Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 1 of 10 PageID #: 1576
`
`IN THE UNITED STATES DISTRICT COURT
`FOR THE EASTERN DISTRICT OF TEXAS
`SHERMAN DIVISION
`
`
`
`
`
`
`
`
`
`
`Civil Action No. 4:23-cv-01147-ALM
`
`Jury Trial Demanded
`
`
`
`
`R2 Solutions LLC,
`
` Plaintiff,
`
`v.
`
`Databricks, Inc.,
`
` Defendant.
`
`
`
`JOINT MOTION FOR ENTRY OF DISPUTED PROTECTIVE ORDER
`
`Plaintiff R2 Solutions LLC (“R2”) and Defendant Databricks, Inc. (“Databricks) hereby
`
`request that the Court resolve a disputed issue as to the protective order to be entered in this case.
`
`The parties have met and conferred and have agreed to all aspects of a protective order except for
`
`the provision regarding what tools shall be available on the source code review computer. The
`
`parties’ respective positions are as follows:
`
`Plaintiff’s Position
`
`The crux of the parties’ dispute is whether the protective order should forbid the installation
`
`of tools that can compile and execute source code (i.e., turn the code into a form that a computer can
`
`run). Databricks believes that such tools should be forbidden. R2 believes that such a restriction is
`
`inappropriate for the reasons set forth below.
`
`The parties have abided by the default protective order in this case since its entry on April 9,
`
`2024. See ECF 18. On July 16, Databricks served its invalidity contentions and made a document
`
`production. Upon inspection, R2 discovered that, with respect to source code, the production
`
`included only replicas of open-source code repositories. It included none of Databricks’ own source
`
`code for the accused instrumentalities, and Databricks did not otherwise make its own source code
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 2 of 10 PageID #: 1577
`
`available for inspection (and still has not done so).
`
`R2 raised this issue with Databricks on August 6 and requested that Databricks’ source code
`
`be made available for inspection. On August 16, over four months after the entry of the default
`
`protective order, Databricks claimed that it would need to amend the protective order before
`
`producing its source code. On August 19, R2 sent Databricks a draft protective order which draws
`
`heavily from Judge Gilstrap’s typical order and mimics the protective order that this Court entered
`
`in R2 Solutions LLC v. American Airlines, Inc., which also involved the patent-in-suit. See No. 4:22-
`
`cv-00353-ALM, ECF 22 (Oct. 17, 2022). After prodding from R2, Databricks finally returned a
`
`redlined version of the draft protective order. Databricks’ redline was essentially a re-write, going
`
`so far as changing the definitions of the various confidentiality designations, imposing a page limit
`
`on source code print outs, unreasonably limiting contiguous lines of source code that may appear in
`
`briefing, and expanding the reach of the prosecution bar, among much more.
`
`Nevertheless, in an effort to resolve these disputes amicably, R2 engaged in a protracted
`
`back-and-forth with Databricks attempting to reach a workable compromise version. The parties
`
`have now been able to agree to all terms of the Protective Order, except for the terms in paragraph
`
`12(g) regarding source code review tools. R2’s proposal is attached as Exhibit A, and Databricks’
`
`proposal is attached as Exhibit B.
`
`Databricks proposes language that would preclude R2 from implementing software review
`
`tools “capable of compiling or executing the code.” See Ex. B at 8. R2 opposes inclusion of such
`
`language. Databricks bears the burden to show good cause as to why such a limitation on discovery
`
`should be imposed. See Saxon Innovations v. Nokia Corp., No. 6:07-CV-490, 2008 U.S. Dist. LEXIS
`
`138985, at *18 (E.D. Tex. Aug. 18, 2008). But it has not, and cannot, demonstrate such good cause,
`
`as case law in this District makes clear. See id. at *24-26 (finding that defendants failed to establish
`
`good cause and that “Plaintiff [could] use compilers to analyze Defendants’ source code.”).
`2
`
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 3 of 10 PageID #: 1578
`
`Precluding review tools that allow compiling and executing code is arbitrary and
`
`unreasonable. As R2 has explained to Databricks multiple times, compiling and executing the code
`
`could greatly expedite R2’s review efforts. Big data analytics platforms like Databricks’ are often
`
`gargantuan, and if R2 could compile and execute the code as part of its review efforts, R2 can process
`
`mock datasets on the review computer with the accused instrumentalities to fast-track identification
`
`of the most relevant portions of the code. This could save hours and thousands of dollars. And
`
`allowing R2 to compile and execute the code would not be burdensome for Databricks—to the
`
`contrary, Databricks would simply have to install one or two more tools on the stand-alone computer
`
`for R2 to accomplish its review.
`
`Databricks has provided three reasons for limiting R2’s discovery in this fashion: (1) such
`
`provisions are “standard” for Databricks; (2) there is purportedly a security risk in allowing R2 to
`
`compile and execute the code; and (3) the code may not compile because (i) only portions of the
`
`source code will be made available, and (ii) the one-computer-setup for the source code review may
`
`prevent the code (which Databricks says requires multiple servers) from compiling. As to (1),
`
`whether a provision is “standard” for Databricks is irrelevant—the question is whether Databricks
`
`has shown good cause to limit R2’s discovery efforts. Regarding (2), there is no security risk—any
`
`compiled/executed code would be confined to the stand-alone, non-networked computer just like all
`
`of the un-compiled source code. And as to (3), whether R2 is able to compile the code is R2’s
`
`problem and does not burden Databricks in any way. See Saxon Innovations, 2008 U.S. Dist. LEXIS
`
`138985, at *18 (“[Defendants] claim that compilers will be ineffective in this case because only
`
`complete source code files can be compiled and Plaintiff will only be inspecting portions of the
`
`source code. This unsupported assertion may turn out to be correct, but at this early stage in the
`
`litigation the Court sees no reason to prohibit compilers, especially since any time and money wasted
`
`
`
`3
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 4 of 10 PageID #: 1579
`
`on their use will be at Plaintiff's expense.”).1
`
`Rather than allow the code to be compiled and executed, Databricks has offered to allow R2
`
`access to a customer-facing version of its platform. This is a wholly inadequate solution—indeed,
`
`providing the customer-facing portion of Databricks platform divorced from the executable version
`
`of the underlying source code is not a workable compromise and does little to assist R2 in its code
`
`review efforts. As explained above, having the compiled/executable version of the code would allow
`
`R2’s expert to track the processing of mock data sets through the code to more quickly identify the
`
`modules, files, and functions germane to the relevant data processing steps. Simply having a
`
`customer facing interface may allow R2 to make queries and even generate results, but it would not
`
`allow R2 to ascertain which portions of the code are involved with accomplishing the relevant
`
`functionality.
`
`At bottom, Databricks is using the protective order issue as a ruse to avoid its discovery
`
`obligations and run out the clock. Databricks was obligated to produce its source code almost three
`
`months ago and still has not done so, and Databricks only raised the protective order issue after R2
`
`insisted that Databricks abide by the rules. Databricks’ gamesmanship has already caused immense
`
`delay. The Court should enter R2’s proposed protective order (attached as Exhibit A).
`
`Defendant’s Position
`
`The present dispute centers around R2’s request to obtain not only source code in an
`
`electronic native format and tools to review that code, but also software tools that it can use to
`
`compile and execute Databricks’ source code.2 R2’s request should be rejected. Databricks’
`
`
`1 To be clear, R2 may additionally (or in the alternative) insist that Databricks make an
`executable copy of the source code available for review. This issue, however, is premature, as
`Databricks has not yet made its source code available for inspection such that R2 can determine the
`sufficiency of the production.
`2 R2 also seeks a complete “executable copy of the source code,” although notes that this
`issue “is premature.” (See supra, fn 1.)
`
`
`4
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 5 of 10 PageID #: 1580
`
`agreement to produce code in electronic native format allow R2 to use tools for reviewing that code
`
`is sufficient for discovery, and its request for execution tools not only goes beyond what is required,
`
`but it would also be futile, given the inability to use such tools with the “portions of the source code”
`
`that R2 acknowledges will be available (see supra, at 3). Databricks thus respectfully submits that
`
`the Court deny R2’s requests for tools to execute Databricks’ highly sensitive source code.3
`
`Courts, including in this District, have found that source code in native electronic format is
`
`sufficient for purposes of discovery, and there is no requirement for a producing party to provide an
`
`environment that would ease the requesting party’s evaluation of the code. See Kelora Sys., LLC v.
`
`Target Corp., No. C 10-04947 CW LB, 2011 WL 6000759, at *3 (N.D. Cal. Aug. 29, 2011)
`
`(rejecting request for a functional build environment for the source code, finding that the “benefits
`
`of making it easier for [plaintiff] to evaluate the Accused Infringers’ source code do not outweigh
`
`the burdens associated with the proposal”); Philips N. Am. LLC v. Glob. Med. Imaging, LLC, 343
`
`F.R.D. 59, 66-67 (N.D. Ill. 2022) (rejecting request to allow the storage of “scripts and data output
`
`by those scripts or other [unidentified software] tools” on the source code computer, finding that the
`
`“convenience of [the proponent’s] expert in speed of review” did not outweigh the possibility of
`
`“create[ing] security concerns for extremely valuable source code”); see also Gree, Inc. v. Supercell
`
`Oy. No. 2:19-cv-00311-JRG-RSP, Dkt. 134 at 1–3 (E.D. Tex. Dec. 17, 2020) (rejecting request for
`
`producing party to “build[] a game environment” for the executable file to run correctly, finding that
`
`it is “unduly burdensome, especially when [plaintiff] has access to the source code and has had access
`
`
`3 Databricks does not concede that this Court is the appropriate venue for this case, or that it
`should adjudicate the present dispute. In re Apple Inc., 979 F.3d 1332, 1338–39 (Fed. Cir. 2020)
`(explaining that “once a party files a transfer motion, disposing of that motion should unquestionably
`take top priority”). Respectfully, Databricks submits that pursuant to its Motion to Transfer (see Dkt
`20), the Northern District of California (“the NDCA”) is the appropriate forum for case, and the
`present dispute regarding Databricks’ source code—which, pursuant to the parties’ agreed protocol,
`will be loaded onto a computer in Northern California, then made available to R2 in Northern
`California—only confirms that the NDCA is the more convenient venue.
`5
`
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 6 of 10 PageID #: 1581
`
`to the functionality itself for nearly two years”).
`
`Consistent with the cases cited above, this Court should reject R2’s requests for tools to
`
`execute the code and “process mock datasets” (see supra, at 3). R2’s only basis for its request is that
`
`these tools would purportedly help it “fast-track identification of the most relevant portions of the
`
`code” and “could save hours and thousands of dollars” (see supra, at 3). But R2 provides no support
`
`or any meaningful explanation as to how it would purportedly use the tools for “fast-track
`
`identification.” Indeed, despite Databricks’ requests for R2 to identify the tools it seeks to use, R2
`
`has not done so. Notably, R2 does not contend that such execution tools are necessary for its source
`
`code review, because they are not. In other words, R2 is requesting these unidentified tools “merely
`
`for the convenience” of its review. This request should be rejected. Kelora Sys., LLC, 2011 WL
`
`6000759, at *3 (noting that discovery rules do not require having the producing party undertake
`
`burdensome measures “merely for the convenience of the requesting party.”).
`
`R2’s request should also be rejected as futile. The protective order does not include any
`
`provisions under which R2 would be permitted to load “mock datasets” for processing onto the non-
`
`networked review computer. Rather, the protective order only provides that R2 may request that
`
`“software tools for viewing and searching Source Code Material” be loaded onto the computer and
`
`that Databricks may load such software, provided certain conditions are met. (Exhibit A at ¶ 12(g);
`
`Exhibit B at ¶ 12(g)) And, as R2 itself acknowledges, “only portions of the source code”—i.e., the
`
`portions relating to the elements of aspects that are accused of infringing—“will be made available.”
`
`(See supra, at 3.) Yet executing the code is not possible without an incredibly voluminous amount
`
`of source code, which would span features, functionalities and other components of the Databricks
`
`platform that are completely irrelevant to any issue in this litigation. Because R2’s requested
`
`execution tools cannot be used without a voluminous amount of irrelevant code, and because any
`
`such request for a voluminous amount of irrelevant code would improperly cause Databricks undue
`6
`
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 7 of 10 PageID #: 1582
`
`burden that far outweighs any legitimate benefit to R2, this Court should deny R2’s request for
`
`execution tools.
`
`Further, R2’s attempts to summarily dismiss Databricks’ objections should likewise be
`
`rejected. R2 argues that its proposal poses no security risk because “any compiled/executed code
`
`would be confined to the stand-alone, non-networked computer.” Again, however, executing the
`
`code in the manner in which it operates in the ordinary course of business would require the computer
`
`to be connected to a larger environment, thereby creating a security risk. To the extent R2 seeks to
`
`generate evidence from executing the code in a non-networked environment, such an attempt would
`
`be improper, as such purported evidence would not reflect how the code actually operates in the
`
`ordinary course of business. Databricks has offered, in an effort to reconcile this dispute, to provide
`
`R2 with access to Databricks’ customer environment for the accused instrumentalities. R2 also
`
`already has access to, and can readily compile and execute, all code for Apache Spark (an open
`
`source project)—which is the only functionality that R2 has accused as infringing. Such access,
`
`combined with Databricks’ source code in native format, is more than sufficient for R2 to investigate
`
`its infringement theories.
`
`Finally, Databricks notes that R2 makes a number of unfounded accusations and statements
`
`in this Motion. (See supra, at 1–2.) While they are unrelated to the present dispute and this Court
`
`need not resolve them, Databricks takes this opportunity to correct the record. First, R2 suggests
`
`that Databricks has not complied with its obligation to produce source code under LCR 3-4(a). That
`
`is incorrect. On July 16, 2024, Databricks produced source code “sufficient to show the operation
`
`of any aspects or elements of” Apache Spark—again, the only accused instrumentality identified in
`
`R2’s initial infringement contentions. Second, R2 appears to take issue with the fact that Databricks
`
`proposed revisions to R2’s initial proposed protective order. But Databricks’ revisions were
`
`reasonable and standard, and R2 itself acknowledges that the Parties worked together to reach a
`7
`
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 8 of 10 PageID #: 1583
`
`“compromise version” on all but the currently disputed provision regarding software tools. Indeed,
`
`Chief Judge Gilstrap entered a Protective Order in another case with modifications very similar to
`
`those that Databricks proposed here, including Databricks’ proposed language for paragraph 12(g)
`
`that is the subject of this dispute. (See Byteweavr, LLC v. Databricks, Inc., No. 2-24-cv-00162-JRG-
`
`RSP Dkt. 32 at 8–9 (E.D. Tex. June 30, 2024)).
`
`For all the reasons discussed above, Databricks respectfully requests that the Court enter
`
`Databricks’ proposed protective order (Exhibit B).
`
` Dated: October 16, 2024
`
`Respectfully submitted,
`
`/s/ Edward R. Nelson III
`EDWARD R. NELSON III
`State Bar No. 00797142
`ed@nelbum.com
`BRENT N. BUMGARDNER
`State Bar No. 00795272
`brent@nelbum.com
`CHRISTOPHER G. GRANAGHAN
`State Bar No. 24078585
`chris@nelbum.com
`JOHN P. MURPHY
`State Bar No. 24056024
`murphy@nelbum.com
`CARDER W. BROOKS
`State Bar No. 24105536
`carder@nelbum.com
`NELSON BUMGARDNER CONROY PC
`3131 West 7th Street, Suite 300
`Fort Worth, Texas 76107
`817.377.9111
`
`COUNSEL FOR PLAINTIFF
`R2 SOLUTIONS LLC
`
`8
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 9 of 10 PageID #: 1584
`
`Dated: October 16, 2024
`
`
`
`
`
`
`
`/s/ Jessica M. Kaempf
`Michael J. Sacksteder
`CA Bar No. 191605 (Admitted E.D. Texas)
`Email: msacksteder@fenwick.com
`Gregory Sefian
`CA Bar No. 341802 (Admitted Pro Hac Vice)
`Email: gsefian@fenwick.com
`Su Li
`CA Bar No. 339374 (Admitted Pro Hac Vice)
`Email: sli@fenwick.com
`FENWICK & WEST LLP
`555 California Street, 12th Floor
`San Francisco, California 94104
`Telephone: 415.875.2300
`Facsimile: 415.281.1350
`
`Vigen Salmastlian
`CA Bar No. 276846 (Admitted E.D. Texas)
`Email: vsalmastlian@fenwick.com
`FENWICK & WEST LLP
`801 California Street,
`Mountain View, CA 94041
`Telephone: 650.988.8500
`Facsimile: 650.938.5200
`Jessica M. Kaempf
`WA Bar No. 51666 (Admitted E.D. Texas)
`Email: jkaempf@fenwick.com Jonathan G.
`Tamimi
`WA Bar No. 54858 (Admitted E.D. Texas)
`Email: jtamimi@fenwick.com
`FENWICK & WEST LLP
`401 Union Street, 5th Floor
`Seattle, WA 98101
`Telephone: 206.389.4510
`Facsimile: 206.389.4511
`
`Attorneys for Defendant Databricks, Inc.
`
`
`
`
`
`
`CERTIFICATE OF SERVICE
`
`I hereby certify that a copy of the foregoing document was filed electronically in compliance
`
`with Local Rule CV-5(a). Therefore, this document was served on all counsel on October 16, 2024.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`9
`
`
`
`/s/ Edward R. Nelson III
`
`

`

`Case 4:23-cv-01147-ALM Document 52 Filed 10/16/24 Page 10 of 10 PageID #: 1585
`
`CERTIFICATE OF CONFERENCE
`
`I hereby certify that counsel for Plaintiff and counsel for Defendant conferred regarding
`
`the relief sought by this motion and that this motion is unopposed.
`
`/s/ Edward R. Nelson III
`
`10
`
`