Oracle Corporationv.Thought, Inc.

Patent Trial and Appeal Board

Apr 23, 2015

10624753 (P.T.A.B. Apr. 23, 2015)

Trials@uspto.gov Paper No. 44
571.272.7822 Filed: April 23, 2015

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

ORACLE CORPORATION,
Petitioner,

v.

THOUGHT, INC.,
Patent Owner.
____________

Case IPR2014-00117
Patent 6,985,912 B2
____________

Before KARL D. EASTHOM, BARBARA A. BENOIT, and
STACEY G. WHITE, Administrative Patent Judges.

WHITE, Administrative Patent Judge.

FINAL WRITTEN DECISION
35 U.S.C. § 318(a) and 37 C.F.R. § 42.73

IPR2014-00117
Patent 6,985,912 B2

2
I. INTRODUCTION
Oracle Corporation (“Petitioner”) filed an amended Petition
requesting inter partes review of claims 1–16 (“challenged claims”) of U.S.
Patent No. 6,985,912 B2 (Ex. 1001, “the ’912 Patent”) pursuant to
35 U.S.C. §§ 311319. Paper 6 (“Pet.”). Thought, Inc. (“Patent Owner”)
did not file a Preliminary Response. On April 25, 2014, we instituted inter
partes review of claims 1–3, 6, and 71 on three grounds of unpatentability
alleged in the Petition (Paper 16, “Dec. to Inst.”). After institution of trial,
Patent Owner filed a corrected Patent Owner Response (Paper 31, “PO
Resp.”), to which Petitioner filed a Reply (Paper 34, “Reply”).
Patent Owner filed a Motion for Observations (Paper 37, “Obs.”) on
certain cross-examination testimony of Petitioner’s declarant, to which
Petitioner filed a Reply (Paper 41, “Obs. Reply”).
An oral hearing was held on November 18, 2014, a transcript of which
has been entered into the record. Paper 43 (“Tr.”).
The Board has jurisdiction under 35 U.S.C. § 6(c). In this Final
Written Decision, issued pursuant to 35 U.S.C. § 318(a) and 37 C.F.R.
§ 42.73, we determine Petitioner has shown by a preponderance of the
evidence that challenged claims 1–3, 6, and 7 are unpatentable.
A. Related Proceedings
The ’912 Patent is the subject of Thought, Inc. v. Oracle Corp., No.
3:12-05601 (N.D. Cal.). Pet. 1; Mandatory Notice (Paper 10).

1 Prior to institution, Petitioner filed an unopposed motion to withdraw its
alleged grounds of unpatentability against all claims except 1–3, 6, and 7.
Paper 15. We granted Petitioner’s motion and, therefore, we considered
only challenges leveled against claims 1–3, 6, and 7. Dec. to Inst. 3.
IPR2014-00117
Patent 6,985,912 B2

3
B. The ’912 Patent
The ’912 Patent is directed to “enhanced database access and
performance when correlating or translating one database to another
database or to an object programming application.” Ex. 1001, Abstract. It
“provide[s] a simple wrapper or facade interface and API to be utilized to
coordinate relational data systems and object data systems and to control
multiple software modules that are utilized to load, insert, update[,] and
delete data from data systems and objects or object models in object
programming systems.” Id. at 3:64–4:2.
One of the objects of the invention is to “provid[e] an improved
caching system for multiple users.” Id. at 4:15. Data or objects populate the
cache when the data or objects first are read from a data source, or the data
may be preloaded as directed by the user. Id. at 4:24–27. Once a unit of
data or object is updated, the cached versions of that unit of data or object
are invalid. Id. at 16:50–51. “[A] cache notification system[] termed a UNS
(Unit Notification System)” provides “notification that a unit of data or
object in a cache is invalid.” Id. at 5:26–31. Upon receiving a notification
of invalid data, the user can set the cache to delete and reload the invalidated
data or wait until the user attempts to read that unit of data before updating
the cache. Id. at 16:52–55.
C. Illustrative Claim
Claim 1 of the ’912 Patent is the only independent claim challenged,
and it recites:
1. A computer system or network comprising a local or
distributed data or object caching system wherein:
IPR2014-00117
Patent 6,985,912 B2

4
(a) data or objects are (i) placed in a user cache in the
caching system when the data or object is first read from a data
source by a user as a lazy cache, or (ii) preloaded as directed by
a user,
(b) each unit of data or object that is placed in a user
cache is registered with a unit notification system (UNS) as
being present in the cache and the UNS is connected to a Java
Messaging Service (JMS) or to any other reliable messaging
service available in the object programming language system
environment,
(c) when any unit of data or object is updated in a user
caches and persisted (saved) to a primary data source the UNS
sends a message to all other user caches that may exist and
contain the updated unit of data or object to register in the user
cache a message that the unit was updated by another user and
is noted as being invalid in the cache, and
(d) other user caches that have been notified as described
in (c) will read from the data source the updated unit of data or
object and update the cache only when a user attempts to access
the invalidated unit or data or object in a user cache, may
automatically update the user cache as soon as notified as has
been pre-set in the caching system by the user of the user cache.
D. Alleged Grounds of Unpatentability Instituted for Trial
The following table summarizes the three challenges to patentability
that were instituted for inter partes review:
Reference(s) Basis Claim(s) Challenged
Franklin2 § 102 1–3

2 Michael J. Franklin et al., Transactional Client-Server Cache Consistency:
Alternatives and Performance, ACM, 1–44 (1997) (“Franklin,” Ex. 1005).
IPR2014-00117
Patent 6,985,912 B2

5
Reference(s) Basis Claim(s) Challenged
Franklin and JVM3 § 103 2
Franklin and Chang4 § 103 6 and 7
II. ANALYSIS
A. Claim Interpretation
In an inter partes review, claim terms in an unexpired patent are given
their broadest reasonable construction in light of the specification of the
patent in which they appear. 37 C.F.R. § 42.100(b); see Office Patent Trial
Practice Guide, 77 Fed. Reg. 48,756, 48,766 (Aug. 14, 2012). Pursuant to
that standard, the claim language should be read in light of the specification,
as it would be interpreted by one of ordinary skill in the art. In re Suitco
Surface, Inc., 603 F.3d 1255, 1260 (Fed. Cir. 2010). Thus, absent any
special definitions, we give claim terms their ordinary and customary
meaning, as would be understood by one of ordinary skill in the art at the
time of the invention. In re Translogic Tech., Inc., 504 F.3d 1249, 1257
(Fed. Cir. 2007). Any special definitions for claim terms must be set forth
with reasonable clarity, deliberateness, and precision. See In re Paulsen,
30 F.3d 1475, 1480 (Fed. Cir. 1994).
For purposes of the Decision to Institute, we construed the terms
(1) “when a user attempts to access the invalidated unit or data or object in a
user cache, may automatically update the user cache,” (2) “lazy cache,”
(3) “unit notification system (UNS),” (4) “reliable messaging service,” and

3 Tim Lindholm et al., THE JAVA™ VIRTUAL MACHINE SPECIFICATION 1–247
(Lisa Friendly ed., Addison-Wesley 1997) (“JVM,” Ex. 1009).
4 U.S. Patent No. 5,627,979 (“Chang,” Ex. 1010), issued May 6, 1997.
IPR2014-00117
Patent 6,985,912 B2

6
(5) “provides transparent persistence.” Dec. to Inst. 7–12. “Patent Owner
agrees with the Board’s construction of the terms construed by the Board [in
the Decision to Institute] and respectfully submits that further interpretation
of that construction is needed in two instances,” the construction of “reliable
messaging service” and “transparent persistence.” PO Resp. 9–10; see id. at
10–17. With the exception of “reliable messaging service” and “provides
transparent persistence,” which we will address below, we discern no reason
to modify or further discuss in this Final Written Decision our constructions
for the remaining claim phrases. For convenience, those claim constructions
are reproduced in the table below.
Claim Phrase Claim Construction
“update the cache only when a user
attempts to access the invalidated
unit or data or object in a user cache,
may automatically update the user
cache as soon as notified as has been
pre-set in the caching system by the
user of the user cache”
“update the cache only when a user
attempts to access the invalidated
unit of data or object in a user cache,
or may automatically update the user
cache as soon as notified as has been
pre-set in the caching system by the
user of the user cache”
“lazy cache”
“a cache in which data or an object
is first placed in the cache when it is
requested.”
“Unit Notification System (UNS)”
“a software component that manages
identifying information of data or
objects in caches and can send
messages to the caches in a
computer system.”

1. “reliable messaging service”
In the Decision to Institute, we construed “reliable messaging service”
to be “a messaging service that sends messages that will be received
IPR2014-00117
Patent 6,985,912 B2

7
according to guarantees made by the service.” Dec. to Inst. 11. Patent
Owner asserts that in order to account fully for the context of the term as
recited in claim 1, element b, we need to construe the phrase “any other
reliable messaging service available in the object programming language
system environment.” PO Resp. 10–11 (emphasis omitted). According to
Patent Owner, a reliable messaging service should be construed as follows:
a service that:
i) provides communication between software components
or applications as a loosely coupled peer-to-peer facility in
which the sender and receiver do not have to be available at the
same time in order to communicate;
ii) sends messages that will be received according to
guarantees made by the service;
iii) has reliability capabilities similar to the JMS
messaging service, namely store and forward capabilities; and
iv) has an interface for object oriented programs to use
the messaging service.
Id. at 12–13 (citing Ex. 2001 ¶ 84).
Patent Owner asserts that a person of ordinary skill in the art would
have understood the recited phrase “any other” to mean that the reliable
messaging service of claim 1 is a reliable messaging service with capabilities
similar to those of the Java Messaging Service (“JMS”). Id. at 11–12 (citing
Ex. 2001 ¶ 83). In support of its position, Patent Owner relies upon the
testimony of Sam Malek, Ph.D. See Ex. 2001 ¶¶ 80–85. Dr. Malek provides
a detailed description of JMS (id. ¶¶ 39–47), however, he does not provide a
sufficient factual basis for his assertion that “any other” would be limited to
services similar to JMS. During his deposition, Dr. Malek testified that JMS
was an example of a messaging service and that there were other messaging
services beyond JMS. Ex. 1016, 48:19–24. He, however, asserts that
IPR2014-00117
Patent 6,985,912 B2

8
because the ’912 Patent mentions JMS, that in itself is an indication that the
claims are directed to a particular form of messaging performed by JMS. Id.
at 50:19–51:7. We note, however, that the ’912 Patent does not limit the
messaging service to JMS; instead it lists JMS as an example of a messaging
service that could be used. Ex. 1001, 5:57–60 (“caching system that utilizes
a messaging system such as the Java Messaging Service (JMS)”), 16:28–30
(“the Java Messaging Service (JMS) or to any other reliable messaging
service”), 16:46–49 (“JMS . . . or other reliable messaging service), 41:22–
25 (“JMS (or similar) notification caching system”). At most, there is one
passing reference in the specification to the messaging system being
“similar” to JMS, but the specification does not describe in what way the
messaging system should be similar to JMS, nor does it state how similar the
messaging system should be to JMS. Thus, we find that the ’912 Patent
does not require JMS or any specific features of JMS. Therefore, we
determine that the broadest reasonable interpretation of “any other” as
recited in claim 1 of the ’912 Patent does not limit the claimed messaging
service to those similar to JMS.
As to the phrase “available in the object programming language
system environment,” Patent Owner argues that this should mean that the
recited messaging service has an interface for object oriented programs to
use the messaging service. PO Resp. 12 (citing Ex. 2001 ¶ 83). In support
of its position, Patent Owner cites Dr. Malek’s testimony. Dr. Malek,
however, provides no analysis or factual support for that assertion outside of
a discussion of JMS’s program language system environment. Ex. 2001 ¶¶
82–83. We determine that this term is not limited to the environment of the
JMS. Therefore, we determine that the broadest reasonable interpretation of
IPR2014-00117
Patent 6,985,912 B2

9
“available in the object programming language system environment” is not
limited to having an interface for object oriented programs to use the
messaging service.
We are not persuaded that express construction is necessary for the
terms “any other” and “available in the object programming language
system environment.” In addition, after reviewing the evidence of record,
we determine that our preliminary construction of “reliable messaging
service” to be “a messaging service that sends messages that will be received
according to guarantees made by the service” is the broadest reasonable
construction of the term and, as such, we do not modify our preliminary
construction.
2. “provides transparent persistence”
In our Decision to Institute, we construed the phrase “provides
transparent persistence” to mean “allows for the storage of data objects that
are not inherently database aware.” Dec. to Inst. 12. The parties do not
dispute the construction itself; instead, the parties dispute the viewpoint used
to evaluate whether transparent persistence has been achieved. PO Resp.
13–17; Reply 8. Patent Owner asserts that transparent persistence should be
evaluated at runtime. PO Resp. 14–17. Petitioner argues that it should be
evaluated from the perspective of the developer of the source code. Reply 8.
Under the Petitioner’s view, transparent persistence may be achieved
if the person writing the source code may create an object without including
persistence code (e.g., SQL statements or JDBC calls), and this remains true
even if a code generator later generates persistence code that may be
embedded in the object. Reply 8. Patent Owner asserts that Petitioner’s
proposal does not go far enough. Patent Owner argues that transparent
IPR2014-00117
Patent 6,985,912 B2

10
persistence requires that the object not have any code in it for implementing
the storage of data in the underlying database. PO Resp. 14. According to
Patent Owner’s view, the existence of persistence code in the object makes
the object “brittle” such that the code can be broken by even minor changes
to the underlying data store. Id. at 4. “Transparent persistence solves these
problems by completely separating the application logic of the object from
its persistence logic. Transparent persistence of objects means that the
objects are not database aware. In other words, the objects do not have any
code that deals with their persistence.” Id. (citing Ex. 2001 ¶ 56).
This is a question of the proper scope of the claim term. Transparent
persistence is not defined expressly in the specification. Therefore, we must
give the term the scope that one of ordinary skill in the art would accord to it
given its ordinary and customary usage. See Paulsen, 30 F.3d at 1480.
Petitioner’s expert, Dr. Antony Hosking, testifies that “transparent
persistence means that an object is not database-aware from the perspective
of the object-oriented application developer that designs the objects.”
Ex. 1022 ¶ 5.5.1.6. Dr. Hosking and Petitioner cite extrinsic references to
support the assertion that one of skill in the art would have understood
transparent persistence to be seen from the viewpoint of the developer. Id.
¶¶ 5.5.1.5–5.5.1.6 (citing Ex. 1019, 1; Ex. 1020, 5 (“By transparent
persistence, we mean that programmers should not need to be aware of . . .
stable storage representations of persistent object.”)); Reply 9 (citing Ex.
1018, 2–3; Ex. 1019, 1). On behalf of Patent Owner, Dr. Malek testifies that
one benefit of transparent persistence is improved programmer productivity,
but that the term, when properly construed, also includes other benefits such
as decoupling the object and its storage code such that they could be
IPR2014-00117
Patent 6,985,912 B2

11
modified independently at runtime without requiring the recompilation of
the source code. Ex. 2001 ¶ 89.
We determine that the ’912 Patent does not limit transparent
persistence to only the situation where persistence code is not present in the
object at all points in time. Petitioner and Petitioner’s expert have provided
persuasive evidence that one of skill in the art would be familiar with
transparent persistence as a tool to ease the development of an object by not
requiring the programmer to be aware of the details of the underlying
database. We are persuaded that, under the broadest reasonable
interpretation of “provides transparent persistence,” the ’912 Patent does not
require all potential benefits of transparent persistence to be achieved in
order to meet the limitations of the claim. Thus, we determine that the claim
limitation may be met if an object is created without persistence code, and
the claim limitation still will be met if a code generator later generates
persistence code and embeds that code in the object.
B. Anticipation by Franklin
Petitioner alleges that claims 1–3 of the ’912 Patent are unpatentable
under 35 U.S.C. § 102 over Franklin. Pet. 11–19. Petitioner relies on claim
charts showing how this reference discloses the claimed subject matter. Id.
Petitioner further relies on a Declaration from Dr. Hosking. Ex. 1014.
For the reasons discussed below, we conclude that Petitioner has
proved by a preponderance of the evidence that claims 1–3 are anticipated
by Franklin.
IPR2014-00117
Patent 6,985,912 B2

12
1. Overview of Franklin (Ex. 1005)
Franklin is a paper titled “Transactional Client-Server Cache
Consistency: Alternatives and Performance.” Ex. 1005. Franklin states that
“[c]lient data caching is a dynamic form of data replication. As with any
form of data replication, care must be taken to ensure that the presence of
multiple copies in the distributed system does not jeopardize the correctness
of programs.” Id. at 2.
Franklin describes several algorithms used to maintain cache
consistency. Id. at 10. One family of algorithms described in Franklin is the
Optimistic Two-Phase Locking (“O2PL”) family. Id. at 23. In O2PL, each
client has a local lock manager from which a lock must be obtained before a
transaction can access a client’s data. Id. After a page is updated in the
client’s cache, the update is sent from the client to the server. Id. The server
then locks the pages to be updated and sends a message to each client with a
cached copy of the page to be updated. Id. The clients then lock the local
cached copy of the page. Id. O2PL-I is a variant of the O2PL family that
always invalidates cached copies of the updated page. Under O2PL-I, the
client deletes the cached copy and releases the lock. Id. O2PL-P is a variant
that always propagates updated page copies to clients that are caching the
updated pages. Id. Under O2PL-P, the client locks the cached page and
sends a message to the server indicating it is prepared to commit the update.
Id. at 23–24. Then, the server sends copies of the updated pages to the
clients that have notified the server that they are prepared to commit the
update. Id. at 24. The clients then update the cached page and release the
lock. Id.
IPR2014-00117
Patent 6,985,912 B2

13
Figure 4 of Franklin is reproduced below.

Figure 4 illustrates a performance model of a client-server database
management system. Id. at 26. The clients and server in Figure 4 each have
a Concurrency Control Manager. Id. at Fig. 4. The server’s Concurrency
Control Manager stores information about the location of page copies in the
system and manages locks. Id. at 26.
2. Analysis
Petitioner contends that Franklin anticipates claims 1–3. Pet. 11–19.
We have reviewed Petitioner’s detailed explanation identifying where each
limitation allegedly appears in Franklin, along with the testimony of
Petitioner’s declarant, Dr. Hosking. See id.; Ex. 1014 ¶¶ 9.2.1–9.2.1.1.7.2;
Ex. 1022 ¶¶ 5.1–5.2.2.2. We also have reviewed Patent Owner’s assertions
and evidence as to why Petitioner’s explanations and evidence are deficient.
PO Resp. 18–34; Ex. 2001 ¶¶ 107–132. On this record, we are persuaded
that Petitioner has shown by a preponderance of the evidence that Franklin
discloses each limitation of claims 1–3.
Claim 1 requires placing data or objects in a cache. Petitioner argues
that Franklin discloses the claimed cache. Pet. 12–13; Ex. 1014
¶ 9.2.1.1.2.2. Element (b) of claim 1 further requires that the data be
IPR2014-00117
Patent 6,985,912 B2

14
registered with a unit notification system (“UNS”) and that the UNS be
connected to JMS “or to any other reliable messaging service available in
the object programming language system environment.” Petitioner asserts
that Franklin’s Concurrency Control Manager is the claimed UNS. Pet. 13;
Ex. 1014 ¶ 9.2.1.1.3.2. Petitioner argues that a reliable messaging service is
disclosed inherently as part of the O2PL algorithms described in Franklin.
Ex. 1014 ¶ 9.2.1.1.3.4. Element (c) of claim 1 requires that when any data
or object is updated, the UNS sends a message to all other caches that may
contain the updated data alerting them that the data has been updated and is
invalid. Petitioner argues that in the O2PL algorithms discussed in Franklin,
once a page has been updated, the server sends a message to each client that
has cached copies of any of the updated pages to inform those clients that
the data has been invalidated. Ex. 1005, 23. Finally, claim 1 requires the
cache to either (1) update the data when a user attempts to access the
invalidated data or (2) update the data automatically. Petitioner argues that
both options are disclosed in Franklin via its discussion of the O2PL-I and
O2PL-P variants of the O2PL algorithm family. Specially, Petitioner argues
that Franklin discloses the O2PL-I algorithm in which a client purges a page
from its cache upon receipt of an invalidation message from the server.
Pet. 16. Dr. Hosking opines that this discloses the claimed updating of the
cache when a user attempts to access the invalidated data because “the next
request for the now purged page will be a miss, and the client will request a
new copy of the page and add the new copy to the cache.” Ex. 1014
¶ 9.2.1.1.5.3. Petitioner notes that the OP2L-P algorithm automatically
updates the client cache in response to an invalidation message from the
server. Id. ¶ 9.2.1.1.5.4; Pet. 15; Ex. 1005, 23–24.
IPR2014-00117
Patent 6,985,912 B2

15
Claim 2 depends from claim 1 and further requires the applications or
users to access the data in a virtual machine environment. Petitioner directs
us to Franklin’s disclosure of examples of object-oriented database
management systems (“OODBMS”) including Gemstone. Pet. 17 (citing
Ex. 1005, 2). Dr. Hosking explains5 that “the example OODBMS described
in Franklin include OODBMS, such as Gemstone, that specifically run in
object-oriented virtual machine environments, including both Java and
Smalltalk virtual machine environments.” Ex. 1014 ¶ 9.2.1.1.6.1. Claim 3
depends from claim 2 and further requires a distributed system with a central
data source. Petitioner argues that Franklin’s Figure 4 depicts a distributed
client-server database management system that discloses the limitations of
claim 3. Pet. 17 (citing Ex. 1005, 25, Fig. 4).
In response, Patent Owner contends that Franklin does not disclose
elements (b) and (c) of claim 1. We address each of Patent Owner’s
arguments in turn.
i. “any other reliable messaging service available in the object
programming language system environment”
Much of Patent Owner’s argument on this element centers on its
proposed construction of reliable messaging service. As noted in § II.A.1 of
this Decision, we do not agree with Patent Owner’s proposal and we
determine that the broadest reasonable construction of this term is “a

5 Extrinsic evidence, such as Dr. Hosking’s Declaration, may be used to
explain, but not expand on, what one of ordinary skill in the art would have
understood from reading the purportedly anticipatory reference. See In re
Baxter Travenol Labs, 952 F.2d 388, 390–91 (Fed. Cir. 1991) (concluding
that the Board properly considered an expert declaration that explained the
makeup of a prior art product).
IPR2014-00117
Patent 6,985,912 B2

16
messaging service that sends messages that will be received according to
guarantees made by the service.” With that in mind, Patent Owner’s
remaining arguments are that Petitioner has not provided sufficient evidence
to show that Franklin inherently discloses a reliable messaging service. PO
Resp. 20. In addition, Patent Owner contends that the two-step commit
disclosed in Franklin can be performed using an unreliable messaging
service. Id. at 22.
As to the inherent disclosure of a reliable messaging service,
Petitioner’s expert Dr. Hosking opines that
it is inherent in the disclosure of Franklin, that the invalidation
messages are sent using a reliable messaging service. . . . In the
context of Franklin, messages are sent to caches to invalidate
copies of data stored in those caches as part of a cache
consistency algorithm. Invalidation messages sent through any
other messaging service other than a reliable messaging service
would result in missed invalidation messages and inconsistent
caches, rendering the system describe[d] by Franklin
ineffective.
Ex. 1014 ¶ 9.2.1.1.3.4. We find Dr. Hosking’s testimony to be well
reasoned. “‘[A]nticipation by inherent disclosure is appropriate only when
the reference discloses prior art that must necessarily include the unstated
limitation.’” Rexnord Indus., LLC v. Kappos, 705 F.3d 1347, 1355 (Fed.
Cir. 2013) (quoting In re Omeprazole Patent Litigation, 483 F.3d 1364,
1378 (Fed. Cir. 2007)). According to our construction of reliable messaging
service, such a service makes guarantees regarding the receipt of messages.
Therefore, we find that one of ordinary skill in the art would understand an
unreliable messaging service to be one that makes no such guarantees. It
follows that if an unreliable messaging service is used, then there is no
guarantee that invalidation messages would be received. The stated purpose
IPR2014-00117
Patent 6,985,912 B2

17
of the algorithms described in Franklin is “to ensure that such caching does
not result in the violation of transaction semantics.” Ex. 1005, 1. Franklin
goes on to note that “care must be taken to ensure that the presence of
multiple copies [of data] in the distributed system does not jeopardize the
correctness of programs.” Id. at 2. The absences of guarantees regarding
the receipt of invalidation messages would render the disclosed algorithms
ineffective for their stated purpose because one cannot ensure that messages
are received if an unreliable messaging system is used. Thus, we find that
one of ordinary skill in the art would read Franklin’s disclosure to require a
reliable messaging service.
We also find Patent Owner’s argument regarding the two-phase
commit to be unpersuasive. Patent Owner asserts that “the evidence shows
that messages sent in a two-phase commit process are not required to be sent
with a reliable messaging service for the commit process to function.” PO
Resp. 22. According to Patent Owner, “one of the purposes of a two-phase
commit process is [to] perform database transactions over a distributed
network where there is no guarantee that a message will be received at its
destination.” Id. at 23 (citing Ex. 2001 ¶ 112). Dr. Malek supports Patent
Owner’s assertion by testifying that “a two-phase commit process has a
mechanism for handling failures in the protocol, namely time-out
mechanisms.” Ex. 2001 ¶ 112 (citing Ex. 2006, 356 (noting that timeout
mechanisms are used in cases where a process may be waiting indefinitely
for a message to be received)). Petitioner argues that only one of the
disclosed O2PL algorithms (O2PL-P) could be performed with a two-phase
commit. Tr. 10:21–25. As described in Franklin, the O2PL-I algorithm
requires “the client [to] simply purges its copies of the updated pages,
IPR2014-00117
Patent 6,985,912 B2

18
releases its local locks on those pages, and then sends an acknowledgement
message to the server, a two-phase commit protocol is not necessary in this
case.” Ex. 1005, 23 (emphasis added). In addition, Petitioner argues that by
including a time out mechanism, the messaging system becomes a reliable
messaging system, i.e., one that makes guarantees. Reply 3; Tr. 11:5–17,
12:11–13:2. Petitioner argues that one of skill in the art would have
understood the timeout mechanism to be a part of the messaging service
“and specifically the part of that service that provides reliability to the
overall messaging service.” Tr. 12:24–13:2. On cross examination, Patent
Owner’s expert testified that “a two-phase commit process is generally a
technique that it can use to achieve reliability . . . because of the fact that it
has [a] time out mechanism in it.” Ex. 1016, 114:21–24. Thus, we
determine that one of ordinary skill in the art would have understood that
messaging performed via a two-phase commit with a time out mechanism to
be a reliable messaging system. We are persuaded that O2PL-P could not be
performed with an unreliable messaging system. Therefore, we determine
that despite the possibility of using a two-phase commit protocol, Franklin’s
disclosed O2PL algorithms necessarily require a reliable messaging system.
ii. “when any unit of data or object is updated . . . and
persisted (saved) to a primary data source the UNS sends a
message . . . that the unit was updated . . . and is noted as
being invalid in the cache”
Patent Owner argues Franklin does not disclose the sequence recited
in element (c) of claim 1. PO Resp. 25–33. We note, however, that “as a
general rule the claim is not limited to performance of the steps in the order
recited, unless the claim explicitly or implicitly requires a specific order.”
Baldwin Graphic Sys., Inc. v. Siebert, Inc., 512 F.3d 1338, 1345 (Fed. Cir.
IPR2014-00117
Patent 6,985,912 B2

19
2008) (citing Interactive Gift Express, Inc. v. Compuserve Inc., 256 F.3d
1323, 1342–43 (Fed. Cir. 2001)). This is particularly true here because
claim 1 recites a computer system and not a method.
According to Patent Owner, the inclusion of the term “when” means
that the invalidation message cannot be sent to the clients before the update
is persisted in the data source. PO Resp. 29. Patent Owner urges that the
term “when” has an ordinary and customary meaning of “at or during which
time.” Id. at 29 n.5 (quoting Ex. 2003, 1424). We do not agree with Patent
Owner’s narrow interpretation. Franklin discloses the sending of
invalidation messages and the commitment to the database to be
contemporaneous activities. Ex. 1005, 23–24. Franklin describes that in the
O2PL algorithms a “transaction enters its commit phase” then it sends
invalidation messages to the server and the client regarding the data to be
updated, and then the data is updated in accordance with the algorithm. Id.
Thus, one of ordinary skill in the art would have understood that the sending
of invalidation messages to be part of the “commit phase.” As such, the
sending of invalidation messages would occur “when” the update is being
committed. Therefore, we determine that this disclosure meets the
limitations of element (c) of claim 1.
Patent Owner also asserts that Franklin does not disclose a message
that notes that the data in the cache is invalid. PO Resp. 33–34.
Specifically, Patent Owner argues that the O2PL-P merely propagates the
pages without a message that the page is invalid, and the O2PL-I algorithm
merely deletes the page without a message that the page is invalid. Id. As
disclosed in Franklin, “the server sends a message to each client that has
cached copies of any of the updated pages.” Ex. 1005, 23. The clients then
IPR2014-00117
Patent 6,985,912 B2

20
obtain locks for those pages. Id. It is only after the locks are obtained that
any page is purged or updated. Id. Thus, the message sent to the client is a
message that alerts the client that the page in question has been invalidated,
and this message is received prior to any action being done to the page. As
such, we determine that the message described by Franklin would have been
understood by one of ordinary skill in the art to be a message that invalidates
the data stored in the caches containing copies of the data that has been
updated. Therefore, we find Patent Owner’s argument to be unpersuasive.
After considering the arguments and evidence presented by Petitioner
and Patent Owner, we determine that Petitioner has shown by a
preponderance of the evidence that Franklin anticipates claims 1–3.
C. Obviousness over Franklin and JVM
Petitioner alleges that claim 2 of the ’912 Patent is unpatentable under
35 U.S.C. § 103 over Franklin and JVM. Pet. 25–26. For the reasons
discussed below, we conclude that Petitioner has proved by a preponderance
of the evidence that claim 2 would have been obvious over Franklin and
JVM.
Claim 2 depends from claim 1 and further recites that in the system of
claim 1, multiple applications or users access data in an object language
virtual machine environment. JVM is the specification for the Java Virtual
Machine. Ex. 1009. According to Petitioner, JVM was published
approximately five years before the priority date of the ’912 Patent. Pet. 25;
Ex. 1014 ¶ 8.5.2.2. The reference describes the requirements and
capabilities of the Java Virtual Machine. Pet. 25 (citing Ex. 1009, 2).
Petitioner asserts that to the extent that we find Franklin to not disclose the
IPR2014-00117
Patent 6,985,912 B2

21
“object language virtual machine environment,” it would have been obvious
for one of ordinary skill in the art to implement Franklin’s caching system in
the virtual machine environment disclosed by JVM. Id. at 25. As discussed
above, we find that Franklin discloses the claimed “object language virtual
machine environment.” We note, however, that “obviousness under § 103
can be based on a reference which happens to anticipate the claimed subject
matter.” In re Meyer, 599 F.2d 1026, 1031 (CCPA 1979); see also Cohesive
Techs, Inc. v. Waters Corp., 543 F.3d 1351, 1364 (Fed. Cir. 2008) (“it is
commonly understood that prior art references that anticipate a claim will
usually render that claim obvious”). Thus, we analyze the disclosures of
JVM to ascertain whether it provides additional teachings that also would
have rendered claim 2 obvious in light of the disclosures of Franklin.
JVM discloses that “[t]he Java Virtual Machine is an abstract
computing machine. Like a real computing machine, it has an instruction set
and uses various memory areas. It is reasonably common to implement a
programming language using a virtual machine . . . .” Id. at 26 (quoting
Ex. 1009, 2). Dr. Hosking opines that “[i]t would have been obvious to one
of ordinary skill in the art to combine the disclosure of JVM with the
disclosure[] of Franklin . . . to implement the caching systems described
therein for an object oriented virtual machine environment.” Ex. 1014
¶ 8.5.2.3. Petitioner asserts that “environments, such as JVM, were widely
known and used in distributed computer systems including caching
systems.” Pet. 25–26 (citing Ex. 1014 ¶ 8.5.2.3).
Patent Owner asserts that Petitioner improperly relied upon hindsight
reasoning to combine JVM and Franklin. PO Resp. 38. We are not
persuaded by Patent Owner’s assertion. We, instead, are persuaded that the
IPR2014-00117
Patent 6,985,912 B2

22
Java Virtual Machine was well-known, and that the implementation of a
caching system, such as the one described by Franklin, in the Java Virtual
Machine environment would have been within the knowledge of one of
ordinary skill, and that such modification would have been motivated by the
knowledge of one of ordinary skill. See In re McLaughlin, 443 F.2d 1392,
1395 (CCPA 1971) (“Any judgment on obviousness is . . . necessarily a
reconstruction based upon hindsight reasoning, but so long as it takes into
account only knowledge which was within the level of ordinary skill at the
time the claimed invention was made and does not include knowledge
gleaned only from applicant’s disclosure, such a reconstruction is proper.”).
Therefore, we are not persuaded of improper hindsight reasoning.
After considering the arguments and evidence presented by Petitioner
and Patent Owner, we determine that Petitioner has shown by a
preponderance of the evidence that Franklin and JVM would have rendered
obvious claim 2.
D. Obviousness over Franklin and Chang
Petitioner alleges that claims 6 and 7 of the ’912 Patent are
unpatentable under 35 U.S.C. § 103 over Franklin and Chang. Pet. 26–28.
For the reasons discussed below, we conclude that Petitioner has proved by a
preponderance of the evidence that claims 6 and 7 would have been obvious
over Franklin and Chang.
1. Analysis of Prior Art Challenge
Chang is a patent directed to a system and method for providing a
graphical user interface for mapping and accessing objects in a data store.
Ex. 1010, [54]. Chang discloses that “[a] user may define a mapping
IPR2014-00117
Patent 6,985,912 B2

23
between object schema and data store schema by use of a high level
language, Schema Mapping Definition Language (SMDL), which is data
store independent, object oriented language independent, and extensible.”
Id. at Abstract. Chang discusses the differences in structure between
objected oriented and relational databases and describes a method and
apparatus for allowing a user to access a relational database from an object
oriented application. Id. at 1:48–4:43. In Chang’s disclosure, “a definition
of a mapping between object schema and conventional data store schema is
represented by a Schema Mapping Internal Representation [(SMIR)].” Id. at
5:30–32. As described in Chang, “data store independence is provided by
the use of the data store independent [SMDL], the data store independent
[SMIR], and the single, uniform, data store independent interface to a Data
Store Manager.” Id. at 5:66–6:4.
Claim 6 depends from claim 1 and further recites that the system of
claim 1 “also provides transparent persistence.” Petitioner asserts that the
mapping and accessing of data discussed in Chang teaches transparent
persistence. Pet. 26–27 (citing Ex. 1014 ¶ 9.3.4.1.1.3). Claim 7 depends
from claim 1 and further requires that the system comprise at least one
Object to Relational Mapping Repository. Petitioner argues that Chang’s
Data Store Manager contains a SMIR that defines the mapping between and
the object oriented and relational databases. Ex. 1010, 5:30–57; see Pet. 28;
Ex. 1014 ¶¶ 9.3.4.1.4.2–9.3.4.1.4.3.
As to claim 6, Patent Owner argues that Chang does not teach storing
objects that are not inherently database aware. PO Resp. 39. Patent
Owner’s argument is premised on its proposal regarding how and when to
evaluate whether an object provides transparent persistence. As discussed in
IPR2014-00117
Patent 6,985,912 B2

24
§ II.A.2, we do not agree with Patent Owner’s proposal. In that discussion,
we determined that the claim limitation may be met if an object is created
without persistence code. Patent Owner directs us to Chang’s disclosure of
code generators 410 and 420 that later insert data access code within objects
enabling those objects to manipulate data in the underlying data store. PO
Resp. 43 (citing Ex. 1010, 37:9–57). As discussed above, however, we
determined that one aspect of transparent persistence may be achieved if an
object may be created in a manner such that the programmer does not need
to be aware of (and therefore, does not write source code to access) the
underlying data store. Thus, under the broadest reasonable interpretation of
the term, we find that one of ordinary skill in the art would have understood
Chang’s disclosure of objects that have persistence code added at a later
point in time by a code generator to be a system that provides transparent
persistence. See Ex. 1022 ¶ 5.5.1.5.
Patent Owner also contends that Petitioner’s rationale for combining
Franklin and Chang is conclusory and rooted in hindsight. PO Resp. 50.
Petitioner asserts that transparent persistence was a well-known concept at
the time of the invention of the ’912 Patent and that it would have been
obvious to add transparent persistence to the disclosures of Franklin. Reply
13. Petitioner asserts that adding transparent persistence, as taught by
Chang, to the caching system of Franklin only requires “applying a known
technique (e.g., transparent persistence) to a known system as disclosed in
Franklin and Chang (e.g., Franklin’s system as modified by Chang and the
knowledge of a skilled artisan) to yield predictable results (providing
transparent persistence in the system).” Id. (citing Ex. 1022 ¶ 5.5.3.4); see
Ex. 1014 ¶ 8.6.2.2. We are persuaded by Petitioner’s argument that
IPR2014-00117
Patent 6,985,912 B2

25
transparent persistence was well-known, that the addition of such to a
system such as the one described by Franklin would have been within the
knowledge of one of ordinary skill, and that such modification would have
been motivated by the knowledge of one of ordinary skill. McLaughlin, 443
F.2d at 1395.
2. Analysis of Secondary Considerations
Patent Owner asserts that evidence of secondary considerations
outweighs Petitioner’s prior art evidence of obviousness of claim 6. PO
Resp. 52–54. Specifically, Patent Owner directs us to its CocoBase product,
which it alleges to embody the features of claim 6. Id. at 52. Patent Owner
asserts that CocoBase received widespread industry recognition and awards,
such as the 2003 WebSphere Advisor Editors’ Choice Award in the field of
Data Integration. Id. at 53 (citing Ex. 2004, 4). In addition, in January
2003, About.com named CocoBase the “Best of the Net.” Id. at 54 (citing
Ex. 2005, 1).
We are not persuaded that Patent Owner’s evidence of industry
acclaim is sufficient evidence of indicia of nonobviousness. As to the
About.com article, there is only a brief mention of CocoBase and very little
discussion as to why this application is praiseworthy. The entire discussion
of CocoBase is as follows: “CocoBase is the leading object to relational
database mapping tool for the Java platforms. It requires no EJB, JDBC, or
SQL coding.” Ex. 2005, 1. The WebSphere article also is very short, and it
provides little detail as to why CocoBase is praiseworthy. Ex. 2004, 4.
Patent Owner provides no evidence by which we may evaluate whether
these web pages should be considered to be industry praise. There is no
evidence of record as to how these awards were given. Nor is there evidence
IPR2014-00117
Patent 6,985,912 B2

26
as to whether these awards were well-known or respected in the industry.
We are not persuaded that these two mentions on web pages constitute
industry praise. See Bayer Healthcare Pharms., Inc. v. Watson Pharms.,
Inc., 713 F.3d 1369, 1377 (Fed. Cir. 2013) (finding that brief discussions of
Patent Owner’s product in journal articles “fall well short of demonstrating
true industry praise”). In addition, Patent Owner only has made a
conclusory statement that CocoBase embodies the limitations of claim 6 and
has not presented evidence of a nexus between CocoBase and the limitations
of claim 6. Therefore, we are not persuaded that Patent Owner’s evidence of
secondary considerations outweighs Petitioner’s evidence of obviousness.
We determine that Petitioner has demonstrated by a preponderance of
the evidence that claims 6 and 7 are unpatentable over Franklin and Chang.
III. CONCLUSION
We conclude Petitioner has demonstrated by a preponderance of the
evidence that
1. Claims 1–3 are unpatentable under 35 U.S.C. § 102 as anticipated by
Franklin;
2. Claim 2 is unpatentable under 35 U.S.C. § 103 as obvious over Franklin
and JVM; and
3. Claim 6 and 7 are unpatentable under 35 U.S.C. § 103 as obvious over
Franklin and Chang.
IV. ORDER
In consideration of the foregoing, it is hereby
ORDERED that, based on a preponderance of the evidence, claims 1–
3, 6, and 7 of the ’912 Patent are unpatentable;
IPR2014-00117
Patent 6,985,912 B2

27
FURTHER ORDERED that, because this is a Final Written Decision,
parties to the proceeding seeking judicial review of the Decision must
comply with the notice and service requirements of 37 C.F.R. § 90.2.

IPR2014-00117
Patent 6,985,912 B2

28
PETITIONER:
Vaibhav P. Kadaba
Steven D. Moore
Kilpatrick Townsend & Stockton LLP
wkadaba@kilpatricktownsend.com
smoore@kilpatricktownsend.com

PATENT OWNER:
J. Warren Lytle, Jr.
Brian K. Shelton
SUGHRUE MION PLLC
jlytle@sughrue.com
bshelton@sughrue.com