by on June 2, 2026
In this philosophical essay, I explore the question: “When (if at all) is it ethically and morally acceptable to use proprietary software in the production and/or improvement of urgently needed copylefted FOSS?”
The question presents a complex conundrum. I attempt herein to rigoriously examine it through both a priori ethical analysis and a posteriori (and folksy) consideration of my personal experience and the shared experiences of the early software freedom movement.
I surprised myself at the outcome of my analysis. I conclude that under some circumstances (of which we have already witnessed in key historical examples), use of proprietary software by FOSS contributors to create/improve FOSS becomes a moral imperative. And, that imperitive often supersedes the moral imperative to avoid using that proprietary software.
I grew up lower middle class in the USA — which is quite privileged by global standards. As such, I never went hungry, but most meals did not include second helpings. My family were early adopters of “extreme couponing”. My earliest childhood memories are climbing through (on Monday mornings) the gigantic bin of recycled Sunday newspapers behind the grocery store. My job in this endeavor was “Sunday insert extraction”. Back then, more than half of all households received the Sunday paper. That Sunday insert was the goldmine. The insert paid for the paper subscription (and much more) through its colorful 30-40-page advertisements filled with coupons. Get your hands on 10–20 more of those inserts freely from the recycle bin, and you could get “Extreeeme!”.
The late 1970s and early 1980s were the wild west of extreme couponing. There were nearly no restrictions on how many identical coupons one shopper could use. A shopper could use 50¢-off coupons for 49¢ trial-sized products. I remember once my mother filled the entire checkout belt with 30+ trial size dishwashing liquids — for which my mother paid ≈65¢ (i.e., just the sales tax).
My young indoctrination to extreme couponing now feels as if it’s part of my DNA. If you’re ever on on mute during a voice call with me and I don’t reply quickly, it might be because I’m currently standing in a grocery store aisle — comparing value per volume of different sizes and cross-referencing that with a coupon’s fine print.
I actually miss having my hands covered in newspaper ink every Monday — as that analog way was superior to digital. For much of my life, I enjoyed “extreme couponing” as one of the few activities that put aside my worries about the future of users’ software freedom, rights, and privacy. Sadly, about ten years ago, everything started to change. Now, using any coupon is a constant battle with proprietary, bait-and-switch spyware.
Major grocery chains in the USA provide proprietary apps — now the primary place providing coupons. Most chains have parallel websites that allow account holders to “clip” them digitally. Printed ads still exist, but are rife with phrases like “Digital Deal only”. Shoppers now collide in the grocery aisles while navigating the Kafkaesque apps and websites to figure out how to (virtually) clip a good coupon.
The once relaxing process of watching television on a Sunday afternoon while clipping paper coupons slowly morphed into a two-hour ordeal of figuring out the minimum proprietary Javascript required, then building a text cross-list noting the discounts, screenshotting the bigger discounts, and then using the list and screenshots to prove to the grocery manager that I really was supposed to get $1-off of 4+ avocados this week. And, yes, I clipped it in the web browser. And, no, I can’t install the app on this device to “check it in the app”. 🙄
Most of my life, I couponed due to absolute financial necessity. I can fortunately now afford to pay full price, but frugality is part of my moral code. While the correct moral action for these grocery chains is to simply reduce prices for all shoppers automatically, they behave unethically. They prefer to use proprietary software to track you, to attempt to manipulate you into buying items you don’t need, and all the other obvious anti-features. In response to their unethical affront to the public, I feel the moral imperative to exploit that system by not falling for the manipulation tactics, maximizing my grocery budget, and acting as a watchdog when the digital coupons just seem to “not work”.
My colleague Karen Sandler and I have keynoted three times regarding our own ethical and moral analysis of software freedom in today’s digitized world. Our primary thesis: one must constantly compromise some software freedom purity to merely participate as a normal citizen. My lifelong grocery shopping experience exemplifies one situation of hundreds where tasks that once included no software interaction now mandate proprietary software usage. I lament this reality; it annoys me. But annoyance isn’t immorality, and software freedom activism was founded on pragmatic idealism. I fear that recent hard-core FOSS rhetoric has forgotten what that means.
As a fan of Kantian moral frameworks, my formulation of pragmatic
idealism focuses on ranking the competing moral imperatives. While “use
no proprietary software myself” is a noble moral tenet, its value
remains paltry when viewed comparatively.
The moral imperative to avoid writing proprietary software
towers over the moral imperative to avoid using it.
The conclusion is similar in a Utilitarian ethical analysis: provided one does not encourage others to join them, using proprietary software harms only the self. Writing proprietary software harms everyone who ever uses it — perhaps into the distant future.
The reverse situation yields similar analysis. Often, an individual, by using proprietary software today, can assure that many others will ultimately use less proprietary software in the future. In such situations, use of proprietary software by the individual is clearly a higher-ranked moral imperative (because it prevents many others from facing similar moral dilemmas). This also serves the greatest good for the greatest number: one person uses proprietary software so that many others won’t.
SFC lives this truth every day we operate. As part of SFC’s fiscal sponsorship of our member projects, we use a lot of annoying proprietary and/or trade-secret software (including banking systems, corporate donation and purchase order systems, third-party donation services, etc.). Yet, if SFC didn’t use those systems on our member projects’ behalf, at least one individual (possibly many) from each project would use them, instead.
Application of pragmatic idealism shows that using proprietary software is occasionally the right ethical choice for at least three reasons: (a) serving a higher moral imperative outside the scope of FOSS, (b) limiting the number of people who use proprietary software (now or in the future), and (c) increasing the efficiency and/or efficacy of urgently needed FOSS advancement.
While my opening example of using proprietary software to increase efficiency for FOSS’ advancement focused on the non-FOSS imperative of frugality, the principle expands well beyond that. In fact, the earliest FOSS contributors clearly believed that other imperatives sometimes justify proprietary software use.
By the time Richard M. Stallman (RMS)1 began to formulate software freedom as an essential right in the mid-1980s, proprietary software was the norm. Most computer users were using a Microsoft, Apple, or proprietary Unix system. RMS chose to write a copylefted Unix-like system (GNU) to replace proprietary Unix on its existing hardware.
But consider how RMS and the other early GNU contributors approached that work. If the goal was to minimize the developers’ use of proprietary software, the obvious approach would have been as follows: write a small bootable kernel on the existing ITS computers for the new hardware. Then, one could painstakingly write the basics of a C library in assembler (because, no FOSS cross-compilers existed in those days). After probably a decade of fulltime work by 3–5 people, there might have then been enough of a C library, a C compiler, a shell, and a text editor that others might also have software freedom.
I’m quite glad that the early GNU contributors did not take this approach. In fact, RMS started by writing the key program he knew best how to write quickly: another Emacs implementation — designed for existing proprietary Unix systems. And, GNU Emacs was working nicely in less than one year!2 Next, the GNU contributors tackled GCC — but not by bootstrapping in assembler. Instead, they used existing, proprietary C compilers to bootstrap. That too brought software freedom to C programming much more quickly than the “pure” approach.
Nearly all readers already know how the story ends: these early contributors quickly brought most of a working POSIX system to users — all licensed under copyleft licenses — except for the kernel. That story is often told, yet so rarely is the most important point the focus: the entire FOSS community used part-FOSS/part-proprietary systems for at least 15 years. Our (correct) focus was to deliver as much software freedom to other users as we could — as quickly as possible. Our overarching moral imperative was “create as much copylefted software as possible” — even if we used proprietary software as part of FOSS’ production.
The users’ software freedom matters much more than the developers’ software freedom. If FOSS developers find a path that accelerates software freedom for their users, then that higher moral imperative demands that path.
We dream of a world without proprietary software — where every byte of code on every piece of hardware provides complete, Corresponding Source to its users. Despite our best efforts, we drift further from that world every day. We race against the wealthy people whose wealth grows faster when they produce more proprietary software. We will not beat them in our lifetimes, but we can make progress. Along the way, we must constantly reconsider: Should we use some proprietary software to accelerate urgent improvements to FOSS? Willingness to sometimes answer that with “yes” (pursuant to rigorious criteria) epitomizes pragmatic idealism!
Every day, I find that I’m using more proprietary software than I did in the early 2000s. I’ve been on that unfortunate trajectory since the mid-2010s. I was one of the last hold-outs who using only X200 and T500 laptops — the last laptops ever made that could run all FOSS code from the up-top to the down-deep. Nevertheless, in November 2025, I switched to a Novacustom V54 that requires so many blobs for the internal devices that 77.5% of the packages contained in the firmware distribution are proprietary.
Today, my personal software freedom on my own laptop (at boot-time)
is measurably 77.5% less than it was when I lugged around the
T500. However, everything I do for SFC now happens quicker. I spend
substantially less time waiting for beancount to load SFC’s
books. I can attach three additional external displays (instead of just
one). I stopped injuring myself every time I travel from the T500’s
weight (as my X200 became too slow many years before). These and so many
other efficiency improvements from my Novacustom laptop made my work at
SFC (at least) 50% more efficient.
I gladly take that trade-off all day long. My job is to preserve, protect, and promote users’ software right to repair. That moral imperative outranks my moral squeamishness every time the blob on my wireless card processes a network packet.
Frankly, the software freedom and rights of computer users of the future deserve priority over activists’ and developers’ preference to use only FOSS. We should reluctantly but doggedly use any proprietary software that assuredly accelerates creation of and improvements to essential and urgently-needed copylefted FOSS.
1 RMS deserves substantial credit for formulating the first ethical framework for software freedom, his invention of copyleft, and his authorship of the earliest GNU programs. I am however not comfortable mentioning RMS and his work without noting his bad behavior that I frequently witnessed and his ongoing refusals to curtail or apologize for that behavior.
2 Free Software, Free Society: Selected Essays of Richard M. Stallman, 3ʳᵈ edition (2015). GNU Press, Boston, MA. Page 21.
Please email any comments on this entry to info@sfconservancy.org.