kemuri 🦀 @kemuri@qoto.org
Joined Nov 2022
kemuri 🦀
boosted
This article from @dangoodin about "juice jacking" is the kind of article which make me happy to be an Ars Technica subscriber.
Actual research and nuanced information, while still making it clear that general fearmongering is unwarrented.
I also really appricate that the article includes the following, which is really key in the disucssion:
"The problem with the warnings coming out of the FCC and FBI is that they divert attention away from bigger security threats, such as weak passwords and the failure to install security updates. They create unneeded anxiety and inconvenience that run the risk of people simply giving up trying to be secure."
kemuri 🦀
boosted
kemuri 🦀
boosted
Did you know that OpenGL on Asahi Linux supports some special features of Apple's tile-based GPUs better than Apple's own OpenGL driver? ✨
On macOS if you want zero-cost custom blending you have to use Metal, but Mesa supports GL_ARM_shader_framebuffer_fetch to do it in OpenGL!
We also support GL_KHR_blend_equation_advanced_coherent, which gives you a lot of common blending modes without having to write any shader code ^^
kemuri 🦀
boosted
How do we decide what a "true threat" is when we have no cultural consensus on what a "reasonable person" is?
https://popehat.substack.com/p/true-threats-and-american-cultural
kemuri 🦀
boosted
Twitter is dying, but my advocacy work helping trans people raise money via fundraisers depends on my five years of work I've spent writing on Twitter, so I need help.
If you're a trans person who needs help raising money for any reason, reply to this toot with your fundraiser. I'll reach out.
If you're a cis person and you can, please boost this post so your trans followers can see it.
kemuri 🦀
boosted
Datpiff, one of the main sources present and historical for mixtapes, is pivoting in a different direction, and contacted the Archive about hosting a library of pretty much all past releases. I naturally said yes.
Expect the collection to come soonish. (They're pointing to the general hiphop mixtape set right now.)
kemuri 🦀
boosted
The highest ranking search result for accessibility auditing tools is accessibilitychecker.org. The design and initial messaging had me suspicious out of the gate, but I didn't see immediate branding for an overlay company.
That is until I submitted a sample test of Google and saw one of the overlay monsters appear as a recommended solution.
Equally as gross, to my mind, is the "legal action" modal that appears when you attempt to use the "back to home" link.
Don't fall for this one, folks.
kemuri 🦀
boosted
kemuri 🦀
boosted
I don't begrudge any writer who wants to keep writing for Substack. Make money however you can do it! But I am tired of people saying that there is nothing scammy or problematic about a publication that hides its financials, hides who it is paying, and then says that we should just trust that it is a neutral platform with no agenda. And on top of that, they now refuse to have any kind of coherent moderation policy.
kemuri 🦀
boosted
We recently moved, so my wife asked for a copy of her medical records from her old hospital. We were expecting a CD that she could hand to her new doctor. We were instead shipped a box that contained over 60,000 printed-out pages.
Our healthcare system is truly insane.
On Rust Trademark Policy Draft
@PointyFluff
I agree that IANAL disclaimer is not excuse to be careless. To that end, I think it is better to be lengthy, to include reasonings and sources for the reader to judge, than to be brief. Of course it's your choice to ignore it on seeing IANAL.
ArchLinux e.g. does obtain a sublicense from Linux Mark Institute to use "Linux" in its name, and you can see that on its website. The Tux penguin is more like Ferris. "C++" has become a generic term via the ISO standard I think, as I have not found anyone claiming trademark rights to it. The Standard C++ Foundation claims ownership to the hexagon C++ logo trademark, which has the unaltered logo clause, although I have not found any published enforcement actions.
I focus on comparison to Python because the Python mark actually experienced a trademark dispute in 2013, so I regard them as the most experienced.
On Rust Trademark Policy Draft
Trademark laws themselves are a mess of nuances, and even though the Rust Trademark Policy draft is generally not out of ordinary, it unfortunately has failed to help illuminate the most important distinction, that is between nominative use and non-nominative use. This draft is heavily based on Model Trademark Guidelines, which has seen no usage in programming language projects before. It's an unfortunate choice of basis in my opinion, as many controversial clauses are verbatim from it. The Rust Trademark Working Group has not reviewed and revised the draft well enough.
That said, the amount of misunderstanding about it propagated by influencers is very disheartening. The hostility and arrogance displayed towards lawyers and legal expertise is in stark contrast to the usual humility the Rust community has towards technological expertise. I hope everyone do better.
I Am Not A Trademark Lawyer, if you are, I welcome your corrections. Keep caveats in mind, be critical but civil, let's discuss the concerns:
"I cannot even use Rust the word any more! How am I gonna write my blog article / book / presentation?"
That's called "nominative use", and is allowed by trademark laws, whatever the Rust policy says. Nominative use is allowed regardless if it's commercial or not.
"The Linux distro packagers cannot use the word Rust for their packages for the Rust toolchain any more!"
That was a concern back in June 2022, was fixed then, and the fix is inherited. Section 4.2.1 explicitly allows this non-nominative use.
"I cannot make any change to the Rust logo? What about my freedom of expression?"
You can freely use your creativity on the image of the Rust logo, as long as you do not use it to refer to Rust. In other words, you can totally publish your altered Rust logo as fan art, but you should not use it in a presentation to stand in for the official logo, or use it as the logo of something related. This "unaltered logo" clause is in effect for every trademark, including even the hexagon C++ logo. Python requires all alteration to be submitted for review, and publishes a guideline with permitted and not permitted examples.
"I have named my project Rust XXX, and I have to change it to something else? No other program language requires this!"
You don't have to change it, you just need to ask Rust Foundation for permission. Perl requires permission for such usage too. Python and Elixir have an exception for freely distributed products, but require permission otherwise too. "node" and "go" doesn't have this problem since their respective trademarks are "Node.js" and "Golang".
"I can't register rustmasters.com!"
You can, but the domain name usage prohibition is a mistake they should fix. Toyota USA v. Tabari (2010) has already decided that domain name can qualify for nominative fair use. All other policies mentioning domain name specifies prohibition only conditional on consumer confusion. Some comments say Rust is "going Oracle", but Rust has actually gone beyond Oracle here.
"I can't use Rust in the name of my user group?"
You can, and the draft policy is wrong again. This is almost certainly also a nominative use like in the domain name case, and Rust should change the unenforceable clauses. Python policy takes care to specify logo usage because logo usage is almost always non-nominative.
The policy draft does have serious issues, not because it can actually prohibit you from doing any reasonable thing, but because the things it tries to prohibit, it cannot. The Rust Trademark Working Group needs to improve their legal review process and not take the Model Trademark Guidelines at its words. I hope later revisions can fix the mistakes above and use more examples to illustrate nominative use and permitted non-nominative use.
@ThePaulMcBride @tosbourn
It is mentioned at the end of the draft that it comes from http://modeltrademarkguidelines.org/ and indeed the draft document does not deviate far from that model. Many controversial terms are verbatim from that model.
kemuri 🦀
boosted
Do you see this? 
I'm like a lion in savannah, and yesterday evening I was like a wild hunter in the bedroom: there was feet outside the bed 
@norbu
I am not sure which version you read, but the current version has 4.2.1, which explicitly grants license for using the Rust name for porting and packaging without asking.
It is also false that every blog post about Rust has to include a trademark disclaimer. Most blog posts, even if they use Rust in the title, are considered "nominative use" and is automatically exempted from trademark dispute by trademark laws, overriding this policy.
As I mentioned, Linux has a similar policy, so Arch Linux actually needs to obtain a sublicense, and show the following on its website:
> The Arch Linux name and logo are recognized trademarks. Some rights reserved.
> The registered trademark Linux® is used pursuant to a sublicense from LMI, the exclusive licensee of Linus Torvalds, owner of the mark on a world-wide basis.
@ThePaulMcBride @tosbourn
IANAL, but after reading a number of trademark policies, the Rust trademark policy is quite ordinary. The same language regarding domain name can be found in the policy of Java (Oracle), Linux, FreeBSD, Apache.
Python and Perl trademark policies do not have language specific to domain names, but they do have language against using them in other trademarks. Most don't realize though that trademarks don't have to be registered -- "common law" rights are conferred in a mark solely by its usage in commerce in the US. "Commerce" is defined very broadly in case laws as encompassing open source too. So it is very likely for a domain name in active use to become a trademark and therefore touch that clause nonetheless.
As for "rust" being a general word mentioned on orange site, "rust" is not a general word for a programming language. You cannot register "rust" as a trademark for an iron oxide product. A brief search on the US Trademark Electronic Search System reveals a handful of live trademarks of the single word "rust", including the game Rust, registered not only for the category of games but also those of several types of merch goods.
@norbu "The Rust Foundation has no desire to engage in petty policing or frivolous lawsuits." Nothing will happen if you register a domain containing "rust" and don't do anything with it. Section 5.3.2 is a general domain name exclusion term that you would find in many trademark policies, including that of Linux, Apache, Java, etc.
@squirrel @AstraKernel
The unfortunate fact online is that when criticisms amount to thousands, even if each one of them may be valid, the volume alone has a depressing effect akin to harassment. e.g. "Your icon is ugly." by one person is an expression of personal opinion, but the same thing by a thousand people would become a depressing social disparagement. (I think your icon squirrel is ugly-cute.)
To summarize the common use case summary of the draft policy, Ferris is always free to use, the shorthand "RS" is always free to use, and "Rust" can be used as long as a disclaimer of no affiliation / no endorsement is present. To that end it is similar to Apache's trademark policy. Regarding usage in presentation specifically, it is specified in 5.1.3:
> You can use the Word Marks in book and article titles, and the Logo in illustrations within the work, as long as the use does not suggest that the Rust Foundation has published, endorsed, or agrees with your work.
I Am Not A Lawyer but I wouldn't worry, because there are so many Apache talks and books out there.
Emotionally I agree that the language of the policy, and legalese in general, to be irksome. Partly it is the language being more specific about intentions and behaviors than daily usage, violating some norms for signaling friendliness. The other part, I think is the new boundary, the divide of the Foundation and Users made aware and bare, and every assertion of control by the Foundation in the policy widens that divide, and it hurts. Despite that, I still think the policy is necessary.
@ecton Rust Foundation probably won't be crazy like Monster Energy in how they go about enforcing the trademark though.
@ecton
IANAL, the following are just my understanding of trademark and copyright laws in the US.
Trademark and Copyright are related but separate issues. As seen from Monster Energy's lawsuit with Glowstick Entertainment over the "Monster" in a game title, in a trademark dispute the material at dispute does not even need to be an adaptation of any copyrighted materials. Even though the CC-BY license allows adaptations, it is only a copyright license. As long as Rust(tm) is a trademark, Rust Foundation HAVE TO go after usages of the logo which may mislead audience about Rust Foundation's involvement. I suppose that is partly why Adacore and Ferrous Systems are giving their verified Rust toolchain an entirely distinct name, Ferrocene, instead of VerifiedRust or something to that effect. (Ferrocene being a better and cooler name notwithstanding.)
For example, Adacore website has a Legal section which specifically attributes trademarks such as HP-UX, Java, and Mac to their respective owners as well as a general attribution for the ones not specifically covered.
Joined Nov 2022
