Wow. In Belvac Production Machinery, Inc. v. Adonis Acquisition Holdings, LLC, C.A. No. 25-166-JLH (D. Del.), the plaintiff moved to compel the defendant to preserve all random access memory ("RAM") copies of certain copyrighted software that is apparently used to operate equipment.
To me, this is a wild motion—although it takes a bit of knowledge about computer memory to show why.
Some Background on Computer Memory
Unlike Nate, who wisely avoided expounding on analytical chemistry, I feel like some background on how computer memory works is needed here, to help understand just what plaintiff is asking for. The parties didn't include this in their briefing—but they probably should have explained things a bit more.
RAM is essentially a computer's short-term memory. Typically, when a computer is sitting without power, its software is stored on longer-term storage such as flash memory or a hard drive. When you turn the computer on, let it boot up, and use it, portions of various software are copied into RAM and executed.
Of course, the software is not just copied to RAM. There are also multiple layers of cache memory in a computer that hold copies of portions of software to be executed, including for example short-term cache memory on the disk and multiple levels of cache on the CPU itself.
And even the portions of software that are copied into RAM may soon be temporarily offloaded from RAM back to a disk or other storage as "virtual" memory. As it transits back and forth, the material in RAM can become fragmented over time—i.e., it may well look like a jumbled mess.
Memory on modern computers is also protected. If everything is working properly, the program you use to send e-mail can't see the password you put in to access your bank account. It's not easy for a program to circumvent that protection. And if it did, the resulting memory dump could contain all kind of things that were in use on the computer, potentially including things like passwords and cryptographic keys.
Finally, the memory dump will likely be next to useless to anyone but an expert. It's not like you can open it up and read it in Microsoft Word.
Plaintiff's Request to Retain Records of RAM
All of that is to say, this is a big request from the plaintiff. So, what is going on here, exactly?
According to plaintiff's brief, it owns the copyright on some software used to operate some equipment. D.I. 68 at 1. It alleges that, when the equipment and a related computer are used, copies of its software are made in RAM—causing copyright infringement. Id. at 1, 3.
The defendant denies that the copying occurs or, in the alternative, argues that it is transitory. Id. at 1. They also say that they have not turned the equipment on or off. Id. at 1, 3-4.
What Plaintiff actually wants is not entirely clear from its motion, but it seems to involve "real-time retention" of all data in RAM on both the machines and the related PCs, whenever the software is used. To me, "real time" means they don't want just a snapshot, but more like a running log of every single thing that goes in and out of memory.
They call this "a limited set of temporary ESI, narrowly tailored in scope to accomplish this necessary objective." Id. at 4. Uh-huh.
Defendant's response was likewise hard to parse. At times they seem to indicate they may agree that plaintiff should get this relief or a subset of it. D.I. 69 at 3 ("The starting point for any preservation provision should, therefore, be no broader than requiring preservation of . . . RAM data . . . ."). At other times they seemed to push back more. Id. at 4. Neither side, though, seems to have a full understanding of how crazy this request is.
The Court Denies RAM Discovery—For Now
The Court ably handled the disjointed briefing by denying the request. Judge Fallon noted that it appeared the parties had not adequately met-and-conferred:
Defendant's response convinces the court that the parties have not fully engaged in a meet and confer that included any exploration of compromise options. Defendant opposes any preservation requirement, but acknowledges that if the court finds good cause for preservation, the "starting point" for the protocol ["]should ... be no broader than requiring preservation of: RAM data["]
Belvac, C.A. No. 25-166-JLH, at 3. In denying the motion, the Court further noted the vague language used by the parties:
Plaintiff's motion for the entry of its proposed ESI protocol is DENIED without prejudice. Plaintiff has not met its burden to show good cause for the preservation of ESI as set forth in its proposed protocol. Neither the proposed protocol nor Plaintiff's letter submission defines phrases such as "real-time retention" or "known-to-Adonis scenarios," rendering the scope of the information to be preserved too broad and vague. Furthermore, the parties have not met and conferred on Defendant's compromise proposal.
Id. at 4. It ordered the parties to engage in another meet-and-confer, including about the potential compromise and the costs of preservation:
Therefore, IT IS ORDERED that the parties shall engage in a meet and confer on or before August 8, 2025 with the goal of reaching a compromise consistent with Defendant's proposal. . . . In addition to discussing the scope of Defendant's ESI preservation obligations, the parties shall also meet and confer with respect to the reasonable anticipated costs for preservation of the ESI. . . .
If no compromise is reached during the meet and confer process, any renewal of the instant motion by Plaintiff should include in the briefing: (1) analogous case authorities, to the extent there are any, regarding the preservation of evidence of the alleged reproduction of copyrighted software; and (2) precise and narrowly contoured language defining the scope of the ESI which must be preserved, with each side's competing proposal clearly identified for any portions of the proposed protocol that remain in dispute
Id. at 4-5.
I'll be very curious to see if this dispute comes back to the Court. To me, it seems like an extremely burdensome and cumulative request that could be satisfied by inspection of the relevant devices and simple forensic images of the disks, rather than by forging new ground on memory capture technology.
If you enjoyed this post, consider subscribing to receive free e-mail updates about new posts.
