Comment by π hyena
Re: "Gemini trademark conflict"
I found out about the gemini protocol only 3 days ago by word of mouth. It was mentioned offhandedly in a lemmy comment.
Even if Google's AI were still named Bard, I doubt I would have ever stumbled across this place through a traditional search. And I have actively looked for information on the indie web.
So, maybe there's a bit of a catch-22: people who are refugees from the large internet will have to use the very products they are trying to avoid if they want to bring new people into this space.
There is a lot more at odds with gemini's growth than the fact that Google is shadowing its name.
Aug 07 Β· 2 months ago
Poll Results
1. Ignore it
ββββββββββββββββββββββββ 77%
2. Contact Google and fight it (Gemini *DID* pre-date Google's Gemini)
ββββββββββββββββββββββββ 12%
3. Rebrand gemini to something else?
ββββββββββββββββββββββββ 11%
91 votes were cast.
49 Later Comments β
π» darkghost Β· Aug 07 at 19:56:
Trademark is a strange realm. As @stack says, nobody registered a trademark on our side. That is the end of our case. But what is the function of a trademark? It protects companies and customers from market confusion. Gemini is a common word. As long as Gemini the protocol doesn't adopt the font and rainbow squished diamond thing, and we are unlikely to be confused for Google Gemini. Having no organization, being small, and being obscure also helps. Google sues who exactly if they are defending Gemini against our protocol? Yeah. Exactly. Its like some company in the 70s trying to protect their trademark on email or hypertext.
π requiem Β· Aug 07 at 21:08:
Social media hashtags have settled on #geminiprotocol, so I tend to just refer to it as that.
π stack Β· Aug 07 at 21:41:
Sooner or later Google will name some technology "Gemini Protocol", and likely, we will become unsearchable in Google. That is great as those who normally use google services without holding their nose likely do not belong here.
And unless they centralize us, there is no one to sue or change anything. Hence my warnings re: gemini foundation, funding of a central project, an authority with a code of conduct, officers, etc. Those who support such things while babbling of mass adoption and corporate backing are not your friends.
π stack Β· Aug 07 at 21:54:
Consider this subspace:
β bbs.geminispace.org/s/Critical_Mass
The mod there routinely deletes my dissenting responses because in their solipsistic little mind they cannot imagine a world where other opinions exist or that they may be on the wrong side.
You will always find me critical of those desiring to seek "mass adoption", killer apps, or other means of destroying our little community by flooding us with hordes of newcomers.
As others have explored, a community can preserve its values while expanding slowly. A rapid inflow does not allow for a gentle transition and accomodation of newcomers, who then rely on other newcomers to guide them. That absolutely destroys the original community, drastically changing everything.
I do not wish for that to happen here.
π‘ Breebee Β· Aug 07 at 22:03:
I wrote a rant blog post about this a little while back
β breebee.smol.pub/fuck_google_gemini_06_23_2025
I would love it if we fought it but I fear even if absolutely everyone who uses the gemini protocol pushed back, google is still unfortunetly the bigger fish
π pista Β· Aug 07 at 22:32:
Props to @stack for being the first person Iβve seen on the Internet to have some clue how a trademark works.
Even if Google managed to trademark Gemini, it would be limited to protecting a brand of AI software. A pre-existing protocol wouldnβt be affected, nor would it be grounds to nullify the mark since they are different classes of products.
The function of a trademark is to prevent confusion of like products.
Apple can be a trademark for a computer company. Apple can also be a trademark for a different company producing cleaning products. But you probably canβt trademark Apple for a brand of apples by an orchard.
Additionally it is easier to trademark a complete fixed image including logo, brand lettering, and the name than it is to trademark a single word or phrase.
While everyone wants the broadest protection possible, most registered trademarks are specific to the complete mark itself and not a lone word or phrase.
π stack Β· Aug 07 at 23:40:
Fun fact: in my previous life I was denied a trademark for 'RealBASIC'; the PTO claimed it was "descriptive and in common use". Somehow, McDonalds got "Lovin' it". It's all about how many lawyers you have on retainer bending reality.
πΎ jecxjo Β· Aug 08 at 02:31:
Honestly y'all should be more upset that Musk took the word Grok. Ruining literature and nerd culture is far more harmful.
π jsreed5 Β· Aug 08 at 03:54:
Since this protocol's use of the term "Gemini" was directly inspired by the NASA program of the same name, I (half-jokingly) propose we differentiate ourselves from Google's product by pronouncing it the same way NASA did: "JEM-uh-NEE."
π Unguided Β· Aug 08 at 04:03:
I kind of like it because there's an inherent level of deceit in saying you've been experimenting with the gemini protocol and some low info hiring manager thinking you're talking about AI technology.
π dimkr Β· Aug 08 at 07:28:
The real question is, are you worried, and if you are, why. The Gemini protocol is something you'll find if you're actively looking for it, or if somebody gives you a shortcut. Even before Gemini AI, it wasn't easy to find *for the masses* because it's something only some people (software minimalists, perma/retro computer folks, solarpunks, early WWW folks, ...) find interesting and useful and can describe using good search keywords.
πΎ jecxjo Β· Aug 08 at 19:54:
Gemini the AI is never going to be so big that Google would ever take action. Even at their size the negative PR for quashing an unrelated community of hobiest would not be worth the effort. They can just as easily cook their search results to make us disappear more than we already are.
Honestly my search issues are more on the gemini side than the http side. Looking for a copy of the Titan spec was hard with our search engines as tranjovian seems to be down and the first 10 pages of results are this bbs.
π stack Β· Aug 08 at 20:05:
I think all AIs are here to stay. New models will replace current ones, there will be another AI winter--my third, there will be no buzz but everyone will use LMs daily.
π pista Β· Aug 08 at 21:58:
@stack Weβve run into similar things in gemcutting.
There are a lot of people shouting about copyrights and how you canβt use their stone design.
Itβs all bullshit. You canβt copyright a gemstone design because it isnβt even in the category of things you can copyright. Itβs like how you canβt copyright the shape of a glyph. You can copyright the font as software used to generate the glyphs, but you canβt copyright the glyphs.
In all of history there is one βcopyrightedβ gemstone design because they submitted it 8 times with larger and larger teams of lawyers and bribes. A book telling how to cut the stone would be something you could copyright. A rendering of the stone in art is something you could copyright. One day this gemstone βcopyrightβ will die in court because it isnβt even in the domain of copyright. That fact no other design has ever been approved is proof of the error.
What you can do is patent a gemstone design. There are many gemstone design patents. But design patents, as opposed to utility patents, are EXTREMELY specific, and the adjustment of a single angle will no longer be violating the patent. Design patents are notoriously hard to enforce.
π stack Β· Aug 08 at 22:27:
Indeed. As a recipient of several patents and trademarks, I can attest to the nonsensical nature of the so called "intellectual property law", a ridiculous construct enforced by men with guns purchased with money from the rich.
Anyone who imagines winning a legal fight with google over the ownership of words is daft, plain and simple.
Rejoice that we as a community are judgment-proof, at least until the Gemini Foundation sponsored by corporate interests subverts us....
π stack Β· Aug 08 at 22:54:
Also, trying a trademark case -- if we had one -- would cost upwards of half a million dollars in the US alone. Now don't be silly, kids.
πΊ kallisti5 [OP] Β· Aug 09 at 01:36:
So.. If Gemini never self-described a trademark (TM), and never registered any kind of trademark (R), (and has never been used in commerce) then we will have little to no fight for when the big-bad Google wolf comes and tries to blow Gemini down.
It's why I said sort it now, get some "go away" money from Google, form a not-for-profit smol web org, and move forward on proper footing.
I get the "we're smol and punk rock" sentiment... we just have to make these descisions (or lack thereof) with the understanding of the ramifications.
Gemini is an eventual dead end without protecting its name, or rebranding every gemini:// link to something else.
π stack Β· Aug 09 at 02:25:
And why would anyone give you money? That is just nuts.
Maybe Bill Gates will give you money for not referring to your windows with that trademarked name?
Also, you don't 'self describe ' a trademark or slap a TM on anything. You have to file an application and while it's pending you may have some protection.
And a loose bunch of users cannot just file for a trademark. You have to do it alone or as a business entity, painting atarget on your back.
Sadly, you are way out of your depth here.
πΊ kallisti5 [OP] Β· Aug 09 at 02:35:
@stack
"Also, you don't 'self describe ' a trademark or slap a TM on anything. You have to file an application and while it's pending you may have some protection."
lol, that's not how US trademark law works.
@stack
I actually manage a logomark and trademark for an org. You're over your depth here.
https://www.marketingmedian.com/tm-vs-r-vs-c/
- Adding a TM is notice you intend to file a trademark in the future.
- Adding a R is notice you already have and it is registed with USPTO
- Logomark is an image of your logo which is trademarked.
- Trademark is a textual form your brand or logo.
@stack
*ANYWAY* if your argument is gemini should roll over and let Google walk on it... sure. That's definitely a path Gemini could take. RIP Gemini
π stack Β· Aug 09 at 02:40:
oh really.
π pista Β· Aug 09 at 04:32:
Why would be a dead end? Why does it need to defend a name no one can take away due to its use to describe a protocol preexisting any similar work? Why would changing the schema be needed and how would that help anything?
You are asserting things and not doing anything to explain why.
This is a silly argument to be having when what we really need is more people creating interesting things here. It will eventually sort itself out.
π RubyMaelstrom Β· Aug 09 at 08:25:
Yeah, I agree, a trademark or any kind of "gemini organization" is a silly ideal. As was mentioned previously in the thread, that just gives people someone to sue. Right now gemini is just a protocol. Is there a gopher org somewhere keeping gopher alive? No. It's just a protocol. Still going, after all these years.
π stack Β· Aug 09 at 12:10:
Just for education purposes, to avoid further spread of silliness:
If you do not file an application but put a TM on your mark, you may get some protection, but it is extremely limited geographically and you would have a hard time defending it, proving your priority date and uniqueness of your mark...
The intent to claim and posess your trademark is expressed by conducting a proper search, usually retaining IP lawyers, and filing an application to get a priority date. Until issued (or denied), your mark can bear a TM designator to show your intent.
Anything else is stupid, like Reddit guys putting TM on phrases for emphasis.
π stack Β· Aug 09 at 13:00:
*anyway* my argument is not that "gemini should roll over". Gemini is not capable of rolling over; it is not an entity, and there is no battle of any sort. Anthropomorphising Gemini is silly; it will not stand in front of a judge.
Google owns the trademark, and that is all there is.
If anyone stood up to represent Gemini in a stupid lawsuit or to claim the name (which is no longer possible, thank god for that), I would do whatever I can -- which is probably nothing other than bitching here -- to resist and educate people on the sheer stupidity of it.
Finally, this being a capitalist world, to get any court to act, you would have to show that Google has caused you a monetary loss or similar harm, and argue to recover it. That is likewise a goofy proposition.
My argument is that Google is doing us a favor by slowing down the influx of people thus preserving the character of Gemini community -- as well as filtering off potentially harmful (and ignorant) users who insist on using the Google search engine.
And by character I mean interesting, passionate and gently persuasive, sweet in tone, kind to fellow users, compassionate, knowlegeable and willing to share yet open and receptive to other people's ideas and opinions, and in every way wonderful, much like myself (and everyone else here).
π» darkghost Β· Aug 09 at 13:30:
@stack Speak for yourself. I couldn't persuade fire to burn gasoline.
π stack Β· Aug 09 at 14:02:
@darkghost, don't be so hard on yourself! You've altered my thinking about a couple of things in our previous disussions.
βοΈ gim Β· Aug 09 at 15:14:
yup, @stack is 100% right, TM is definitely not for organizations (and definitely not for non-profit organizations).
Defending TM in (various) courts is π°
π§οΈ candycanearter [βοΈ] Β· Aug 09 at 15:37:
i do agree with the sentiment that its protecting the ignorant users, but that also feels like a double edge sword? like obviously yes having a flood of common users would crowd the protocol, but having a LITTLE bit of discoverability would be nice (i like many people found out about it somewhat through word of mouth, since sdf has a gemini server). more of a personal issue, but i hate having to specify every time i talk about "gemini" that i mean the internet protocol and not the stupid ass ai garbage, and i do wish there wasnt a name conflict there...
πΎ jecxjo Β· Aug 09 at 17:21:
to @candycanearter point, I'm a little annoyed that the first few hits on "smolweb" tend to be this page promoting minimal css/js pages. to me this isn't the idea of what smolweb represents. its protocols and standards that are so minimal its damn near impossible to add junk to it.
π pista Β· Aug 09 at 20:20:
1. Make HTTPS Web content talking about the smolweb and how to get on it
2. Get people to talk about and link to this guide
3. ???
4. PROFIT! (not really, but it solves the problem)
If no one is writing about it or talking about it and getting other people to link to them talking about it, the search engines think it doesnβt exist.
Got a social media following? Link to stuff on gemini. Link to guides to how to get on gemini.
All these problems are solvable by the community just choosing to make it happen.
βοΈ sbr Β· Aug 10 at 04:55:
Moving on a bit. I kind of wish it was just stp:// or something (small text/transfer protocol) and perhaps sstp for a secure version, ie Gemini whilst the former was spartan. Titan would be ttp://
But in a stickler for short (often boring) names, typing out Gemini:// irks me.
βοΈ gim Β· Aug 10 at 06:31:
@sbr do you EVER type http(s)
(I think every browser now has a switch to always force https, a.k.a. HttpsEverywhere)
π§ the_mantelman Β· Aug 10 at 07:09:
I have recently begun to use more niche web browsers like Netsurf and Servo alongside the more established ones, and they don't always have https only. However, I think this discussion is going off-topic.
βοΈ sbr Β· Aug 10 at 10:53:
I wanted to explore some of what option 3 might look like, out of curiosity as I and most everyone voted for 1. Talking about copyright was getting dull.
And yes, I write the protocol in all the time, when writing content that links to other content outside my relative path.
π stack Β· Aug 10 at 11:54:
This discussion made zero sense in the first place as there is no trademark conflict.
π stack Β· Aug 10 at 16:20:
As a final thought, I love that we are using a Google-trademarked name and they can't do a thing about it.
Why would we change anything?
p.s. Copyright was not mentioned once here!
πΎ jecxjo Β· Aug 10 at 17:04:
@sbr your STP protocol exists, its nex://
π¦ zzo38 Β· Aug 10 at 18:04:
I always enter the scheme when it is not a relative URL, because I have it configured to treat entered URLs as relative when the scheme is omitted. I don't use HTTPS everywhere either (and I have changed it so that HSTS will not work).
π norayr Β· Aug 12 at 22:12:
what is google?
π kleuinrg33 Β· Aug 12 at 22:52:
I vote ignore it too. good people will eventually find this place.
π norayr Β· Aug 12 at 22:57:
to me the name 'gemini' and link to the space program is a little bit too much american patriotic.
gopher i understand as grassroots or network of underground tunners - that would describe what we have as gemini well.
perhaps i would prefer something agnostic like http oc xmpp: mdpp (just invented it): minimalistic data publishing protocol.
oh @sbr already said it, and suggested better abbreviations.
aaa look what i found, IANA Service Name and Transport Protocol Port Number Registry, geminiβlm is indeed registered for port 1590 on both TCP and UDP protocols.
by Tom Sawer
It is not our port or our gemini.
by Tony Sawyer, i misread the name (:
what is
β smallworld datastore server
Dude just ignore it
βοΈ sbr Β· Aug 13 at 04:26:
@norayr was also thinking similar things regarding american/space program connection. I like short names but am also not sure how I feel about celebrating a multi billion dollar pissing contest that involved settings tons of fossil fuels on fire and developing the military industrial complex, Instead of improving the lives of people on this planet.
π» darkghost Β· Aug 13 at 10:27:
I think it was a cute attempt to show that Gemini is halfway between the moon (http) and those first protocols (gopher et al). Gemini is less rigid than gopher but much less capable than http. Of course the sequencing is all wrong but whatever.
π² gonzosnose Β· Aug 27 at 07:53:
It's frustrating having to dig a little to find the actual gemini protocol specifications, but I guess nothing can be done. Even if Google's Gemini dies it'll still be clogging search results for ages.
π pista Β· Aug 27 at 13:37:
First two google results for βgemini internet protocolβ (without quotes) just now are:
1. Wikipedia article about us
2. Gemini Quickstart (https://geminiquickst.art) which tells how to get a gemini browser and get into gemini space.
Thatβs pretty darn good.
If you want to compete for just the word βgeminiβ you are literally fighting the stars.
πΊ kallisti5 [OP] Β· Sep 05 at 15:06:
@pista yeah, it's still reasonably easy to find gemini if you add on the correct search modifiers. It *is* however getting more difficult to find as Google's Gemini takes off.
π omorrigan Β· Oct 07 at 03:26:
the advocates for the protocol should be clear on the talks, we are taltkin about Gemini Protocol and not the Google dogfood
π stack Β· Oct 07 at 13:21:
Would you please lock this nonsensical thread
Original Post
Gemini trademark conflict β Anyone else notice that Google is using the Gemini branding for their AI crap? It's taking over the search results for gemini. Getting harder and harder to find things Gemini-related without stumbling over Google's AI rubbish. What should Gemini do?
π¬ kallisti5 Β· 66 comments Β· Aug 07 Β· 2 months ago Β· π³οΈ
Source