Compulife Software, Inc. v. Rutstein

UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF FLORIDA COMPULIFE SOFTWARE, INC., Plaintiff, Case No. 9:16-CV-80808-BER v. BINYOMIN RUTSTEIN A/K/A BEN RUTSTEIN, DAVID RUTSTEIN A/K/A DAVID GORDON A/K/A BOB GORDON A/K/A NATE GOLDEN AND JOHN DOES 1 TO 10, Defendants. COMPULIFE SOFTWARE, INC., Plaintiff, Case No. 9:16-CV-81942-BER v. MOSES NEWMAN, DAVID RUTSTEIN, BINYOMIN RUTSTEIN AND AARON LEVY, Defendants. SUPPLEMENTAL FINDINGS OF FACT AND CONCLUSIONS OF LAW Compulife Software, Inc. (“Compulife”) offers life insurance price quotes. One way for a person to get a quote is to fill out an online form, which triggers Compulife’s proprietary database to generate the quote. In computer-ese, that online form is a “dialog box.” Compulife’s HTML source code causes the dialog box to appear on the user’s computer screen. Defendants copied large portions of that HTML source code, stole Compulife’s database, and began competing against Compulife. After two trials and two trips to the Eleventh Circuit, the remaining legal question is whether copying the HTML source code amounted to copyright infringement. More specifically, the Eleventh Circuit has directed me to make findings about whether the arrangement of the HTML source code as a whole has copyright protection. For the following reasons, I conclude that it does not. I. PROCEDURAL HISTORY This litigation involves two consolidated cases. The first (Case No. 16-80808) alleged direct copyright infringement (Count I), contributory copyright infringement (Count II), unfair competition under the Lanham Act (Count III), federal theft of trade secrets (Count IV), Florida theft of trade secrets (Count V), violation of the Florida Computer Abuse and Data Recovery Act (“CADRA”) (Count VI), Florida unfair competition (Count VII), and violating the Florida Deceptive and Unfair Trade Practices Act (“FDUTPA”) (Count VIII). The second (Case No. 16-81942) alleged Federal trade secrets theft (Count I), direct copyright infringement (Count II), contributory copyright infringement (Count III), Lanham Act unfair competition (Count IV), Florida trade secret theft (Count V), CADRA (Count VI), and Florida unfair competition (Count VII). On March 12, 2018, final judgment was entered for Defendants on all claims in both cases. Compulife appealed. The Eleventh Circuit affirmed the judgments on the Lanham Act, FDUTPA, CADRA, and Florida unfair competition claims. Compulife Software Inc. v. Newman, et. al., 959 F.3d 1288 (11th Cir. 2020) (Compulife — 2 — I). It remanded the trade secret and copyright infringement claims for further findings. After a second trial, final judgment was entered for Compulife on its trade secret claims (Counts IV and V in the ’08 case and Counts I and V in the ’42 case). Final judgment was entered in favor of Defendants on the direct and contributory copyright infringement claims (Counts I and II in the ’08 case and Counts II and III in the ’42 case). Both sides appealed the final judgment. The Eleventh Circuit affirmed the final judgments against Defendants on the trade secret claims; it reversed the judgment on the copyright claims and remanded for further proceedings. 111 F. 4th 1147 (11th Cir. 2024) (Compulife II): Although the district court considered the selection and arrangement of Compulife's code to some degree, the district court never identified the entire arrangement of these variables in the code as a constituent component of the code. For example, the district court expressly evaluated the arrangement of the birth month, birthday, and birth year variables before filtering. But it didn't look at the arrangement of all the variables together. And, relying on Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 , 347–48, 111 S.Ct. 1282 , 113 L.Ed.2d 358 (1991), the district court recognized that factual compilations, “like the ones performed by Compulife's software in compiling facts ... to generate a quote” can be protectable. But the arrangement of the code itself can be protectable, not just the results produced by the software. Given that the arrangement of the code may be protectable, we agree with Compulife that the district court should have abstracted the “arrangement” as something to be analyzed at the subsequent filtration step. Compulife II at 1158. — 3 — II. LEGAL PRINCIPLES Mandate Rule I first acknowledge the limitations imposed by the mandate rule and the law of the case doctrine. As the Eleventh Circuit has explained: “The mandate rule is a specific application of the ‘law of the case’ doctrine which provides that subsequent courts are bound by any findings of fact or conclusions of law made by the court of appeals in a prior appeal of the same case.” Friedman v. Mkt. St. Mortg. Corp., 520 F.3d 1289, 1294 (11th Cir. 2008) (quotation marks omitted). “The law of the case doctrine and the mandate rule ban courts from revisiting matters decided expressly or by necessary implication in an earlier appeal of the same case.” AIG Baker Sterling Heights, LLC v. Am. MultiCinema, Inc., 579 F.3d 1268 , 1270–71 (11th Cir. 2009). It has its greatest force when a case is on remand to the district court. “When an appellate court issues a clear and precise mandate, ... the district court is obligated to follow the instruction. Neither the district court nor any party is free to ignore the law of the case.” Litman v. Mass. Mut. Life Ins. Co., 825 F.2d 1506, 1516 (11th Cir. 1987). “A district court when acting under an appellate court's mandate, cannot vary it, or examine it for any other purpose than execution; or give any other or further relief; or review it, even for apparent error, upon a matter decided on appeal; or intermeddle with it, further than to settle so much as has been remanded.” Id. at 1510–11 (quotation marks omitted). Winn-Dixie Stores, Inc. v. Dolgencorp, LLC, 881 F.3d 835, 843 (11th Cir. 2018). These consolidated cases were remanded for the limited purpose of making factual and legal findings about whether the arrangement of the HTML source code is protectable under the federal copyright laws. I therefore will not reopen the record; I will consider only the evidence and arguments presented at trial. Even if the mandate rule allowed me to reopen the record, I would decline to do so. Both sides had a full opportunity at the trial to develop the facts, to argue their legal positions, and to submit proposed findings of fact and conclusions of law. Neither — 4 — side argued on appeal that I erred by excluding any evidence or by limiting legal arguments. Under these circumstances, the interests of justice do not require that the parties be given a “second bite at the apple.” Moreover, neither party asked to supplement the record after the mandate issued. Copyright Infringement To succeed on its copyright infringement claim, Compulife must prove (1) it held a valid copyright in the HTML source code and (2) Defendants factually and legally copied constituent elements of the work that are original. Compulife II at 1156. There is no remaining dispute that Compulife held a valid copyright registration for the HTML source code or that Defendants factually copied portions of that code. Compulife I, 959 F.3d at 1301-02 . It is the law of the case that the individual variables in the HTML source code were not legally copied. As discussed more fully below, some, but not all, of the arrangement of the code was factually copied. Compulife bears the ultimate burden of proving legal copying. Compulife I at 1301. “‘Legal’— or ‘actionable’ — copying occurs when ‘those elements of the [copyrighted work] that have been copied are protected expression and of such importance to the copied work that the appropriation is actionable.’” Id. at 1302 (brackets in original) (cleaned up). That is, “the portion of the copyrighted work actually taken [must] satisfy the constitutional requirement of originality.” Bateman v. Mnemonics, Inc., 79 F.3d 1532, 1542 (11th Cir. 1996) (citing Feist Publications, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 345-46 (1991)). And, there must be a — 5 — “‘substantial similarity’ between the allegedly offending program and the protectable, original elements of the copyrighted works.” Bateman, 79 F.3d at 1542 . To decide whether the legal copying of a computer program is actionable, the Court employs a three-step process: (1) abstraction, (2) filtration, and (3) comparison. Bateman, 79 F.3d 1543 -46. For the abstraction step, the Court “break[s] down the allegedly infringed program into its constituent structural parts.” Id. at 1544 (quoting Computer Assocs. Int’l v. Altai, Inc., 982 F.2d 693, 706 (2d Cir. 1992)) (bracket in original). For the filtration step, the Court extracts the unprotectable portions of the copyrighted work. “It is at this step that the structural components at each level of abstraction are examined ‘to determine whether their particular inclusion at that level was ‘idea’ or was dictated by considerations of efficiency, so as to be necessarily incidental to that idea; required by factors external to the program itself; or taken from the public domain and hence is nonprotectable expression.’” Bateman, 79 F.3d at 1544 (citing Altai, 982 F.2d at 707 ) (cleaned up). At the filtration step, the defendant bears the burden of proving unprotectability; he must identify “the species of unprotectability that he is alleging and [must] present supporting evidence where appropriate.” Compulife I at 1306. The plaintiff then faces “the manageable task of responding to the appropriately narrowed issue.” Id. (cleaned up). “That is not to say that the defendant must always introduce evidence in order to enable the district court to filter. The defendant may sometimes — 6 — be able to demonstrate by argument alone that an element of a copyrighted work is unprotected.” Id., n.8 As relevant here, there are two ways that the arrangement of information can be unprotected. First, it may not be sufficiently original. Compulife I, 959 F.3d at 1304 . “The sine qua non of copyright is originality.” Feist, 499 U.S. at 345 . “[C]opyright protection extends only to a work’s expressive elements, not to any underlying ‘idea, procedure, process, system, method of operation, concept, principle, or discovery’ expressed therein.” Compulife I at 1304 (citing 17 U.S.C. § 102 ). “Copyright infringement occurs only if one copies protected elements of a copyrighted work; in other words, the portion of the copyrighted work that is copied must ‘satisfy the constitutional requirement of originality as set forth in Article I, § 8, cl. 8.’” MiTek Holdings, Inc. v. Arce Eng'g Co., 89 F.3d 1548, 1554 (11th Cir. 1996) (quoting Bateman, 79 F.3d at 1542 ). [O]riginality is not a stringent standard; it does not require that facts be presented in an innovative or surprising way. It is equally true, however, that the selection and arrangement of facts cannot be so mechanical or routine as to require no creativity whatsoever. The standard of originality is low, but it does exist. As this Court has explained, the Constitution mandates some minimal degree of creativity, and an author who claims infringement must prove “the existence of ... intellectual production, of thought, and conception.” Feist, 499 U.S. at 362 (internal citations omitted). For a compilation, there must be “a creatively original selection of facts” to warrant copyright protection. BUC Int'l Corp. v. Int'l Yacht Council Ltd., 489 F.3d 1129, 1141 (11th Cir. 2007) (emphasis in original). — 7 — Second, material may be unprotected based on the merger doctrine because there are too few ways of expressing the relevant idea. Id. And, while the selection and arrangement of information may satisfy the originality requirement in certain circumstances, the merger doctrine precludes a finding of originality where there are “so few ways of expressing an idea.” Id. at 1143 . For the comparison step, the Court compares any remaining (i.e., unfiltered) protectable material to the allegedly infringing work. Where, as here, the copied materials are literal elements of the computer program, the plaintiff has the burden of showing that the two works are “substantially similar.”1 Compulife I at 1302. Substantial similarity has both qualitative and quantitative components. Id. Put simply, a person can copy only a little bit of code, but if that code is qualitatively significant, there is an infringement. Id. 1 Source code is a “literal element” of a computer program. Compulife I, 959 F.3d at 1302 n.6. In contrast, nonliteral elements include “‘the products that are generated by the code’s interaction with the computer hardware and operating program(s),’ of which ‘screen displays and the main menu’ are illustrative examples.” Id. (quoting MiTek Holdings, Inc. v. Arce Eng'g Co., 89 F.3d 1548 (11th Cir. 1996)). A dialog box is a nonliteral element. — 8 — III. FINDINGS OF FACT2 A. User Input Form and Parameter Blocks 1. Compulife offers internet-based insurance quotes through its Terms4Sale.com website. ECF No. 309 at 119:14-19.3 2. HTML source code defines how a web page looks to the user. ECF No. 309 at 130:6-12. 3. Chris Bruner wrote the HTML source code for the Terms4Sale.com website by himself. He did not copy any parts of it from anyone else. ECF No. 309 at 120:9-13. 4. Compulife’s HTML source code displayed the following dialog box — a fill-in-the-blank form — to the website user: 2 Federal Rule of Civil Procedure 52 requires “the court must find the facts specially and state its conclusions of law separately.” Fed. R. Civ. P. 52(a)(1). 3 All citations in this decision are to the filings in Case No. 16-cv-80808. — 9 — State Califomia we Birthdate June w 15 □ 1975 ~ Gender @ Male © Female Smoker / Tobacco @No O Yes Health Class Preferred Plus vw Type of Insurance 20 Year Level Term Ww Amount of Insurance $500,000 Powered by COMPULIFE® PX 550 (recited as 505 (ECF No. 304-29 at 9).4 5. The dialog box was used to “gather[] information from the end user to skip parameters for the [search] engine, things like their age, their sex, smoking status, their health, and various other things.” ECF No. 309 at 121:9-17. 6. The HTML source code then sent the collected information to the search engine, which generated a template showing the user the price quote. Jd. at 121:1417. 7. Within the HTML environment, the dialog box was a table with multiple rows. ECF No. 310 at 130-131. In HTML programming language, the tag <tr> stands 4 “PX” refers to Plaintiff's Exhibits, which are docketed at ECF Nos. 295 and 304. Defendants’ Exhibits (“DX”) are docketed at ECF No. 298. — 10 — for “table row.” Id. at 131. It starts a new row in the table. The end of each row was coded with the tag </tr>. ECF No. 311 at 70:9-11. 8. The HTML source code included mandatory programming commands necessary to generate the form. ECF Nos. 310 at 145:15-146:8; 311 at 4-6. 9. The Compulife HTML source code also included comments, which are non-functional portions of the program.5 Each comment began with the “<!--” tag and ended with the “-->” tag. https://web.mit.edu/thecity/resources/html cheat.html (last visited Dec. 12, 2024). For example, beginning on the first page of PX 542 (ECF No. 295-58), the program contains a block of code beginning with “<!-- Select a Template File For Your Results” followed by an explanation of the options and how they must be formatted. The block ends on the next page with the line TEMPLATE_MORE_INFO.HTM -->.” PX 542 at 1-2. Similarly, before the code block for the NewCategory parameter, there is a commented discussion about how that parameter works, including how to allow the user to select multiple categories. Id. at 7. 10. The table could be formatted using tags such as <td> and <div>. ECF No. 310 at 8:15-19; 131. The <td> tag indicates the beginning of a data element. ECF No. 311 at 70:3-5. 5 One Court described comments as “non-executable appendages to lines of executable code.” Tradescape.com v. Shivaram, 77 F. Supp. 2d 408, 418 (S.D.N.Y. 1999). — 11 — 11. Some rows were a static title that was visible to the user, such as “State,” “Birthdate,” and “Gender.” 12. Some rows were designed to obtain information from the user through either a dropdown menu or radio button. The information selected by the user was assigned to a variable.6 13. It might be necessary to have multiple lines of HTML code to generate a visible row in the table. For example, multiple lines of HTML code were needed for each dropdown menu and radio button. 14. A radio button could be added to a row using the <INPUT type=“radio”> command. Each radio button had a unique value that, when selected, was assigned to a corresponding variable. For example, the following two lines of HTML code created the radio buttons for whether the user was a smoker: Yes&nbsp;<INPUT type=“radio” name=“Smoker” value=“Y”> No&nbsp;<INPUT type=“radio” checked name=“Smoker”value=“N”> See PX 542 at 6.7 Depending on the user’s selection, the parameter “Smoker” was assigned either the value “Y” (for “yes”) or “N” (for “no”). ECF No. 309 at 149:19-150:3. 15. To include a dropdown menu in a row, the “<Select Name>” command was used. Each item in the dropdown menu was associated with a unique value, using the “<Option>” command. When the user selected an item, the unique value for that 6 I use the terms “variable” and “parameter” as synonyms in this Order. 7 Mr. Bruner testified that “&nbsp” was a formatting code for “non-breaking space.” ECF No. 309 at 146:21-22. — 12 — selection was assigned to the parameter associated with the “<Select Name>” command. For example, the dropdown menu for the user’s state of residence included 56 items. Whichever item the user picked assigned a unique value to the parameter “State” based on the command <Select name = “State”>. ECF No. 309 at 146:15-19. 16. The HTML source code contained 11 active variables: State, BirthMonth, Birthday, BirthYear, Sex, Smoker, Health, New Category, ModeUsed, SortOverride1, and FaceValue.8 17. I will use the term “parameter block” to refer to the multiple lines of HTML code necessary to generate a variable’s value (the HTML code between each set of <tr> and </tr> tags). For example, the parameter block for the “State” variable comprises approximately 62 lines of HTML code. See PX 542 (ECF No. 295-58) at 23; PX 149 (ECF No. 295-21) at lines 514-571. The parameter block for the “Sex” variable is seven lines long. PX 542 at 6; PX 149 at lines 735-741. 18. The user-defined variables BirthMonth, Birthday, and BirthYear are all generated by the same row bounded by <tr> and </tr>, so they are part of a single parameter block of over 160 lines of code. PX 149 (ECF No. 295-21) at lines 572-734. 8 The Compulife HTML source code has a non-functional comment (beginning with <!-- <tr> and ending with <tr> -->) containing a block of code that would have generated the variable “CompRating.” PX 542 at 9. It has a different non-functional comment containing HTML code that would generate a row titled “Payment Option” with a dropdown menu for the variable ModeUsed. PX-542 at 8. Defendants did not copy either of these commented parameter blocks. — 13 — 19. The Compulife HTML source code also contained two variables: ModeUsed and SortOverride1, which had fixed, preset values that could not be changed by the user. PX 149 at lines 507-08; PX 542 at 8; ECF No. 309 at 181:12-17. 20. The ModeUsed and SortOverride1 variables did not require user input, so they were not assigned to a row in the table (that is, they are not within a code block bounded by <tr> and </tr> tags). Nevertheless, I will treat them each as a parameter block because Compulife argues they are a constituent component of the HTML source code. Compulife Appellate Brief, 2022 WL 13821778 at *10-11. 21. The Compulife HTML source code had the following nine (9) sequential parameter blocks (PX 542): Row Title Selection Options Underlying Variable(s) State Dropdown list of state State abbreviations Birthdate Dropdown list of months BirthMonth Dropdown list of 1-31 Birthday Dropdown list of years BirthYear Gender Radio Button: male or Sex female Smoker/Tobacco Radio Button: yes or no Smoker Health Class Dropdown list of health Health categories Type of Insurance Dropdown list of NewCategory different terms and kinds of insurance coverage Not assigned to a row None ModeUsed Not assigned to a row None SortOverride1 Face Amount Dropdown list of amount FaceAmount of coverage requested — 14 — 22. The relevant portion of Defendants’ HTML code contained the following nine (9) sequential parameter blocks (PX 149): Row Title Selection Options Underlying Code Lines Variable(s) Not assigned to a None ModeUsed 507 row Not assigned to a None SortOverride1 508 row State Dropdown list of state State 514-571 abbreviations Birthdate Dropdown list of months BirthMonth 572-588 Dropdown list of 1-31 Birthday 590-623 Dropdown list of years BirthYear 625-734 Gender Radio Button: male or Sex 735-740 female Smoker? Radio Button: yes or no Smoker 741-746 Health Class Dropdown list of health Health 747-758 categories Type of Dropdown list of different NewCategory 758-793 Insurance terms and kinds of insurance coverage Face Amount Dropdown list of amount FaceAmount 794-836 of coverage requested 23. The dialog box also included a button where the user could submit the relevant information to the Compulife search engine. The Compulife HTML source code showed the user a button with the text “Compare Now.” Within the HTML code, that button was coded using <INPUT type = “submit”> and assigned a value to the parameter “CqsComparison.” 24. There are no programmer’s comments explaining why Mr. Bruner organized or arranged the blocks of code in a particular order. — 15 — 25. Most life insurance companies and products require the same input parameters that Compulife used. ECF No. 309 at 28:18-25. 26. All of the variables were independent. That is, the value selected for one variable did not affect the value of any other variable. 27. The Compulife HTML source code did not analyze or process the variables. ECF No. 309 at 122:10. B. Quote Generating Process 28. Once the user clicked the “Compare Now” button, the HTML source code sent the user’s information to the internet search engine using the <FORM action=> command and the “method=’Post’” command. ECF Nos. 309 at 155:24-156:5; 311 at 12:13-20; PX 542 at 1. 29. The internet search engine is on Compulife’s server. ECF No. 310 at 18:24-19:2. 30. The HTML source code would send the user’s information to CQSL.CGI, which was an executable script that ran on Compulife’s server. ECF No. 309 at 155:24-156:5; ECF No. 310 at 15:7-9. 31. All of the variables were sent to the search engine simultaneously. 32. The search engine is written in C++ programming language. ECF No. 309 at 121:7-8. 33. The names and formatting of the variables being sent by the HTML source code had to match the names and formatting of the variables in the search — 16 — engine. Otherwise, the search engine would not be able to understand what was being passed. ECF No. 309 at 121:18-122:3; 142:15-24; ECF No. 310 at 27:24-29:3. 34. Values for the variables State, BirthMonth, Birthday, BirthYear, Sex, Smoker, Health. NewCategory, ModeUsed, SortOverride1, and FaceAmount all had to be passed to the search engine to get a quote. ECF No. 309 at 146:4-8. 35. Assuming the variables matched, the search engine queried the underlying Transformative Database of insurance rates and sent a price quote back to the end user on a templated form. ECF No. 310 at 27:10-23. — 17 — 36. In sum, the process worked like this: END USER User inputs person al information and desired kind of insurance into HTML-created form; Template with insurance price quote variable values ass igned displayed to end user COMPULIFE WEBSITE HTML source code passes variables and Price quote generated and inserted into accompanying values to search engine template INTERNET SEARCH ENGINE C++ code queries database based on values Rate information returned to C++ assigned to variables search engine INSURANCE RATE DATABASE 37. If the parameter blocks had been arranged differently, the user would have seen the same rows in the table, but in a different arrangement. For example, the row for “State” might have appeared after the row for “Gender.” Regardless of this rearranging, the same variable values would have been sent to the search engine, which would have generated the same insurance price quote. — 18 — 38. The order of the variables (and hence of the parameter blocks) in the HTML source code did not matter to the functioning of the quote-generating process. ECF No. 309 at 142:13-14. This final factual finding — that the variables did not have to be sent to the search engine in a particular order — warrants more extended discussion because it differs from my original findings of fact. Upon further review of the record, I erroneously adopted Compulife’s position that the order of the parameter blocks in the source code mattered to whether the HTML code could obtain a price quote from the internet search engine. I said, “All the parameters must be present, spelled correctly and provided in the correct order for the software to produce quotes.” ECF No. 314 at 8, ¶17. I cited to the following italicized testimony from Mr. Bruner, which was the portion of the record cited to support the proposed finding of fact: Q: Now, turning your attention to the HTML code, can you describe what that code does? A: It gathers information from the end user to skip parameters for the engine, things requested like their age, their sex, smoking status, their health, and various other things, it can go into health analysis and everything else. And from that, the HTML calls the engine which does the information lookup, goes to the template files and produces the results back through the template files. Q: Are there any requirements for the HTML to work correctly with the engine? A: The parameters need to be spelled correctly and there are certain things that have to be there in order to get valid results. Q: So the engine is expecting the information to come to it in a particular way? A: Yes. Q: And if it doesn’t come to it that way, it won’t operate? — 19 — A: No. It will give an error. ECF No. 309 at 121:9-122:3. Mr. Bruner did not testify that the variables had to be provided to the search engine in a particular order. There was no testimony that the search engine would have worked differently if the parameter blocks had been rearranged, so long as all parameter names were properly spelled and formatted. In fact, Mr. Bruner testified that putting the {SortOverride1, ModeUsed} parameter block at a different place in the sequence did not affect the quote generating process. ECF No. 309 at 142:13-18. Mr. Bruner also testified that the HTML source code did not process or analyze the variable values; it merely passed them over to the search engine. When the user clicks the “Compare Now” button, it activates the following command: <FORM action= “/web/201001160033856/http://www.compulife.net/cgibin/cqsl.cgi” method= “POST”> The <method= “POST”> part of this command sends the parameter values simultaneously to the search engine. ECF No. 311 at 12:13-20. There is only one such command in the source code, so the circumstantial evidence in the record shows that all variable values are passed simultaneously. There was no evidence to the contrary. See also https://developer.mozilla.org/en-US/docs/Web/HTTP/Methods/POST (last visited January 4, 2025) (POST command sends a single, concatenated array of all variable values). — 20 — I find as a matter of fact that the arrangement/order of the parameter blocks does not affect the HTML source code’s ability to communicate with the C++ search engine. So, it was not necessary to have the parameter blocks in any particular order. IV. COMPULIFE’S ARGUMENTS As summarized by the Eleventh Circuit, “Compulife argues that the arrangement of its various components of source code — state, birth, month, birthday, birth year, sex, smoking status, health classification, insurance type, payment options, sorting — is creative and therefore protectable.” 111 F.4th at 1157 (italics in original). In his opening statement (which is not evidence), counsel for Compulife said: After Mr. Barney’s testimony, Mr. Bruner, who wrote the code, both the code that operates the software and the HTML that resides on websites that communicates with the software, he will explain how that was written by him originally, how it is creative, and he will also explain how it was taken by the Defendants in two ways; one, in the first case, and one in the second case.” ECF No. 308 at 19:10-16 (emphasis added). Compulife’s counsel also said: And what we will explain to your Honor is that this original creative material is divided into several different categories, code concerning state selection and numbering, code concerning birthdates, code concerning gender, code concerning smoker or nonsmoker status, different types of health codes that relate to insurance, different types of category selection codes that relate to insurance, mode used codes, overwrite codes, and face amount codes. And what the witnesses will explain to your Honor is that this is protectable code. This is not in the nature of unprotectable facts or very simple alphabetization; rather, this code was the result of very specific creative choices and original choices that Compulife made in the course of writing the code. You will hear from the author of the code, Mr. Bruner, about those. — 21 — Id. at 22:22-23:11 (emphasis added). In his closing argument, counsel for Compulife said that Ms. Miracle, Mr. Barney, and Mr. Bruner gave evidence showing that the arrangement of the parameter blocks was protected: The HTML code, as you heard from Mr. Bruner and Mr. Barney, and from our expert, Nancy Miracle, the HTML code has parameters in it that are necessary to communicate with the software, with the server software. Without formatting requests in the right way, the software won't respond. The copying was necessary in order for it to function. . . . Plaintiff presented expert testimony regarding all of the different elements of the HTML code, that the HTML code that was used didn't need to be organized the way it was, it could have been organized in a number of different ways. Ms. Miracle testified about that and I will go to her PowerPoint in a second. In the second case, the copying that occurred was of, again, necessary HTML code in order to scrape the quotes from the Term4Sale website, and again, there was no legal copying dispute offered by the Defense. ECF No. 313 at 6-7 (emphasis added). In its Proposed Findings of Fact and Conclusions of Law, Compulife proposed the following relevant findings of fact, based on the evidence it asserted was in the record: To accomplish the task of looking up insurance quotes for a user, the HTML code contains different blocks of code: each block of code relates to the different information needed from a user to produce an insurance quote: state selection code, (Vol. 2, 139:10-20); birthdate and birth month selection code (Vol. 2, 139:21-140:6); birth year code written in camel case (Vol. 2, 140:7-9); gender selection code (Vol. 2, 140:11-17); smoker/tobacco code (Vol. 2, 140:18-21); health class code1 (Vol. 2, 140:22-141:7); insurance category code (Vol. 2, 141:8-142:12); mode used — 22 — code (Vol. 2, 154:9-14); the code for sorting output information (Vol. 2, 155:1-11); and face amount code (Vol. 2, 142:19-143:8). In order for the Compulife HTML code to communicate with the Compulife internet engine, the HTML code must send the engine the correct parameters and variables defined in the HTML code so that the internet engine can look up the insurance rates in the Compulife software and produce an insurance quote; it must be a one-to-one match. (Vol. 3, 27:24-29:3). The parameters are also contained in the Compulife internet engine source code written in C++. (Vol. 3, 24:25-25:11, PX 542). The Compulife internet engine is expecting the parameters and variables to come to it in the particular way it is written in the Compulife HTML, and if it does not come to the engine in that way the software will spit out an error and not produce results. (Vol. 2, 121:18122:3). The parameters, which correspond to the blocks of HTML code, are: “State”, “BirthMonth”, “Birthday”, “BirthYear”, “Sex”, “Smoker”, “Health”, “NewCategory”, “ModeUsed”, “SortOverride1”, and “FaceAmount.” (Vol. 2, 146:4-8, PX 542, 567). All the parameters must be present, spelled correctly and provided in the correct order in order for the software to produce quotes. (Vol. 2, 121:18-22). ECF No. 307 at 9-11 (¶¶ 11-12). Compulife proposed the following relevant conclusions of law on the issue of whether it held a valid copyright on the HTML source code: Even were the validity of Compulife’s copyright disputed by defendants, Compulife demonstrated that its 2010 HTML code is original creative authorship entitled to copyright protection. Bruner, Compulife’s programmer, wrote the 2010 HTML code himself and did not copy it from anyone else. Compulife’s expert, Miracle, opined that the 2010 HTML code contained numerous creative elements including a creative way to identify and organize variables related to requests for insurance quotation information. Defendants offered no evidence demonstrating Compulife’s 2010 HTML code “does not satisfy the constitutional requirement of originality as set forth in Article I, § 8, cl. 8,” Bateman, 79 F.3d at 1542 , a requirement that “variously has been characterized as ‘modest,’ ‘minimal,’ and ‘a low threshold.’” Original Appalachian Artworks, 684 F.2d at 824. Therefore, based on either the presumption Compulife’s registration certificate, the doctrine of law of the case, or the evidence at trial, Compulife owns a valid copyright in its 2010 HTML code. — 23 — ECF No. 307 at 39-40. In the Legal Copying section of the proposed conclusions of law, Compulife’s entire argument was: Compulife also proved both the qualitative and quantitative significance of the HTML code copied by the defendants. The Compulife 2010 HTML code totals 347 lines. (Vol. 3, 147:4). Thirteen of those lines are mandatory and have to be in the code for it to function. (Vol. 4, 4:15-5:5). Of the remaining code, defendants copied 282 lines of Compulife’s HTML code, or 84% of the total lines of code. (Vol. 3, 144:7-146:22). The 282 lines of code defendants copied perform the key functions needed to produce an insurance quote from user input: the state selection code, the birthdate and birth month selection code, the birth year code written in camel case, the gender selection code, the smoker/tobacco code, the health class code, the insurance category code, the mode used code, the code for sorting output information, and the face amount code. In reaching this conclusion I credit the opinion of Compulife’s expert Nancy Miracle that defendants copying of Compulife’s HTML code was qualitatively and quantitatively significant. (Vol. 4, 27:3-8). Defendants failed to meet their burden to show that the elements they copied from Compulife’s HTML code were unprotectible. Defendants offered no evidence on this point, only speculation. Defendants presented no expert testimony either. Compulife proved legal copying of its 2010 HTML code in the ’08 case. ECF No. 307 at 41. Finally, in its appellate brief, Compulife argued "the organization and arrangement of Compulife’s HTML code was original and minimally creative to entitle Compulife to copyright protection.” 2022 WL 13821778 , at *17. It said: The Compulife 2010 HTML Code that defendants copied is made up of eleven code blocks: “State”, “BirthMonth”, “Birthday”, “BirthYear”, “Sex”, “Smoker”, “Health”, “NewCategory”, “ModeUsed”, “SortOverride1”, and “FaceAmount.” The order they appear in above is the order they must appear in in the copied code for the HTML to work properly and return quotes. Compulife's programmer Bruner chose the organization of those code blocks himself. The defendants failed to offer any evidence to show that the organization or arrangement of those code — 24 — blocks was anything other than original and creative. The organization and arrangement of the eleven parameters for the variables required to return an insurance quote is not dictated by any outside source, nor was there any testimony or evidence to demonstrate that, for example, “State” must come first always, followed by “BirthMonth”, followed by “Birthday” and then “BirthYear”, etc. The eleven code blocks in Compulife's HTML could have been arranged differently in up to over 39 million permutations without repeating any of the code blocks. Id. at *17-18 (emphasis added). V. DEFENDANTS’ ARGUMENTS In their proposed findings of fact and conclusions of law, Defendants argued that the online form was not creative because “there are very few ways to create a form to collect the necessary information to generate a [life insurance] quote.” ECF No. 306 at 19. On appeal, they said the arrangement of the parameter blocks was not sufficiently original. Defendants’ Appellate Brief, 2022 WL 17960736 at *14-21. More specifically: “The relevant inquiry is not whether there is some imaginable, although manifestly less useful, method of” arrangements. A finding of originality merely because Compulife “could have arranged” its code blocks in different way [sic] “misapprehends the question.” “The pertinent inquiry is whether [Compulife] has demonstrated originality, the ‘sine qua non’ of copyright, in its arrangement or coordination.” Compulife fails in this regard, and the mere fact that its programmer created the Source Code is of no moment. . . . While in theory there are numerous ways to organize this information, in practical, the district court correctly found that “there are only [a] few methods by which Compulife can gather and compile the information — 25 — needed to generate a life insurance quote.” Indeed, it would make no sense to put the code blocks concerning insurance category, mode, and face amount before those concerning a user’s biographical information (state, birth date, month, and year, gender) and smoking and health status. Therefore, the way Compulife selected and arranged the Source Code is not entitled to copyright protection. Id. at *15 , 18 (citing BellSouth Advert. & Publ’g Corp. v. Donnelley Info Publ’g, Inc., 999 F.2d 1436 (11th Cir. 1993) (pincites omitted) (second bracket in original). Defendants also argued they had elicited affirmative testimony of unprotectability on cross-examination of Mr. Barney, Mr. Bruner, and Ms. Miracle. Id. at *21. VI. TRIAL EVIDENCE Mr. Barney Despite counsel’s expectations at the time of opening statements, Mr. Barney did not testify about the arrangement of the HTML code. Moreover, Compulife did not cite to Mr. Barney’s testimony in support of the proposed findings of fact or conclusions of law related to the arrangement of the code. Mr. Bruner Similarly, contrary to counsel’s expectations, Mr. Bruner did not testify that he made specific creative choices about how to organize the blocks of HTML code. At best, his testimony was that (1) he wrote the code from scratch, (2) he did not copy any of it from anyone else, and (3) it could have been written differently. He did not testify whether he had considered other ways to arrange the blocks of code. He also did not include any comments in the code to explain why he arranged it as he did. — 26 — Ms. Miracle As best I can discern, there are three relevant portions of Ms. Miracle’s trial testimony that touched on the arrangement of the HTML source code. None support a finding that the arrangement of the parameter blocks gets copyright protection. The first was: Q. Have you done any analysis, come to any conclusions about the originality and creativity of the code or the organization and arrangement of that HTML code? A. Yes. Q. And what is that? A. The arrangement of the -- you have to understand HTML is a whole document, okay, so I am not talking about little bits and pieces, but the arrangement of the pieces within the document, the use of the HTML language in an expressive fashion, and the elements that are uniquely picked, such as the data values and the data names, constitute a unique whole. I haven't seen anything quite like it. ECF No. 310 at 127:18-128:4. Ms. Miracle was offering a factually-unsupported conclusion. There is no explanation for why or how the “arrangement of the pieces within the document” is unique or how the HTML language is being used “in an expressive fashion.” Here, too, at best the evidence is that the parameter blocks could have been arranged differently. The second relevant part of Ms. Miracle’s testimony was a non-responsive answer to the question “And so the record is clear, if you can please verbally explain the percentage of code that has been copied and the number of lines that you found were copied?” ECF No. 310 at 145:9-11. As part of her answer, she testified: — 27 — The part that is discretionary, and the part where the programmer's creativity shows -- don't laugh at me, programmers are creative -- is the part where Chris Bruner went ahead and described what he wrote, how he wrote it, why he picked those names, and the fact that those names are not only required here, the names are required in the C++ code to which this HTML is the front end, and to which this HTML prepares input to be able to invoke it correctly. Those are unique and, you know, I look at so much HTML, and there are things that are always the same, the head tags are the same, the body tags, etc., but the place that the creativity shows is in the choices like names, structures, and stuff, and those are the things that were copied. ECF No. 310 at 146:9-21 (emphasis added). She did not identify which “structures” were unique or creative, nor did she explain why they were unique or creative. Finally, Ms. Miracle was asked, “You also testified that you believe the HTML source code is original. What do you base that on?” ECF No. 311 at 43:17-18. She responded, “I base that on the HTML as a document, and the names and the content and the values in the document.” Id. at 43:19-20. She did not reference the arrangement of the variables or the parameter blocks. In addition to her testimony, Ms. Miracle’s expert report was admitted into evidence. I give no weight to her expert report. That report primarily sought to issue legal opinions. See, e.g., ECF No. 304-14 at 12 (“Compulife’s software is protected by copyright. The arrangement and selection of the compilation of information concerning the term life insurance market, term life products, and term life rates used in NAAIP’s software is also protected by copyright.”). I declined to consider these opinions. ECF No. 314 at 28 n.21, (I “disregard Ms. Miracle’s opinions regarding the creativity and originality of Compulife’s code because these constitute legal — 28 — conclusions.”); Compulife Appeal Brief at *48 n.13 (conceding that I “correctly disregarded Nancy Miracle’s legal conclusions.”). The portion of Ms. Miracle’s report that addressed the HTML code discussed only the selection of the parameter names and values. It did not address the arrangement or structure of the overall HTML code: The copyright for the HTML source code of 2010 specifies the parameters that are used to format a request for information from the Compulife database. Each parameter has a distinct name and distinct values associated with that distinct name. The combination of the names of all the parameters is unique, as are the data values. The names are purely arbitrary and not dictated by any external factor. ECF No. 304-14 at 17-18 (using the SortOverride1, State, HealthClass, and NewCategory parameters as examples). It is the law of the case that Defendants did not infringe on protected parameter names. Ms. Miracle’s report is not germane to whether Defendants legally copied the arrangement of the parameter blocks. Compulife’s proposed findings of fact did not cite anywhere in the record that Ms. Miracle “opined that the 2010 HTML code contained numerous creative elements including a creative way to identify and organize variables related to requests for insurance quotation information.” As discussed above, Ms. Miracle provided only conclusory factually-unsupported testimony that the arrangement (as compared to the naming convention) of the variable blocks was creative. In sum, as the finder of fact, I reject the argument that Ms. Miracle’s testimony provides evidentiary support for a finding that the arrangement of Compulife’s — 29 — HTML source code was expressive or original. I give her testimony no weight in analyzing whether the arrangement of the HTML source code is protected. VII. CONCLUSIONS OF LAW Compulife says I should conclude as a matter of law that the arrangement of the parameter blocks is protectable because: (1) Mr. Bruner did not copy it from anywhere else, (2) it was not dictated by industry standards, (3) the parameter blocks had to be in a specific order to allow communication with the search engine, (4) the parameter blocks could have been arranged many different ways, (5) Ms. Miracle said the arrangement was creative and original, and (6) Defendants did not offer evidence showing a lack of originality or creativity. The evidence at trial does not support this legal conclusion.9 A. Abstraction The Eleventh Circuit said I should have “identified the entire arrangement of the[] variables in the code as a constituent component of the code [then] look[ed] at 9 There is an inconsistency between Compulife’s arguments that (1) the parameter blocks had to be in a particular order so they could communicate with the search engine and (2) the parameter blocks could have been arranged 39 million different ways. If the parameter blocks in the HTML source code had to be in a particular order to communicate with (that is, be compatible with) the search engine, then that order was dictated by the C++ code; if so, there was no choice in how to arrange them, so their arrangement arguably was not creative or original. See Bateman, 79 F.3d at 1547 . In contrast, as discussed more fully below, if the parameter blocks could have been arranged in any order, there is no credible evidence that Mr. Bruner made a creative choice. I need not resolve this inconsistency because I find as a matter of fact that the parameter blocks were not required to be in a particular order. — 30 — the arrangement of all the variables together.” Compulife II at 1158. I interpret the reference to the “variables in the code” to mean the parameter blocks. For the software to function, it needed to display dropdown menus and radio buttons to the user. Each of these functional elements was generated by a block of code bounded by <tr> and </tr> commands. It is these blocks of code, not merely the parameters themselves, that could be re-arranged. Therefore, the parameter blocks are the relevant “constituent parts” of the code for abstraction purposes. The parameter blocks were arranged as follows: Compulife Defendants State ModeUsed BirthMonth SortOverride1 Birthday BirthYear Sex State Smoker BirthMonth Birthday BirthYear Health Sex NewCategory Smoker ModeUsed Health SortOverride1 NewCategory FaceAmount FaceAmount The Compulife HTML source code was divided into three subgroups of parameter blocks: (1) user demographic information, (2) information about available insurance products, (3) hidden parameters. The hidden parameters were in the middle of the group related to the insurance products. — 31 — The Defendants’ code had the hidden parameters first, then the demographic data, then the insurance product data. Compulife’s code also had comment blocks interspersed between some of the parameter blocks whereas Defendants’ code did not. Filtration/Factual Copying It is the law of the case that Defendants factually copied blocks of Compulife’s HTML code. Neither the Eleventh Circuit nor this Court has addressed whether there was factual copying of the arrangement of those blocks. I find that Defendants factually copied the following arrangements of parameter blocks: • {State, BirthMonth, Birthday, BirthYear, Sex, Health, Smoker, Health, NewCategory} grouped together; • {ModeUsed, SortOverride1} grouped together; • {FaceValue} as the last parameter block. Defendants did not factually copy the arrangement as a whole because they put the {ModeUsed, SortOverride1} parameters first in the overall arrangement, whereas Compulife put these parameter blocks in the middle of the overall arrangement. So, there was factual copying of parts of the arrangement, but not the entire arrangement. Filtration/Legal Copying/Creativity There was no legal copying. A preponderance of the evidence does not show that the organization and arrangement of the HTML source code rises to the level of protectable expression. Compulife has not shown “the existence of ... intellectual — 32 — production, of thought, and conception” (supra, at 7) in the order or arrangement of the parameter blocks. Rather, the evidence is consistent with Mr. Bruner having organized the parameter blocks essentially at random. As an initial matter, Compulife argues that BirthMonth, Birthday, and BirthYear are three separate blocks of code. Because I define “parameter block” to be the lines of code bounded by a pair of <tr> and </tr> tags, I treat the code associated with these three variables as a single parameter block. That distinction is not material. As discussed below, there are only six ways to organize these three variables.10 And, the ordering of {Month, Day, Year} for someone’s date of birth is hardly original. So, even if they are separate parameter blocks, their arrangement is not protectable. 1. The Direct Evidence It is undisputed that (1) the parameter blocks could have been arranged in more than one way, (2) the arrangement of the blocks did not affect the functionality of the software, (3) Mr. Bruner selected this arrangement from among the many possible arrangements, (4) Mr. Bruner did not copy the arrangement of the blocks from anyone else, and (5) the arrangement is not dictated by industry standards. 10 The six possible arrangements are: {BirthMonth, Birthday, BirthYear} {BirthMonth, BirthYear, Birthday} {Birthday, BirthMonth, BirthYear} {Birthday, BirthYear, BirthMonth} {BirthYear, BirthMonth, Birthday} {BirthYear, Birthday, BirthMonth} — 33 — These facts do not, by themselves, establish the requisite originality or creativity for an arrangement to be protectable. See BellSouth, 999 F.2d at 1443. Compulife cites BUC Int’l as support for a finding that the arrangement of the parameter blocks was original. In BUC Int’l, the parties disputed whether the plaintiff’s online yacht listing directory was an original work of authorship. The plaintiff’s principal testified “that he evaluated what he thought was the most critical information for the [yacht] brokers and structured BUC’s compilation based on those assessments.” 489 F.3d at 1144 . And, a yacht broker with knowledge of the relevant marketplace testified that the plaintiff’s directory was arranged differently from the other information available to brokers and was therefore original. Id. The Eleventh Circuit held that this direct testimony was sufficient for the jury to find originality of authorship. Id. at 1145 . Here, there is a lack of credible evidence that the arrangement of the parameter blocks was sufficiently creative to satisfy the constitutional standard for copyright. There is no direct testimony about how or why Mr. Bruner chose to arrange the parameter blocks in the way that he did. For example, there was no testimony that Compulife had a reason for asking users for their state of residence before asking for their gender. Maybe people are more likely to complete the entire form if they are “eased in” with less personal questions? No evidence of that. Or, maybe users report a more pleasant customer experience if they are asked for demographic information before being asked what kind of insurance they want? No evidence of that. Maybe the HTML source code runs faster, more efficiently, or more securely based on the — 34 — ordering of the parameter blocks? No evidence of that. In fact, there was no direct evidence of any intellectual production, thought, conception, or creativity for the way Mr. Bruner arranged the parameter blocks.11 2. The Circumstantial Evidence Sometimes order matters. Mathematically, whether order matters affects the number of possible outcomes. Combinatorics is the branch of mathematics that involves calculating the number of unique permutations of a set. That number depends on whether the order of the permutation matters. For example, in the game of blackjack, getting dealt an ace then a ten is the same as getting dealt a ten and then an ace. In set theoretic terms {Ace, Ten} = {Ten, Ace}. Likewise, in a Pick-5 lottery, the selection order of the balls does not matter: {1, 3, 5, 10, 17} = {3, 10, 5, 1, 17}. On the other hand, when walking in Manhattan, the order of turns matters; turning {left, left, right} is different from turning {right, left, left}. More generally, for a set of n numbers, if order matters, there are n choices for the first element of the permutations, n-1 for the second, n-2 for the third, and so on. The total number of unique permutations is “n-factorial” (written “n!”) which is the product of all numbers from 1 to n. So, for the set of three elements {A,B,C} or 11 I do not suggest that a subjective intent to create an original work is enough. Ultimately, the work must be objectively “original” to warrant copyright protection. — 35 — {BirthMonth, Birthday, BirthYear}, if order matters, there are six unique permutations (3 x 2 x 1): {ABC, ACB, BAC, BCA, CAB, CBA} In contrast, if order does not matter, there is one permutation of the set of three elements. ABC is the same as ACB, BAC, BCA, CAB, and CBA. Just like in blackjack or the lottery. Put differently, there is only one way to select three items from a set of three.12 What conclusions can we draw? When order matters, one can choose from among multiple unique arrangements. So, there is a natural implied circumstantial inference that the order was chosen for a reason that may be creative and original. 12 As this explanation shows, Compulife’s assertion that there are 39 million possible permutations of the 11 variables in the HTML source code is misleading. There are 39 million unique permutations only if it matters what order the parameters occur in the HTML source code. If order matters, there are 11! = 39,916,800 possible permutations. But, here the order does not matter. So, there is 1 unique permutation of the HTML variables. In combinatorics, when order does not matter, the way to calculate the number of ways to select a subset of k items from a set of n items is “n choose k,” represented as 𝑛𝑛 𝑛𝑛! � � = For example, if order does not matt𝑘𝑘er, th𝑘𝑘e!r(e𝑛𝑛 a−re𝑘𝑘 )s!ix ways to select two items from a set of four objects. That is, . Those subsets are: {1,2}, {1,3}, {1,4}, 4! 24 — 4 — {2,3}, {2,4}, {3,4}. And, ther�e a�r=e 2!(4−2!) = 4 = 6 — 2 — — 52 — � � = 2,598,960 possible hands in 5 card draw poker. Here, because the order of the pa5rameter blocks does not matter, there is only one way to arrange the HTML variables: 11 choose 11 equals 1. — 36 — In other words, if A is different from B and someone chooses A, there is an implied inference that A was chosen for a reason. Accord Key Publications, Inc. v. Chinatown Today Publishing Enters., Inc., 945 F.2d 509 , 513 (2d Cir. 1991) (“[s]election implies the exercise of judgment in choosing which facts from a given body of data to include in a compilation”) cited in Warren Pub., Inc. v. Microdos Data Corp., 115 F.3d 1509, 1518 (11th Cir. 1997). Hamlet’s famous “to be or not to be” speech makes no sense if you re-order the words into “to to be be not or.” It is therefore reasonable to infer that Shakespeare chose the order of the words for an original, creative reason. No such natural inference arises when order does not matter. If A and B are the same, there is no inherent reason to pick one instead of the other. Put differently, when all permutations are the same, there are no unique alternatives to choose from. So, the person can get the identical outcome by choosing any one of the multiple permutations. In the absence of other evidence, there is no inference that this choice was creative or original. Applying this concept to the current case, order does not matter in the arrangement of the HTML source code’s parameter blocks, so there is no natural circumstantial inference that Mr. Bruner made an original, creative choice about how to arrange them. Whether order matters can also depend on whether a particular process is simultaneous or sequential. In a sequential process, by definition, events must happen in a particular order. For example, in making Cuban coffee, one whips the espuma before pouring in the espresso. https://www.homegrounds.co/how-to-make37 cuban-coffee (last visited December 19, 2024). But, in making cappuccino the foamed milk is poured on top of the espresso. https://athome.starbucks.com/recipe/classiccappuccino (last visited December 19, 2024). Similarly, when painting, primer goes on before the topcoat. Because in every sequential process it matters what order things happen, there is a circumstantial inference that a person arranges the steps in the correct order so that the process works properly.13 The Compulife HTML source code as a whole is arranged to collect the user information simultaneously and then pass it to the search engine simultaneously. This structure does not circumstantially imply that Mr. Bruner made a creative or original decision how to arrange the parameter blocks. This same analysis applies to whether the arrangement of blocks of computer code affects the functioning of the computer program. When arrangement affects functionality, it implies that the programmer engaged in an act of “intellectual production, of thought, and conception.” Otherwise, the software would not work. For Compulife’s HTML source code, the order of the parameter blocks does not affect the code’s functionality, so there is no inference that Mr. Bruner made a creative or expressive choice.14 13 Of course, the chosen sequence nevertheless may not to be copyrightable if it is dictated by other factors or if there are a sufficiently limited number of ways to sequence the steps. See supra n. 9. 14 To be clear, I understand that function cannot be copyrighted. I consider functionality only insofar as it creates a circumstantial inference of creativity and originality. — 38 — 3. Defendants’ Burden Finally, Compulife argues that Defendants failed to meet their burden of offering affirmative evidence on the issue of legal copying. But, as the Eleventh Circuit explained in Compulife I, a party can satisfy its burden of proving unprotectability even if it does not introduce affirmative evidence. 959 F.3d at 1306 n.8. Compulife’s argument also is contrary to the correct understanding of the role of an evidentiary burden. To say that a party has the burden of proof means that, looking at the totality of the evidence, a preponderance of that evidence must be in the party’s favor. It is sometimes more accurately described as the “burden of persuasion.” “[T]he burden of persuasion [is] the notion that if the evidence is evenly balanced, the party that bears the burden of persuasion must lose.” Dir., Office of Workers' Comp. Programs, Dep't of Labor v. Greenwich Collieries, 512 U.S. 267, 272 , 114 S. Ct. 2251 , 129 L. Ed. 2d 221 (1994). Where the evidence is not evenly balanced, the party whose side is supported by a preponderance of that evidence wins. That situation can be described as one party or the other “meeting” or “not meeting” its burden of proof, but the relevant question is whether the matter at issue is, or is not, proven by a preponderance of the evidence. The distinct, but related, concept of the “burden of production” refers to “a party’s obligation to come forward with evidence to support its claim.” Id. Unless it has a burden of production, a party does not need to introduce affirmative evidence to sustain its burden of persuasion. It can meet that burden by showing that the — 39 — evidence in the record (whoever introduced it) establishes (or fails to establish) a matter at issue. For example, the defendant bears the burden of proof on the affirmative defense of statute of limitations. If the plaintiff fails to introduce any evidence of conduct within the statutory period, the defendant wins, even if the defendant has not offered any evidence of its own. Or, as the Compulife I Court explained, “[N]o evidence would be necessary to convince a court that alphabetization is an entirely unoriginal method of arranging data and thus unprotectable as a structural element of a work.” 959 F.3d at 1306 n.8. Here, there is no credible evidence, direct or circumstantial, that the arrangement of the HTML source code is original. Therefore, Defendants were not required to introduce contrary evidence to show unprotectability. VIII. CONCLUSION The Copyright Act defines the term compilation as “a work formed by the collection and assembling of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship.” 17 U.S.C. § 101 (emphasis added). Compulife has not shown that the arrangement of its HTML source code as a whole is an original work of creativity or authorship. So, it has failed to meet its burden of showing legal copying. By separate Order, final judgment will be entered in favor of Defendants on the direct and contributory copyright infringement claims (Counts I and II in the ’08 case and Counts II and III in the ’42 case). — 40 — DONE and ORDERED in Chambers at West Palm Beach, Palm Beach County, in the Southern District of Florida, this 10th day of January 2025. BRUCE E. REINHART UNITED STATES MAGISTRATE JUDGE Al