Oracle America, Inc. v. Google Inc.

Filing 1085

RESPONSE (re #1080 MOTION in Limine to Exclude Evidence Regarding Compatibility Testing Suite ) filed byOracle America, Inc.. (Peters, Marc) (Filed on 5/6/2012)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

MORRISON & FOERSTER LLP
MICHAEL A. JACOBS (Bar No. 111664)
[e-mail]
KENNETH A. KUWAYTI (Bar No. 145384)
[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]
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.

18
19

UNITED STATES DISTRICT COURT

20

NORTHERN DISTRICT OF CALIFORNIA

21

SAN FRANCISCO DIVISION

22

ORACLE AMERICA, INC.

23

Plaintiff,

24

v.

25

GOOGLE INC.

26
27

Case No. CV 10-03561 WHA
ORACLE AMERICA, INC’S
OPPOSITION TO GOOGLE’S
MOTION IN LIMINE TO EXCLUDE
EVIDENCE REGARDING
COMPATIBILITY TESTING SUITE

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

28
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE TO EXCLUDE EVIDENCE
REGARDING CTS
CASE NO. CV 10-03561 WHA
pa-1527443

1
2
3

Oracle opposes Google’s motion to exclude evidence regarding the Android Compatibility
Test Suite (“CTS”).
I.

THE CTS PROVIDES PROOF OF INFRINGEMENT OF THE ’104
PATENT

4

Google’s Compatibility Definition Document (TX 2802) requires Android devices to meet
5
certain performance characteristics in order to be certified as Android-compatible: “Compatible
6
implementations must ensure not only that applications simply run correctly on the device, but
7
that they do so with reasonable performance and overall good user experience. Device
8
implementations MUST meet the key performance metrics of an Android 2.3 compatible device,”
9
as specified in the document. (TX 2802 at 122; see also Brady Topic 9 Dep. 121:6-128:1,
10
attached as Exhibit A; Morrill Dep. 167:3-169:20, attached as Exhibit B.)
11
The CTS ensures that devices meet these characteristics through testing. Some CTS tests
12
test the performance of Android phones—these tests are highly relevant to Oracle’s infringement
13
claims. The fact that OEM device manufacturers must pass these tests before calling their phone
14
“Android” suggests that they have not changed Android’s symbolic reference resolution
15
functions, which would significantly and negatively affect performance. The Android
16
Gingerbread source code (TX 47) includes a number of CTS performance tests:
17
18



0047\gingerbread23 - GOOGLE-0000000527\cts\tests\tests\performance\src\android\performance\cts\MultiAppStartu
pTest.java (Tests that restart of calculator takes less time than start of calculator)



0047\gingerbread23 - GOOGLE-0000000527\cts\tests\tests\performance2\src\android\performance2\cts\AppStartup.ja
va (Tests for average MusicBrowserActivity in com.android.music startup time of
less than 500 milliseconds)



0047\gingerbread23 - GOOGLE-0000000527\cts\tests\tests\performance3\src\android\performance3\cts\AppStartup.ja
va (Tests for average BrowserActivity in com.android.browser startup time of less
than 1300 milliseconds)



0047\gingerbread23 - GOOGLE-0000000527\cts\tests\tests\performance4\src\android\performance4\cts\AppStartup.ja
va (Tests for average ui.ConversationList in com.android.mms startup time of less
than 700 milliseconds)

19
20
21
22
23
24
25
26
27
28

ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE TO EXCLUDE EVIDENCE
REGARDING CTS
CASE NO. CV 10-03561 WHA
pa-1527443

1

1


2
3
4
5
6
7

0047\gingerbread23 - GOOGLE-0000000527\cts\tests\tests\performance5\src\android\performance5\cts\AppStartup.ja
va (Tests for average AlarmClock in com.android.alarmclock startup time of less
than 650 milliseconds)

The CTS thus provides circumstantial evidence that OEMs include on their devices the infringing
functionality that Google provides to them.
II.

THE CTS PROVIDES PROOF OF INFRINGEMENT OF THE ’520
PATENT

8

The Court should allow Oracle to present evidence regarding the Android CTS for the

9

’520 patent because the CTS directly tests the accused behavior in Android. Specifically, the

10

CTS includes a test for the fill-array-data instruction that the Android code accused of

11

infringement creates. (See TX 47 at 0047\gingerbread23 - GOOGLE-00-00000527\cts\tools\vm-

12

tests\src\dot\junit\opcodes\fill_array_data\Test_fill_array_data.java.) That OEMs pass these tests

13

suggests that they do not change Dalvik’s ability to execute the fill-array-data instruction, a

14

Dalvik instruction used for array initialization. Indeed, the final step in claim 1 of the ’520 patent

15

is “interpreting the instruction by a virtual machine to perform the static initialization of the

16

array.” By requiring OEMs to pass the CTS to certify their devices as Android-compatible,

17

Google actually requires them to perform a step of the patented method (a step that Google has

18

refused to admit that its licensees perform, the Court may recall). This evidence is highly

19

probative, and the jury should hear it.

20

Google’s only argument to exclude CTS evidence for the ’520 patent is that “the CTS

21

includes no tests directed at the Android Software Development Kit.” That is flat wrong. The

22

CTS includes tests directed to confirm the proper functioning of dx tool’s simulation of Java

23

bytecodes that participate in static array initialization. (See, e.g., bytecode tests in

24

0047\gingerbread23 - GOOGLE-00-00000527\cts\tools\dx-tests\src\dxc\junit\opcodes (opcodes

25

newarray, iastore, lastore, fastore, dastore, aastore, bastore, and sastore).) The dx tool is part of

26

the Android SDK, and is part of the software accused of infringing the ’520 patent. If any

27

developer or OEM were to change the dx tool, speculation for which there is no evidence, the

28

CTS tests suggest that the manner of creating the fill-new-array instruction would not be changed.
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE TO EXCLUDE EVIDENCE
REGARDING CTS
CASE NO. CV 10-03561 WHA
pa-1527443

2

1
2
3
4
5
6
7
8

III.

GOOGLE’S OBJECTIONS ARE INSUFFICIENT TO EXCLUDE THIS
EVIDENCE

Google claims that Professor Mitchell did not utilize performance tests to prove
infringement. In fact, Professor Mitchell refers to the performance tests of the CTS in paragraph
188 of his opening report, which Google attached to its motion as Exhibit C.
Google’s argument that testimony in this area would duplicate Phase I testimony (ECF
No. 1080 at 5) is false. Mr. Morrill’s testimony during Phase I did not cover tests for the Dalvik
Virtual Machine, let alone performance tests. Oracle expects Mr. Morrill’s and Mr. Brady’s
testimony to confirm the presence of the patented technology.

9
IV.
10
11

CONCLUSION

For the reasons stated above, the Court should deny Google’s motion to exclude the CTS
evidence from trial.

12
13

Dated: May 6, 2012

MORRISON & FOERSTER LLP

14
By: /s/ Marc David Peters
15
16

Attorneys for Plaintiff
ORACLE AMERICA, INC.

17
18
19
20
21
22
23
24
25
26
27
28
ORACLE’S OPPOSITION TO GOOGLE’S MOTION IN LIMINE TO EXCLUDE EVIDENCE
REGARDING CTS
CASE NO. CV 10-03561 WHA
pa-1527443

3

EXHIBIT A

Highly Confidential - Attorneys' Eyes Only

1

UNITED STATES DISTRICT COURT

2

NORTHERN DISTRICT OF CALIFORNIA

3

SAN FRANCISCO DIVISION

4
5

_______________________

6

ORACLE AMERICA, INC.,

7

Plaintiff,

8

vs.

9

GOOGLE, INC.,

10
11

)
)
) No. CV 10-03561 WHA

Defendant.

)
)

_______________________)

12
13
14

HIGHLY CONFIDENTIAL - ATTORNEYS' EYES ONLY

15
16
17

Videotaped Deposition of DANIEL MORRILL,

18

taken at 333 Twin Dolphin Drive, Suite

19

400, Redwood Shores, California, commencing

20

at 9:43 a.m., Tuesday, July 12, 2011,

21

before Leslie Rockwood, RPR, CSR No. 3462.

22
23
24
25

PAGES 1 - 248
Page 1
Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

MS. TERAGUCHI:

2

Foerster for plaintiff Oracle America.

3
4

MR. WEINGAERTNER:

Scott Weingaertner of

King & Spalding for Google.

5
6

Yuka Teraguchi of Morrison &

MR. KAMBER:

Matthias Kamber of Keker & Van

09:41:11

Nest for Google.

7

MR. HWANG:

8

THE VIDEOGRAPHER:

9

The witness will be sworn in and we may

10

Thank you, Counsel.

proceed.

09:41:21

11
12

Renny Hwang of Google.

THE REPORTER:

Would you raise your right

hand, please, Mr. Morrill.

13

You do solemnly state that the evidence you

14

shall give in this matter shall be the truth, the whole

15

truth and nothing but the truth, so help you God.

16

THE WITNESS:

17

THE REPORTER:

18
19

I do.
Thank you.

EXAMINATION
BY MR. MUINO:

20

Q.

Good morning, Mr. Morrill.

21

A.

Good morning.

22

Q.

Can you please state your full name for the

24

A.

My full name is Daniel Lawrence Morrill.

25

Q.

And what is your home address?

23

09:41:34

record.

09:41:43
Page 5

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

do you know if one of the optimizations performed by the

2

dexopt tool is to change symbolic references in the DEX

3

bytecode to the memory location of the data referred to

4

by those references?

5

A.

I don't know either way.

6

Q.

Okay.

15:33:39

With respect to the dexopt tool, do

7

you know if one of the optimizations that it makes is to

8

change certain bytecode instructions into in-line native

9

code?

10

A.

11
12
13

I don't know either way.
MR. WEINGAERTNER:

Q.

BY MR. MUINO:

15:33:54

Object to the form.

Okay.

Let's go back to the

Android 2.2 compatibility definition.

14

A.

Uh-huh.

15

Q.

And take a look at page 10, Section 5.

And

16

this is the application packaging compatibility.

17

entitled "Application Packaging Compatibility."

18

first sentence says:

19

install and run Android.APK files as generated by the

20

AAPT tool included in the official Android SDK."

15:34:06

21

It's
The

"Device implementations must

15:34:28

Do you see that?

22

A.

I do.

23

Q.

Second paragraph says:

"Device

24

implementations must not extend either the .APK, Android

25

manifest, the Dalvik bytecode formats in such a way that

15:34:42
Page 165

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

would prevent those files from installing and running

2

correctly on other compatible devices.

3

implementers should use the reference upstream

4

implementation of Dalvik and the reference

5

implementations package management system."

6

Device

15:34:56

Do you see that?

7

A.

I do.

8

Q.

Did you author this -- those two paragraphs?

9

A.

In this form, yes, but I did not originate

10
11

this section.
Q.

15:35:14

Okay.

The last sentence that I read there

12

with respect to the reference upstream implementation of

13

Dalvik, does that refer to the Dalvik virtual machine?

14
15
16

A.

It would refer to the source code of the

Dalvik virtual machine, yes.
Q.

Okay.

15:35:27

And pursuant to the CDD, Google is

17

instructing the device implementers should use the Dalvik

18

source code; is that correct?

19

A.

Device as it is written, device

20

implemented -- -- device implementers should use the

21

reference upstream implementation.

22

Q.

Okay.

23

A.

15:35:47

It's probably worth noting that this is

24

merely a reinforcement of the language in section --

25

well, in Section 1 in the introduction where we refer to,

15:36:06
Page 166

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

again, the reference implementation and the upstream

2

Android Open Source Project.

3

Q.

Okay.

Let's refer to Section 10 now.

4

is on page -- page 18.

5

look at Section 9 on page 17.

6

page 17.

7

I'm sorry, I misspoke.

This

Let's

It's the bottom of

15:36:35

And the first paragraph, the first sentence

8

there says:

"One of the goals of the Android

9

compatibility program is to enable consistent application

10

experience to consumers.

11

ensure not only that applications simply run correctly on

12

the device, but that they do so with reasonable

13

performance and overall good user experience."

14

Compatible implementations must

15:36:49

Do you see that?

15

A.

I do see that.

16

Q.

Did you write that portion?

17

A.

I actually think I did not, but I don't

18
19
20
21

15:37:03

recall clearly.
Q.

Okay.

Do you have an understanding of what

"reasonable performance" means here?
A.

15:37:23

In context it would refer to the contents of

22

the table immediately following, but I do not have a

23

precise answer for what "reasonable performance" would

24

be meant -- or would mean here.

25

Q.

Okay.

Is it a requirement of the CDD that

15:37:48
Page 167

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

Android devices meet the performance thresholds that are

2

shown in the chart below in this Section 9?

3

A.

Section 9 is included in the CDD, and the CDD

4

is the definition of a compatible device, yes.

5

Compatible devices must meet the performance

6

specifications in the table -- or in Section 9.

7

Q.

Okay.

15:38:08

And if you look at the table there,

8

the first row of the table, well, second row, the first

9

row of data is application -- says "application launch

10

time."

And the second column there says:

11

applications should launch within the specified time:

12

Browser less than 1300 milliseconds, MMS-SMS less

13

than 700 milliseconds, alarm clock less than

14

650 milliseconds."

15

"The following

Do you see that?

16

A.
Q.

Okay.

15:38:46

I do.

17

15:38:24

It's required for an

18

Android-compatible device that these applications launch

19

in these specified times in order to be compatible under

20

the CDD; is that right?

21

A.

That is correct.

22

Q.

15:38:59

Why is -- why is speed important to Google

23
24
25

for the launch of applications?
A.

Because an end-user might obtain a phone and

unknowingly purchase a, you know, poor quality phone,

15:39:23
Page 168

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

such as it might have an obsolete processor in it or it

2

might have a, you know -- excuse me -- a poor driver

3

implementation or some other defect that makes it

4

unreasonably slow or at least slower than its competitors

5

in its class.

15:39:43

6

This user would then install applications on

7

it, you know, such as from Android market and then judge

8

the quality of those applications in a negative light

9

because the device is slow.

In other words, the device's

10

poor performance would reflect -- in the user's eyes,

11

would reflect poorly on the application.

12

user were more informed and knowledgeable, they would

13

15:40:00

know that the blame should properly be placed on the OEM.

Whereas if the

14

The intent of this section in the CDD is to

15

make sure that Android devices meet a minimal threshold

16

of performance to rule out the scenario that I just

17

described.

18

applications will launch in a reasonable amount of time

19

and that the user will not blame third-party developers

20

for the errors or implementation issues of an OEM.

21
22

Q.

15:40:17

So that we can rely on the fact that

15:40:36

Is the launch speed of applications something

that's important to consumers?

23

A.

In the way I just described, yes.

24

Q.

Okay.

25

And are you aware of portions of

Android or elements of Android that are designed to

15:41:01
Page 169

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

increase the speed of the launch of applications?

2

MR. WEINGAERTNER:

3

THE WITNESS:

Objection to form.

Yes.

We have several

4

performance optimization avenues that we routinely pursue

5

in launching a device.

6
7
8
9

Q.

15:41:21

BY MR. MUINO:

Okay.

And tell me the ones

that you're aware of.
A.

Sure.

One example is the specific type and

nature of the -- you know, would we call the flash part

10

or the storage chip.

11

some storage types or classes are faster than others.

12

Some models are faster than others;

15:41:40

Another example is the -- the file system in

13

use on that partition.

In some cases, the -- for

14

example, I believe today we use the X4 -- EXT 4 file

15

system because it tends to be faster and more reliable

16

than the file systems we've used previously.

17

We have a variety of deferred loading, you

18

know, and load-on-demand techniques that we use in the

19

software, and we have a variety of pre-caching or

20

pre-loading techniques that we use, one of which is the

21

dexopt tool that you referred to previously, and another

22

is the Zygote technique that you also referred to

23

15:42:08

previously.

24

Q.

Okay.

25

A.

There are others, but those are two of the

15:42:31

15:42:43
Page 170

Veritext National Deposition & Litigation Services
866 299-5127

EXHIBIT B

Highly Confidential - Attorneys' Eyes Only

1

UNITED STATES DISTRICT COURT

2

NORTHERN DISTRICT OF CALIFORNIA

3

SAN FRANCISCO DIVISION

4
5

--------------------------

6

ORACLE AMERICA, INC.,

7

Plaintiff,

)
)

8

vs.

)

9

GOOGLE, INC.,

)

10
11

Defendant.

No. CV 10-03561 WHA

)

--------------------------

12
13
14

-- HIGHLY CONFIDENTIAL, ATTORNEYS' EYES ONLY --

15
16

Videotaped Federal Rule 30(b)(6), Topic 9,

17

deposition of PATRICK BRADY, taken at the law

18

offices of King & Spalding LLP, 333 Twin Dolphin

19

Drive, Redwood Shores, California, commencing at

20

2:24 p.m., Thursday, July 21, 2011, before

21

Leslie Rockwood, RPR, CSR No. 3462.

22
23
24
25

PAGES 1 - 149
Page 1
Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

Flexner for plaintiff Oracle America.

2
3

MR. KAMBER:

Matthas Kamber of Keker & Van

Nest for Google, Inc.

4

THE VIDEOGRAPHER:

Thank you.

5

Will the reporter please swear the witness.

6

THE REPORTER:

7

You do solemnly state that the evidence you

8

shall give in this matter shall be the truth, the whole

9

truth and nothing but the truth so help you God --

14:26:21

Raise your right hand, please.

10

THE WITNESS:

11

THE REPORTER:

12

THE VIDEOGRAPHER:

13

Please proceed.

14
15

I do.

09:38:38

Thank you.
-- thank you.

EXAMINATION
BY MR. NORTON:

14:26:39

16

Q.

Good afternoon.

17

A.

Good afternoon.

18

Q.

As you heard, my name's Fred Norton.

19

represent Oracle.

20

record, obviously you were already deposed on another

21

subject matter earlier today, and so I'm going to try not

22

to go over things that were already covered.

23

time to time I'll have to ask you a question just to set

24

up the context.

25

time as I can be and get out of here as quickly as we can

I

As I mentioned before we went on the
14:26:48

But from

But I'll try to be as respectful of your
14:27:07
Page 4

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

Q.

And do you recall when this was?

2

A.

I want to say that these conversations came

3

up.

Things with Vodafone move slow at times.

4

I want to say it spanned the second half of 2009 and

5

2010.

6

So this --

17:53:48
Q.

Okay.

And if you turn to the prior page,

7

still in Exhibit 230, there's a Section 9 called

8

"Performance Compatibility."

9

A.

Yep.

10

Q.

Would you just read aloud the three sentences

11

that appear there?

12

MR. KAMBER:

13

THE WITNESS:

14

Object to the form.
The three sentences that appear

directly underneath 9 -- the section heading?

15
16

17:54:06

Q.

BY MR. NORTON:

Yes.

Just read it out loud

17:54:22

please.

17

MR. KAMBER:

18

THE WITNESS:

Same objection.
As the document reads:

"One of

19

the goals of the Android compatibility program -- of the

20

Android compatibility program is to enable consistent

21

application experience to consumers.

22

implementations must ensure not only that applications

23

simply run correctly on the device but that they do so

24

with reasonable performance and overall good user

25

experience.

17:54:30

Compatible

Device implementations must meet the key

17:54:50
Page 121

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

performance metrics of the Android 2.2 compatible device

2

defined in the table below."

3
4

Q.

BY MR. NORTON:

And then there's a table that

has criteria set forth there; right?

5

A.

Yeah.

17:55:04

6

Q.

Now, is there any testing that Google does to

7

ensure -- and does the CTS test to see whether devices

8

satisfy the criteria set forth in Section 9?

9

A.

I believe it does test some of these, yeah.

10

Q.

All right.

And the specific criteria that

11

are set out here are application launch time and then

12

17:55:25

simultaneous applications?

13

A.

Yes.

14

Q.

And under application time, the standard is

15

that the following applications should launch within the

16

specified time, the browser less than 1300 milliseconds

17

MMS/SMS less than 700 milliseconds and alarm and clock

18

less than 650 milliseconds; is that right?

19

A.

Alarm clock is a single application here.

20

Q.

Oh, thank you.

17:55:37

21
22

17:55:58

And so that's -- in the case of the browser,
1.3 seconds to connect, 1300 milliseconds?

23

A.

If my math is correct, yes.

24

Q.

And why does Google care how quickly these

25

things launch?

17:56:15
Page 122

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

A.

Well, when -- one of the things that we found

2

when we had the OEMs making Android devices -- porting

3

the Android platform to their devices and running this is

4

sometimes the -- the -- this talks about user experience.

5

But it's also developer experience.

6

17:56:38

So Rovio, for example, is very upset when

7

their app runs very slowly.

And they get complaints.

8

they'll get returns, if it's a paid app.

9

will be rated poorly.

Or

Or their app

And, you know, the app developer

10

has basically, you know, nothing much they can do; right,

11

if the device is just too slow?

12

17:56:57

And so here we're trying to put some, you

13

know, we think fairly conservative and not onerous

14

requirements on OEMs to maintain some minimum level of

15

performance for users and application developers so that

16

it's a reasonable experience.

17

Q.

Okay.

17:57:20

Other than the concern that developers

18

will have a bad experience because they might get

19

complaints or returns, are there other reasons why Google

20

wants to -- has as a goal to enable consistent

21

application experience to consumers?

22

A.

17:57:41

It's really -- from a compatibility

23

standpoint, it's really about developers.

So I'll

24

state -- you know, consumers indirectly drive, you know,

25

the developer experience, I guess I should say.

The

17:58:02
Page 123

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

consumer experience drives the developer experience, if

2

that makes sense.

3

So when a device -- you know, an application

4

developer creates an application that -- a great

5

application, but the device executes that application

6

poorly, the developer has a bad experience, because they

7

may get rated poorly.

8

may, you know, have upset users.

9

what we're trying to ensure doesn't happen here.

10

Q.

All right.

They may suffer returns.

17:58:18

They

And so that's generally

And at the risk of pushing too

11

hard on the obvious, why does Google care about whether

12

or not the developers have bad experiences with their

13

17:58:37

applications running on a given device?

14

A.

Because, I mean, generally, any ecosystem

15

needs developers -- an app ecosystem needs developers.

16

And developers who have a bad experience leave.

17
18

Q.

17:58:55

The other performance criterion here is

simultaneous applications.

19

A.

Yep.

20

Q.

And the standard there in the compatibility

21

definitions is that when multiple applications have been

22

launched, relaunching an already running application

23

after it has been launched must take less than the

24

17:59:10

original launch time.

25

A.

This is an interesting metric.

Yes.

17:59:24
Page 124

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1
2

Q.

So in other words, when you relaunch it, it

will start faster than it did the first time?

3

A.

Yes.

I think maybe relaunching it isn't -- I

4

don't know -- an exact term here.

5

Android does -- or that Android developers do in Android

6

is, you know, launch one application from another.

7

One of the things that
17:59:43

So I may have an application that I can

8

launch out to send an email or do something, and then I

9

hit "back."

And Android has this notion of an affordance

10

for the user to go back.

And you'll go back to the

11

application you were in.

So it's kind of integration

12

between applications.

13

18:00:03

In this case, what we're trying to say is if

14

I open whatever application it may be, and it has an

15

option to email someone, and I hit the email button and

16

it launches the email application, when I go back, I

17

don't want that -- you know, that old application to take

18

longer than it initially took to launch.

19

sense?

20

Q.

18:00:19

Does that make

So I'm in the browser, and I click on an

21

email address.

It opens the email program.

22

to send an email right now, and I want to go back to the

23

browser.

24

to wait for the browser to take launch as long as it took

25

the first time; right?

18:00:33

So I go back to the browser.

I don't want

And I don't want

18:00:50
Page 125

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

A.

Right.

2

Q.

Okay.

3

So again, this is about -- it's a

particular application of speed, but it's about speed?

4

A.

Yes.

5

Q.

And, in fact, both of these criteria for

6
7

performance are about -- I'm sorry.
A.

Go ahead.

18:00:58

Sorry.

Well, I'd say it's about speed, but it's

8

really -- here, I mean, it's about the ability to run

9

multiple applications at once.

10

Q.

Okay.

18:01:11

11

A.

So a big part of Android is multitasking.

12

And, you know, when trying to define that, what we're

13

trying to figure out is, well, okay, do you need to be

14

able to run -- you know, in defining the requirements,

15

able to run multiple applications at once?

16

default user experience in Android is a single foreground

17

application.

18

here is, say, well, when you go back to a previously open

19

application, it should launch faster than if it was from

20

a cold start.

21

it's about, performance.

22

Q.

Well, the

18:01:32

And so the way we're trying to codify that

It's perhaps a -- but yes, that's what

18:01:49

And strictly speaking, what Section 9 is

23

doing with it is not compatibility, per se, but just how

24

well the device performs when consumers have them.

25

MR. KAMBER:

Objection to form.

18:02:14
Page 126

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

THE WITNESS:

No.

I think it is

2

compatibility.

3

for a long time.

4

perspective, you can have two devices that may be

5

technically able -- you know, one device that runs an

6

application well and one device that technically executes

7

the byte code but does it extremely slowly.

8

And this is something that we've debated
From an application developer

18:02:30

And, you know -- and we need, you know, from

9

an application developer's perspective, they maybe, you

10

know, don't consider those two devices to be compatible,

11

because they can't write their applications the same

12

way -- they can't use the same application on both

13

devices.

14

18:02:45

And so what we're doing here is putting, in

15

all honesty, a very low bar on -- on, you know, the

16

minimum performance requirements you must have.

17
18

18:02:59

So, I mean, from our sense, it really is
about compatibility here.

19

And, you know, we constantly get complaints

20

from the developer that, "Hey, when I -- when I want to

21

play a sound on this given device, sometimes there's a

22

2-second lag before the song gets played," you know, "and

23

on this other device, there's no lag."

24

the same -- the same application is executing on both

25

devices, but they're incompatible from a user experience

18:03:19

That's, you know,

18:03:36
Page 127

Veritext National Deposition & Litigation Services
866 299-5127

Highly Confidential - Attorneys' Eyes Only
1

perspective.

2

Q.

And that's a problem for developers.
All right.

So stepping back a little bit,

3

this -- the OEM is free to customize Android in various

4

ways, but one of the things that they have to do is make

5

sure that when they customize Android, they do not impair

6

the performance of the platform with respect to its

7

ability to launch programs quickly and multi-task?

8
9

A.

18:03:58

According to these specific provisions, yes,

they need to meet those.

10

Q.

And these particular applications that are

11

called out here, the browser, MMS and the alarm clock,

12

those are just proxies for the general ability of the

13

platform to meet overall speed expectations.

14

18:04:15

fair?

15

A.

Yes.

Is that

And I believe those applications would

16

have been adjusted over time.

17

trying to establish some standard reference that we can

18

use to have an objective, you know, assessment across

19

different devices.

20

Q.

18:04:38

But yes, I mean, we're

So the idea is if it can launch the browser

21

as fast as we need to launch the browser, it will

22

18:04:52

probably run Angry Birds okay?

23

A.

I wish that was a valid assertion.

But, you

24

know, I mean, at some point you have to be realistic;

25

right?

And we're not going to be able to test these two

18:05:08
Page 128

Veritext National Deposition & Litigation Services
866 299-5127