At 2026-11-08T02:17:17-0600, Rob Landley wrote: > On 11/6/26 22:32, G. Branden Robinson wrote: > > > > I can't repro the file1 source file being moved on any upstream > > > > version. Nor can I repro on mv 9.7 from trixie ? > > > > > > Likewise, it does not reproduce for me either on both. The code > > > looks fine for it as well. > > > > > > I hope we get a real reproducer, or that Rob reconsiders his > > > quickness to make statements like this [1]: > > > > As a leopard doesn't change its spots, a tech bro does not change > > his thoughts. > > I ignored the https://wondermark.com/c/1062/ who jumped into a > mastodon thread
I don't use Mastodon, so this must refer to someone else. > he wasn't part of and then followed to a mailing list to tell me I'm > not worthy to send a bug report about a regression that broke existing > tests, but if you feel dogpiling is a good use of your time... ...because "dogpiling" is when people who don't know each other independently criticize of your mode of engagement. Moreover, you are here pulling a classic tech bro move: conflating criticism of your mixture of derogatory statements about GNU software in general with your defect report as a critique of your person. Whether such conflation is warranted is up to you: if you decide that you'll never, ever report a defect to a GNU project without taking a rhetorical shit on that project and everybody who contributes their volunteer labor to it--well, then, yeah, such a criticism _does_ become one of your person, for you to do any other would be to change who you are. Let the readers know! > In > https://blog.geekitude.com/2020/09/20/linucon-2005-pros-and-cons-of-the-gnu-general-public-license/ > I was the pro-gpl panelist, a full decade before I navigated 0BSD > through the SPDX approval process and spent weeks getting it into > github's choose-a-license pulldown. Okay. I don't see how that's relevant. Should the reader be impressed that you sat a conference panel with Eric Raymond and his spouse? I don't think whether you were pro-GPL in 2005, 2020, or 2026 is relevant here. > Just an example of something I changed my mind about in response to > new information. _That_ you changed your mind is easy to verify. The _reason_ you changed it is necessarily opaque. Sometimes it's easier to assimilate oneself into the culture of Cupertino, Sunnyvale, or Mountain View. > My opinion about the FSF has also changed over the years. I drove > cross country to Boston to meet Richard Stallman in 2001, and remember > I made an analogy to the movie 1776 that made him so angry he ran out > of the room. (I saw his role in the community as John Adams rather > than Jefferson or Franklin.) Stallman is not coterminous with the FSF, nor with the GNU Project. Maybe that fact is one that he struggles to accept himself. Nevertheless, plenty of other people have contributed not only labor but vision--"leadership", if you will--and illustrated that there are ways to pursue the missions of the organization, including a robust defense and promotion of copyleft licensing, without carrying along the founder's preferences on immaterial matters, like his selection of text editor or favored genre of music. For example, several groups of developers have bucked the FSF's copyright assignment policy, and eventually, someone (I forget who) revised the GNU Maintainer's Guide to drop the implication that such assignment is always and everywhere mandatory. (A claim that was never 100% true in the first place, as glibc long and openly incorporated materials copyrighted by others, and there is also the semi-famous story about the Japanese government contributing some sort of localization support to GNU Emacs, and as a sovereign nation-state, claiming inability under international treaties to alienate its own copyright. Whether such inability is actually true, I don't know, but can imagine that being a counterparty to, in principle, representatives of the Japanese emperor made Stallman feel important.) A bench of non-Stallman leadership talent is an essential thing for the FSF and GNU to possess if these organizations are to survive him. And thus we see how maintenance of the pretense that the FSF and GNU are mere appurtenances to the mind and body of Richard Stallman necessitates the erasure of contributions large and small by numerous other people over the past 42 years. Such erasure brings us to... > After a lot more study of computer history I now believe his main > skill was taking credit for the work of others > (https://landley.net/notes-2010.html#19-07-2010 was 15 years ago, This piece manifests poor reasoning straight out of the gate, with its first paragraph. Like all human activity that affects anyone else, engineering is inherently political. (A pig farmer might decide, "haw, I'mma just rig up this here wood'n sluice to drain awl muh hawg waste offa my property and down to the Joneses' crick, haw haw haw.") I reject your theory that the Hurd project failed to acquire a large number of users because Stallman was politically-minded or promulgated an ideology. That theory is inconsistent with your acknowledgment of the success of GNU Emacs, GCC, and the (L)GPL family of licenses. If Stallman's trumpet-blowing crippled the Hurd project, why didn't it cripple the others? I think the Hurd failed for simpler reasons: Mach had a bad reputation because its working set was too big for memory caches of the 1980s, an empirical fact that Linus Torvalds and his many disciples inflated into an eternal, invariant property of microkernels, ignoring the work of Liedtke and others. Making Linux a kernel that ensured that its own size reliably outpaced the rapidly growing caches of subsequent decades also helped defend the principle that There Was No Alternative. Nothing produces "market success" more reliably than market failure.[I] But also, Hurd broke with the usual GNU project pattern, exhibited by GCC, glibc, {file,text,sh-}utils (later coreutils), groff, and others, of taking existing, manifest Unix software and cloning it with enhancements (such as limit removal). By implementing its own concepts and models of data flow in ways that only somewhat resembled familiar Unix ones, Hurd changed the way that systems software developers and administrators interacted with the kernel. As a rule, engineers love changing only their own software. They want everything they interface with to stand stock still, except when they encounter a bug that personally affects them, at which point the narcissists among them get apoplectically angry and demand instant fixes. There _is_ a way to disrupt engineers' accustomed habits dramatically with a redesign of the system layer and compel adaptation: I give you systemd. Contemplate the resources that have been invested in such compulsion, the coupling involved in IBM/Red Hat's simultaneous provision of GNOME and bundled distribution products, and the fractures created in the community by the same. One could say that Raymond used his glib essays as a fife, bravely leading the rats out of the GNU "cathedral", abandoning them in the village bazaar to contemplate what he'd do with his multi-million-dollar windfall,[E] as IBM steadily enclosed the bazaar in a new cathedral. Thus do we enjoy the true freedom that only having a new master can bring. And you helped![F] > before Gosling said > https://www.reddit.com/r/programming/comments/dhrcxw/james_gosling_on_how_richard_stallman_stole_his/) > and that his larger project has probably been a net negative. As I noted before, defending the work of the GNU Project and burnishing Richard Stallman's legacy are, and must be, two different things. But that statement is incompatible with the worldview you're professing. > But then so > was Windows, and sometimes you still have to engage with it, ala > https://knowyourmeme.com/photos/1259257-we-should-improve-society-somewhat . That's true, which is why I am counseling you with respect to how to better engage with the coreutils developers--which I'm doing entirely on my own initiative, as an admirer of the work Pádraig and Collin are doing. They have not asked me to do so and I'm not entirely sure they find my intervention helpful. But I dislike bullying. I think bullies should be opposed at every opportunity. Thus, I oppose you. > I'm surprised that you're surprised I'm not a fan given > https://lwn.net/Articles/202106/ and all, I make no appearance in that article or the attached comments. I don't recall having or expressing any opinion on whether Busybox stuck with GPLv2 or migrated to GPLv3. Perhaps you can aid my memory. > but I guess I'm not surprised that you can only work with people who > like you and like your project. Leaving aside projects arising from employment, and one for which I have administrative responsibilities, I've contributed to, among others: * ncurses (permissively licensed) * procps-ng * GNU findutils * Plan 9 from User Space (permissively licensed) * NetHack (permissively licensed) * Chocolate Doom Those are just the ones I can think of off the top of my head. I don't recall experiencing a conflict with any of their developers, and they represent a 50/50 split between copyleft and BSD-ish licenses. I expect you find some or all of those projects frivolous--two are games. Perhaps we undertake engineering labor for different primary goals: myself, for the pleasure of it; you, because like Linus Torvalds, you want in on the world domination game, and Android's Unixy userspace seemed like a good place to squat. And it is, until they pull an Ubuntu and throw over toybox for uutils. Ain't bundling great? > (My condolences, that must be very rough.) Your crocodile tears are superfluous. That's another example of you manifesting the totalizing worldview of which you accuse Stallman: anyone who disagrees with you must incapable of getting along with anyone who disagrees with _them_. In other words, you're projecting: all the reasonable people in the world agree with you, and those who don't agree with you must be unreasonable. > > > > I'm trying to get all the toybox tests to pass "test_host", but > > > > gnu/crap is a moving target. > > > > > > Hopefully Google doesn't endorse your views. :) > > > > With appropriately structured internal policies, endorsement is > > superfluous. > > This is why I bothered to reply: I've never been a Google employee nor > have I ever claimed to speak for them. They paid me for support for a > couple years (through an intermediary that took 10%) but that ended in > 2023 a few months after the 12k round of layoffs. So I appreciate the > "I'm gonna tell on you to your boss" vibe, but you're a bit late. If you perceive a "I'm gonna tell on you to your boss" vibe from my message, perhaps you should have recruited Gemini's aid in interpreting my mail. Or a disinterested human's. My argument, to be clear, is that your abusive communications regarding the GNU Project are at worst a zero in achieving Google's aims, and at best inculcate a Google-cooperative attitude in young engineers reading your messages. And you're now doing that labor for free, so even better! Do you feel Google compensated you excessively, and you're working off a debt that only you perceive? > > Rob's hostile remarks are at worst a PR problem, and not much of one > > at that, as there is no conceivable future where Google will wish to > > preserve the copyleft commons. > > A PR problem... for Google? Yes--theoretically. But as I immediately noted, that problem is only meaningful in a world where Google wishes to be perceived as a "good citizen" in the world in which it operates. There's plenty of evidence that creating such a perception has no value to its leadership,[A][B] or is adequately in hand by dint of its management of its users' "experience", for instance through personalization of search results, or the mandated provision to children of an AI assistant that reinforces only correct views[C] according to the Board of Directors--more likely, their deputies on staff, as Directors are otherwise occupied bringing down their handicap at Shoreline. > I blogged about potentially replacing my android phone with an iphone > when they locked down the play store back in > https://landley.net/notes-2024.html#05-05-2024 and a quick mbox search > says I first used the phrase "Epstein ballroom" to describe Google's > actions in a message cc'd to both the toybox mailing list and github > back in 2025: > https://lists.landley.net/pipermail/toybox/2025-October/009842.html ...okay. Relevance? You're sometimes critical of Google, therefore your hostility to GNU projects is warranted? I clearly lack your intellect, so I'll need you to make the implied syllogism here explicit. > I'm still not sure how the sealion found the mastodon thread he jumped > into because if he followed me there he'd presumably have noticed a > bit of a theme: > > https://mstdn.jp/@landley/115656919238909490 > https://mstdn.jp/@landley/116178674562572992 > https://mstdn.jp/@landley/116263418959968608 > https://mstdn.jp/@landley/116270515250337867 > https://mstdn.jp/@landley/116576305502700188 > https://mstdn.jp/@landley/116579997705706871 > https://mstdn.jp/@landley/116603227436372619 Again, not clear on the relevance. The "sea lion" wasn't me, as I don't use Mastodon, and it's unclear from your argument what role the "sea lion" plays in this discussion thread, unless...well, unless what's at work here is yet another classic tech bro personality trait. You believe yourself a genius, and therefore all the dunces of the world are united in confederacy against you. > I'm half a year behind on editing and pushing my local notes.txt file > to my web page (I keep forgetting anybody else reads it, it's a lab > notebook and diary) but https://landley.net/notes.html#17-03-2026 is > the most recent "published" entry in that genre (also linking to a > mailing list entry). My most recent public complaint about Google > where the sealion could have seen it was > https://mstdn.jp/@landley/117565679738416023 which motivated me to > switch mkroot off their 8.8.8.8 DNS resolver. Because I am a dunce, I struggle still to discern the relevance. > And yet I still try to interact professionally with developers trying > to have technical discussions about software, I advise you to try harder, specifically when engaging the GNU coreutils developers. > even though I think the company they work for was permanently damaged > by fallout from https://www.wheresyoured.at/the-men-who-killed-google/ > plus all the LLM psychosis and kneeling to maga. I don't rub their > engineers noses in my dissatisfaction with their employer when they > report a bug or send a patch because complaining to them about > something they can't fix isn't useful. No, but you will rub people's noses in something you term "GNU/crap". > Heck, the "AI policy" ... I don't like it, and have clear boundaries, > but a bug report's a bug report. I agree with that much, and have expressed my own annoyance recently with a spate of LLM-generated bug reports against groff.[D] The problem is not that a bug was detected--I'm glad of that--but that, as with an inexperienced human reporter who does a poor job filing a bug, the report made misrepresentations. Each "minimal reproducer" was bloated beyond minimality by a factor of one hundred to one thousand; most analyses were inaccurate; and the fixes proffered were themselves defective. If your `mkroot`-based semi-automated CI tool processes are as non-deterministic as this thread has shown them to be, how are they, operationally, improvements over LLM-generated reports? A foundational element of corporate strategy is asset stripping, a procedure entirely consistent with the lodestars of business leadership as set forth in the Powell Memo[G] and the Friedman Doctrine.[H] LLM-based coding agents are one means of stripping the assets of copyleft projects. But you got there first. Congratulations! > (If you mean it's a PR problem for gnu... did you miss my "rise and > fall of copyleft" talk at > https://archive.org/details/OhioLinuxfest2013 entry 19? I did not mean that it's a PR problem for GNU. And I certainly did miss your talk; I have paid little attention to the progress of your career. > Kirk McKusick was in the audience and said he enjoyed it afterwards. Why would you expect anything but a positive reaction from him? > Again, you seem _surprised_. You know enough to have an opinion, > yet...) Everywhere I disagree with or am critical of you, I must be ignorant or stupid. These are the only possibilities. That black turtleneck looks great on you. > > And his comments have a salutary purpose in transmitting the culture > > to younger generations of engineers in a way that sinks in much > > better than a HR-mandated onboarding training module. > > I honestly have no idea what you mean by that. (No I'm not asking.) I mean that, to Google, you are doing the Lord's work in signaling the virtue of shitting on GNU projects indiscriminately, as with your coinage "GNU/crap". > > > [1] https://lists.landley.net/pipermail/toybox/2026-November/010047.html > > You have to make an effort to read my blog or my project's mailing > list. You have to go out and FIND things to be mad about. _I_ didn't. Collin saved me the trouble. I myself speak freely, and encourage others to do the same, so I won't say you shouldn't. _But_, in an atmosphere of free discourse, one has to be prepared to account for oneself, particularly when one appears to be aggressively indulging biases. In the instant case, human unreliability in manual execution of complex CI procedures combined detrimentally with your categorical regard of GNU coreutils as "GNU/crap", leading you to presume a software defect where really the problem was operator error. That sort of experience _should_ teach humility. I wonder if it will. > > > [2] https://mastodon.social/@[email protected]/117571045071997674 > > You can block on mastodon. _I_ can't. No account. But I also don't have the problem of directly encountering Mastodon content I might wish to block in the first place. > I blocked the sealion so he won't be bothered by my posts again > instead of responding to his message when I first saw it. This sort of Twitter/Facebook-esque tactic in our gamified social media environments is not one that I find productive for human intellectual development, not least my own, and I'm sure you'd agree that I'm sorely in need of such growth. > If you only want bug reports from true believers, you do you. I welcome _valid_ bug reports from anyone, and I think the GNU coreutils developers do too. I've followed this list for years and have never seen Pádraig or Collin engage anyone abusively as you do. Still, should I ever receive a defect report from Ignatius J. Reilly, I'll simply have to be prepared, as with the LLM submissions I got last month, to strike out the voluminous irrelevancies. Regards, Branden [A] https://en.wikipedia.org/wiki/Timnit_Gebru [B] https://www.theguardian.com/technology/2026/may/20/ai-engineer-says-google-unfairly-sacked-him-after-he-protested-against-work-for-israel https://www.washingtonpost.com/technology/2026/04/27/google-employees-letter-ai-pentagon/ [C] https://www.nytimes.com/2026/11/03/us/google-search-schools-restrict-ai.html [D] https://lists.gnu.org/archive/html/groff/2026-10/msg00050.html [E] https://www.zdnet.com/article/eric-raymond-how-ill-spend-my-millions/ [F] https://www.reddit.com/r/vintageads/comments/195nauf/shake_n_bake_and_i_helped_commercial_1968/ [G] https://scholarlycommons.law.wlu.edu/powellmemo/ [H] https://www.nytimes.com/1970/09/13/archives/a-friedman-doctrine-the-social-responsibility-of-business-is-to.html [I] https://en.wikipedia.org/wiki/Arrow%E2%80%93Debreu_model
signature.asc
Description: PGP signature
