the public Android code repository that it owns and controls. (Bornstein at RT 1832:3-

9

O A K L A N D ,

copyright notices on all of its files (see, e.g., TX 623.2), and that Google still makes the files

8

B O I E S ,

fact that Google knew that the Noser engineers were “super shady” (TX 281), that Sun put obvious

7

L L P

6

flat-fee contracts actually may have given them an incentive to cheat. (See TX 30, 70, 74 (Noser

12

statements of work).)

13

Google also ignores the fact that the purpose of the disgorgement provision of the Copyright

14

Act is to make copying “worthless” to the infringer. Bucklew v. Hawkins, Ash, Baptie & Co., LLP.,

15

329 F.3d 923, 933 (7th Cir. 2003) (“Remember that the purpose of allowing suit for the infringer’s

16

lost profits is to make infringement worthless to the infringer.”). Neither copying rangeCheck nor the

17

decompiled files was worthless to Google.

18

IV.

19

There Is No Undue Prejudice to Google In Allowing Oracle to Present Its Infringer’s
Profits Case.

20

The only ostensible prejudice that Google identifies is that it might have to disclose its non-

21

public Android financial statements – some of which are already in evidence. That a defendant’s

22

revenues will be offered into evidence is, under the Copyright Act, a natural and inevitable

23

consequence of engaging in copyright infringement. Moreover, the Court has told both parties “[t]he

24

mere fact that it revealed something about your finances, that’s nothing. Come on. The public can see

25

that all day long.” (RT 5:23-6:16) If Google believes, after the close of evidence, that there is no

26

causal nexus, then it may move for judgment as a matter of law at that time.

27

V.

Conclusion

28
7
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA

1

Many months ago, the Court reminded the parties that “Motions in limine made at this stage

2

must be ripe, not premature or dependent upon the way evidence develops at trial.” (Order

3

Regarding Pretrial Filings (Dkt No. 384).) Google was told unambiguously that it would have to wait

4

until the close of evidence to bring this motion. Unsatisfied with that ruling, Google then repackaged

5

the exact same arguments into a motion in limine. The Court should not entertain this attempt to

6

resurrect a denied motion.

7
8

&

11
C A L I F O R N I A

S C H I L L E R
B O I E S ,

10

O A K L A N D ,

F L E X N E R

L L P

9

Dated: May 14, 2012

BOIES, SCHILLER & FLEXNER LLP
By: Steven C. Holtzman
Steven C. Holtzman
Attorneys for Plaintiff
ORACLE AMERICA, INC.

12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
8
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE RE: ANDROID REVENUES
CASE NO. CV 10-03561 WHA