Why the 'Small Internet' Movement Wants to Revive Gopher (gemini.circumlunar.space) 111
Long-time Slashdot reader lee1 shares a new article from Linux magazine:
The danger and irritations of the modern web have unleashed a movement dedicated to creating a safer and simpler alternative. The old Gopher network and the new Gemini protocol have emerged as building blocks for this new "small Internet."
Anyone who has used the World Wide Web (WWW) lately knows that something bad is happening to it. It does not resemble the WWW of the early years, with enthusiastic amateurs freely sharing ideas and information. These things still exist, and the web is still an indispensable medium connecting the world. But the web experience is now encumbered with advertising, invasions of privacy in the form of pervasive tracking, enormous file sizes, CPU straining JavaScript, the danger of exploits, and door slams asking you to subscribe to a newsletter before viewing a site.
This unpleasant environment has led to a backlash. There are now some communities of developers and computer users who still desire a connected information system, but who seek a refuge from the noise, danger, and increasingly resource-hungry WWW. They feel that web technology does too much, and that since it makes various forms of abuse too easy, no lasting reform is possible.
The solution is to use or create a separate protocol that is simply not capable of supporting the technologies that enable advertising networks, user fingerprinting, or the myriad of other things that exploit users rather than helping them. This small movement has approached the problem from two directions that in practice are often merged: the revival of the Gopher protocol and the creation of a new protocol called Gemini.
Gemini would support its own lightweight hypertext format, and would co-exist with Gopher and HTTP as an alternative client-server protocol with built-in privacy-assuring features like mandatory Transport Layer Security and a "Trust On First Use" public-key security model. ("Connections are closed at the end of a single transaction and cannot be reused," notes the Project Gemini home page.) "You may think of Gemini as 'the web, stripped right back to its essence,'" explains its FAQ, "or as 'Gopher, souped up and modernised just a little', depending upon your perspective..."
"Gemini is also intended to be very privacy conscious, to be difficult to extend in the future (so that it will *stay* simple and privacy conscious), and to be compatible with a 'do it yourself' computing ethos."
Anyone who has used the World Wide Web (WWW) lately knows that something bad is happening to it. It does not resemble the WWW of the early years, with enthusiastic amateurs freely sharing ideas and information. These things still exist, and the web is still an indispensable medium connecting the world. But the web experience is now encumbered with advertising, invasions of privacy in the form of pervasive tracking, enormous file sizes, CPU straining JavaScript, the danger of exploits, and door slams asking you to subscribe to a newsletter before viewing a site.
This unpleasant environment has led to a backlash. There are now some communities of developers and computer users who still desire a connected information system, but who seek a refuge from the noise, danger, and increasingly resource-hungry WWW. They feel that web technology does too much, and that since it makes various forms of abuse too easy, no lasting reform is possible.
The solution is to use or create a separate protocol that is simply not capable of supporting the technologies that enable advertising networks, user fingerprinting, or the myriad of other things that exploit users rather than helping them. This small movement has approached the problem from two directions that in practice are often merged: the revival of the Gopher protocol and the creation of a new protocol called Gemini.
Gemini would support its own lightweight hypertext format, and would co-exist with Gopher and HTTP as an alternative client-server protocol with built-in privacy-assuring features like mandatory Transport Layer Security and a "Trust On First Use" public-key security model. ("Connections are closed at the end of a single transaction and cannot be reused," notes the Project Gemini home page.) "You may think of Gemini as 'the web, stripped right back to its essence,'" explains its FAQ, "or as 'Gopher, souped up and modernised just a little', depending upon your perspective..."
"Gemini is also intended to be very privacy conscious, to be difficult to extend in the future (so that it will *stay* simple and privacy conscious), and to be compatible with a 'do it yourself' computing ethos."
Veronica my love (Score:5, Funny)
Long time no see.
Re: (Score:2)
Paywalls too (Score:3, Interesting)
> "door slams asking you to subscribe to a newsletter before viewing a site."
Interesting position for a site that paywalls the article.
Re:Paywalls too (Score:5, Insightful)
Re: (Score:1)
This site already has a critical problem when it comes to RTFA. Sure, paywalls aren't bad in and of themselves but to direct a site like this to a paywalled article is a particulalry shady form of advertising IMO.
Re: (Score:2)
Re:Paywalls too (Score:4, Insightful)
I have no problem with the site charging for content, I have a problem with editors linking to pay-walled articles from a summary.
Re: (Score:2)
Says the author (Score:5, Informative)
But you can still wait a few week for the article to become free.
Why didn't you wait a few weeks to post the article to slashdot? Oh, of course. You're the author, which is a fact you conveniently failed to disclose.
Author(s): Lee Phillips
Self-promotion is one thing, but I think everyone is more that a bit peeved at the irony, which you don't seem to grasp, of posting an article about "door slams asking you to subscribe to a newsletter before viewing a site" when that's exactly what you've done.
Use Express-Checkout link below to read the full article (PDF).
Buy this article as PDF
Express-Checkout as PDF
Price $2.95
(incl. VAT)
Buy Linux Magazine
Honestly, do you think anyone will pay $2.95 for this single article of yours? That's a rhetorical question, although I wouldn't be surprised if someone paid and then copied the entire article here, just to spite you.
Re: (Score:2)
Re: (Score:2)
You're allowed, but we're also allowed to say it's a bullshit move. I wouldn't pay 25 Dogecoins to read your article.
Re: (Score:3)
So we're not allowed to post our own articles?
Of course you can submit your own articles.
It just seems disingenuous to link to an article that makes you money without disclosing your interests in the original submission.
Re: (Score:2)
Re: (Score:2)
One need not be paid directly for clicks for there to be a conflict of interest. For example, one might hope to get paid to write yet another article.
Re: (Score:2)
Re: (Score:2)
Well the English language gets abused once again to make something sound worse than it is, and give a moral boost to the position a poster is trying to sell us. People have an innate need to feel superior, to even their worse behavior.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Slashdot locks discussions after 14 d (Score:2)
Many people here probably already have a subscription
I have a subscription. Linux Magazine happens not to be included in my subscription.
the article becomes free after a while, as well.
Slashdot archives and locks discussions after 14 days. How does "a while" compare to that?
Re: (Score:2)
If following citations in documents leads to 100 different Gemini sites each of whose operators asks for a separate direct payment, how are users of Gemini going to be able to afford that many subscriptions?
Let it be dead in peace (Score:2)
Ever since gopher had itâ(TM)s ass KOâ(TM)d by the www in a most undignified manner, itâ(TM)s fans have been trying to revive it, asking for a round 2 or a rematch. But it got knocked the fuck out in Round 1 man. Itâ(TM)s proverbial ass was handed to it. Itâ(TM)s dead, Jim. Time to stop the CPR man. Thereâ(TM)s no reviving this parrot.
Re:Let it be dead in peace (Score:4, Interesting)
I remember fondly using the gopher interface for my uni's library catalog system. It worked very well and was super fast. And bonus I could access it over telnet remotely. Also the registration system was on gopher and it worked very well.
But no, the web did not knock it out in round 1. Not even close. It literally took 10 years for the web-based system that replaced it to come anywhere near the speed and functionality of that original gopher systems. The web in those days was super clunky. Now of course web-based interfaces are fairly slick, but they are definitely bloated.
But yes I agree this is not something worth pursuing.
Re: (Score:3)
Why shouldn't it? Are people really in love with the negative consequences of unnecessary complexity?
Re: (Score:2)
Back in the day we only had telnet and ftp in DOS on some old 286 computers without hard drives, in a Novell network. I could access gopher from telnet and it was decent. Archie for ftp, IRC could be used directly with telnet (I think this is still possible if one is patient enough), lots of MUDs, chat rooms, games (like FIBS). The Internet was very different.
One day I saw on MTV some addresses starting with http, I didn't know what to do with them, telnet would not connect, turns out I needed a "browser
Re: (Score:2)
Worth having, but a lost cause...unlike my fight against the Case-IH merger! Red and Orange shall never mix!!
Re: (Score:2)
Re:Let it be dead in peace (Score:4, Funny)
2400 baud??? You were LUCKY. When I was a lad, we had to write down the command on a cardboard box, walk 3 miles to the central server room, enter in the command (using punchcards!) and finally write down the response before returning. Of course we had to do this several times a day.
But you know, we were happy in those days, though we were poor.
Re: (Score:1)
When I was young...
Re: (Score:2)
Re: (Score:2)
2400bps is pure luxury! My CoCo2 computer with 64KB plugged into a regular colour TV connected at 300bps for years, you insensitive clod!
Re: (Score:2)
You could afford 2400? You were posh, man. I had a Smartmodem 2400 (I was rich enough to afford a real Hayes, at least), but connecting to servicd at 2400 cost an arm and a leg. 1200 was usually all I could afford.
Re: (Score:2)
Re: (Score:2)
Re:Let it be dead in peace (Score:5, Funny)
Reader: Is that...
backslashdot: My comment? Yeah.
Reader: Do you always write them encoded?
backslashdot: Well, you have to. The iOS Smart Punctuation system works with the input fields. But there’s way too many UTF-8 quotes for Slashdot. You get used to it. I don’t even see the code. All I see is straight quotes, curly quotes, em dash...
Re: (Score:1)
>smart
emoji quotes go home
Re: (Score:1)
There's no such thing as "emoji quotes". Slashdot is stuck in 1997 and can't display proper quotes, that's all.
Re: (Score:2, Insightful)
Re: (Score:2)
“ ”®±¼½¾Ø
Der König der Löwen
500 Hz
French style quotations
Nørsk
Re: (Score:2)
(Yes, we need a time limited edit button; maybe 1-5minutes timeout)
The comment section seems to support filtered unicode well enough. They have to be doing some handling to allow non ISO-8859 smart-quotes; and the EuroSign.
Cent, Pound, Yen: £¥
Accented Letters (European). ß à á â ã ä å ç è é ê ë ì í î ï ð ñ ò ó ô õ ö ø ù ú û ü ý
Non-ISO-8859:
Re: (Score:2)
not gonna help (Score:2)
Assuming this thing ever became popular, it would still have static ad banners and still track users through their IP addresses and clients.
I really like the idea of it, but they'd just keep adding features and eventually it would become the web again. It would need more intentional design to resist the shitty aspects of the web.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
First of all, trying to click on the link in the summary gives me this:
Secondly, there are comments above that say the article is behind a pay wall.
So no, people are not reading the freakin' article.
Re: (Score:2)
Re: (Score:2)
Yes I admit, I clicked the link just to see where it lead to. But don't be afraid, I wasn't planning on reading anything that would have displayed on the page even if it hadn't failed.
Re: (Score:2)
You must be new here.
Re: (Score:2)
It's just a bunch of sheep and parrots now. I mean, it was even in decline before the advertising started.
Re: (Score:1)
I like the idea of going back to straight HTML/CSS and FTP. IMHO any compute load should be on the *server* side. I guarantee that Java and JS would be a lot smaller/more efficient if it was. Most sites would be way more efficient, faster and easier to use.
Dang straight (Score:1)
Anyone who has used the World Wide Web (WWW) lately knows that something bad is happening to it.
Yep. What could that be?
It does not resemble the WWW of the early years, with enthusiastic amateurs freely sharing ideas and information.
Well, yes. Why is that?
We used to say - right here on Slashdot - that "the internet interprets censorship as damage and routes around it". That was until it was our desired censorship though, now we love censorship. Can't have any "misinformation" out there.
Re: (Score:2)
Re: (Score:2)
A simple internet could be censored as easily as a complex one.
Comment removed (Score:4, Insightful)
Not quite enough (Score:3)
I believe they're on the right track, but it's not quite enough. You *also* have to avoid letting big daddy G (or F) hoover up all the data or build social graphs from it.
I've heard a theory of a social network powered by Etherium where you would use contracts and mining to reward people for staying part of a content-addressable P2P network of encrypted blocks that dispose of themselves once the content is delivered to all recipients. Where everything is only available to direct connections by default, but can be +1'd out further to connections of a connection after being anonymized and then rehosted to boost it further (something that can also enable a sort of anonymous posting by sharing it as if you'd heard it from another connection), though you can also choose to ignore +1s from people you're connected to if you wish (or disconnect, for that matter). Content can only be opened by the recipients, because you share half a keypair with each connection, so it's just sharing stuff with your friends or possibly their connections. And connections spread by people requesting introductions from someone already connected to them.
At least that's the vision, I have no idea if this will ever get past the vaporware stage, as I have a deep suspicion of blockchains solving all the problems.
Re: (Score:2)
Something like Polkadot.
https://www.investopedia.com/t... [investopedia.com]
Re: (Score:2)
Content can only be opened by the recipients, because you share half a keypair with each connection, so it's just sharing stuff with your friends or possibly their connections.
This is key management hell.
First and foremost, you seem to expect end users to maintain their own public/private keypairs. This shit works at an instantaneous and automated level quite well, where a connection itself is only used a single time. But long-term public and private keys, synched across devices and all that shit, and they dont let their private ever leak?
The end user also has to maintain a database of other peoples public keys, the ones they are "connected to", synched across all devices?
OK, and NNTP? (Score:2)
Re: (Score:2)
Spam killed a big part of it.
Re: (Score:2)
Re: (Score:2)
Can you email me connection information so I can get a feed ?
fizball!maple!trek!serves!apple!orange!berry!rain!sunny!dazed!faded!kennel!house!john
Re: OK, and NNTP? (Score:2)
News.aioe.org or nntp.aioe.org
It's free and no account required, just like usenet of old but it has posting and crossposting limits which might be an issue for some.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
We don't need another platform, we need something that's not centralized, where users are back in control again. We need something that lets people connect to those they actually want to connect to, not the entire world and every random kook on the internet. We need something where all your data isn't going to get sucked up by the Googles of the world, sold to the highest bidder, and used to manipulate people or sell you crap.
And every part of that is difficult. Hard to coordinate. Hard to pay for.
Re: (Score:2)
You can check my Status if I allow you to.
I can check your Status if you allow me to.
Nobody is presented with an unrequested Status.
The only permissions that you can check are that you have permission to view anothers Status, and never that someone else has permission to view a Status.
Not sure how you make that a protocol. You can make it software, but software just reintroduces the problems trying to be solved.
Re: (Score:2)
Its protocols, isn't it. Back in the day, people created open protocols and shared them to allow open communications. TCP, etc to allow basic network communication, dns for addressing stuff, ftp for sharing files, irc for chat, nntp/ for open discussions, smtp/pop for private discussions, http for online documents, html to create documents, etc, etc, etc. No-one owned these things, because they were seen as the infrastructure that allowed the communications. Now nobody wants to create protocols because mone
Its own lightweight hypertext format? (Score:2)
Will Gemini be based on HTML 3.2 or HTML 5? Will it support some basic CSS? What about Javascript? I know the goal is to avoid the "mistakes" we're in right now, but are they talking about going back to the equivalent of HTML 1.0 here?
All the focus of the summary is about history, security and communication protocols, but the "own lightweight hypertext format" part is the most important part of the project to make it fast to download and lightweigh
Re: (Score:3)
Re:Its own lightweight hypertext format? (Score:5, Informative)
Will Gemini be based on HTML 3.2 or HTML 5? (...) are they talking about going back to the equivalent of HTML 1.0 here?
Neither of those. The default format, text/gemini, is a stripped down version of Markdown. Yes, Markdown has too many features, so the specification cuts down on even those. What this formatt allows are exclusively:
a) Pure text lines without formatting, not even Markdown formatting, equivalent to p;
b) Monospaced text for code and ASCII art, equivalent to pre;
c) Hyperlink lines, with exactly one clickable link per line (URL and title), equivalent to a;
d) 3 levels of header lines, equivalent to h1, h2 and h3;
e) Lists (unordered only), equivalent to ul/li;
f) Quoted text, equivalent to blockquote.
All of this using the UTF-8 charset exclusively, and with choice of formatting decided by the client. Also notice there's no option for embedding images or anything else. It's a pure text format.
Clients, by the way, are expected to support only plain text/* and this special text/gemini content-types, and even so they're required to support special formatting only for items a, b and c. Items d, e and f can receive special formatting or not, as even if they don't receive any special formatting they're still understandable by anyone reading the text.
An important detail is that the Gemini protocol itself allows for other content-types to be provided, such as actual Markdown, HTML, binary files or whatever, but it has no facility to inform clients what the length of that content is, so whatever file is provided keeps being download by the client until the TLS connection itself closes with with either a success or failure message. The FAQ explains this is so to prevent the protocol from becoming extensible, which would open it to anti-privacy feature abuse via custom header tags and the like. And since files are small there's no problem with clients not knowing how much they're going to download. For larger files, specially binary ones, they recommend linking to content in other protocols.
And what about support for people who need special features for accessibility?
The header allows to specify the language of the content-type so that screen readers can adjust their pronunciation accordingly.
Re: (Score:3)
Any time someone comes up with a so-called "privacy protecting" proposal you have to look at it very carefully and ask "ok, what's in it for them?"
True, and they actually address this in the FAQ. They opted to use TLS in this first implementation mainly because one goal of the project is to allow easy writing of servers, and since there are many free existing and fully debugged TLS libraries for several languages, they had to balance easy of writing vs. the more robust security of, e.g., the Noise Framework, but this is subject to change moving forward as more libraries for the later become available.
That said, implementations are required to support
Yes, please bring back.. (Score:3)
Oh Mother (Score:2)
All hail The Mother Gopher! She returns! All eyes turn to MN.
Remember kids: Minimalism is an illness. (Score:1)
It's the same thing as zen buddhism.
When people can't handle reality, they can either act extroverted, and force reality into a more comforting version (like Abrahamic terrorists, for example),
or they can remove themselves from reality.
Not everyone goes as far as lighting themselves on fire,
but sleeping on a mat on the floor, in an empty room, and obsessing over simplicity, from fixie bicycles to iPhones to this here, is pretty typical.
But you can cut away the cancer without cutting away all your limbs too.
Re: (Score:3)
Almost, but not quite (Score:3)
I spent about 10 minutes looking at the Gemini documents. They have a great goal for simple sites, and would suffice for a lot of what I use the web for. However, it has a fundamental problem - every blank line, paragraph, or link requires a separate connection and TLS session setup. This has huge costs in both network latency as well as processor performance.
The Modern Internet is latency-limited, not bandwidth limited, for small transfers. Their fundamental design forces lots of tiny packets, which run into a speed wall due to the latency issues. Every Line/Paragraph/Link in their text-only pages requires a half-dozen network round-trips - establish a connection and TLS Session, request data, close connection. A reasonable Gemini page with 5 paragraphs, and 10 links (perhaps 2KB of data) would require 5*6(each paragraph) + 5*6 (each blank line between paragraphs) + 10*6 (each link) = 120 round trips to the server. Assuming a network latency of 30 ms, 120 round trips is almost 4 seconds It seems idiotic to me that, for a TEXT only protocol, they don't serve the ENTIRE page at one time (or in client-specified chunks). Getting a 2KB page this way should take on the order of 5 round trips - 150 ms. Getting a huge 200KB page would take about 20 ms longer on a 100 mbps network connection.
The protocol seems to be set up to enable small clients - an Arduino or somesuch. That's a great goal, but the TLS session setup overhead (specifically the Public Key Encryption operations) will bring that platform to it's knees, and doing that on every single line/paragraph/link is simply hideous.
They should really consider network effects in the next version of their protocol.
From a purely personal aesthetics POV, I would have enabled PNG in-line because, next to text, pictures are the backbone of the web. I understand their desire to avoid tracking and 1x1 or transparent images; so I guess my desire could be satisfied with a client app that had a setting of "Download safe pictures", where "safe" might mean "Bigger than 16x16, visible, and coming from the same domain", a purely client-side decision and implementation.
Re: (Score:2)
Re: (Score:1)
> The protocol seems to be set up to enable small clients
and doesn't remotely scale. Taking your example, that was just one user. How is anything supposed to cope if there are 100, 1000 or more simultaneous requests? All hail the Slashdot effect, just for the next generation. No need for DDoS when a single ACK flood lasting less than a second will do to take it down.
What's so bad about reusing an established connection? It's like they are gleefully ignorant of the trials and tribulations of the internet
Re: (Score:1)
> every blank line, paragraph, or link requires a separate connection and TLS session setup
Not sure where you're getting this from (maybe that confusing line break in section 1.1 of the spec?). Files, including text/gemini documents, are sent in one go.
Re: (Score:3)
every blank line, paragraph, or link requires a separate connection and TLS session setup.
This is not true. You might have gotten this impression from the text/gemini document format, where every link and paragraph gets its own line, but the document is retrieved all in one gulp.
The protocol is simple, so can test this on the command line:
$ echo gemini://gemini.circumlunar.space/ | openssl s_client -connect gemini.circumlunar.space:1965 -ign_eof -quiet
Re: (Score:2)
Any image can be tracked. It doesn't matter if it's a 1x1 trac
gated communities (Score:2)
The "gated communities" like FB, IG, LinkedIn, Twitter, Pinterest and all other social media sites that all want to be 'the place to be' that won't allow content consumption without an account are really the worst of the web nowadays.
Geeky barriers to experience entry (Score:2)
are needed because "lusers" poisoned the internet and are lucrative targets-I-mean-audiences.
Democratization begets idiocracy and when an experience has a "filter" it keeps out human pollution.
An analogy is the off-road vehicle hobby world where you really need to build your own AND even if you pay someone for the vehicle still have a high technical understanding. The live steam engine world (the big ones) is another example.
When everyone can do a thing easily the worst are guaranteed to shit in the punch b
Protocol of the week (Score:2)
No protocol survives first contact with its userbase. The needs of your users tend to change over time. Security wasn't a huge overriding concern when SMTP and POP3 first became popular. If your organization was bitten by the recent Exchange bug [slashdot.org], you might find yourself asking, "why doesn't email support an easy, universal, cross-organizational E2E encryption?" When protocols can't be extended, they can't grow to meet new needs. It is a foolhardy endeavor to release a spec that can't easily be extended.
We s
Re: Protocol of the week (Score:2)
"why doesn't email support an easy, universal, cross-organizational E2E encryption"
Because email operators want to read your email. For legit reasons and otherwise. E2E encryption isn't a feature your employer wants on your work correspondence. E2E isn't a feature Google wants on their ad supported client. So, of course, if you want E2E encryption you need to add that.
Re: (Score:2)
Spam basically. I used to run email servers for a business and later for customers, and without spam checking, e.g. content parsing, it would have been impossible to provide a useful service. Its been 15 years, but I imagine that's still true.
Re: Protocol of the week (Score:2)
But if spammers had to find your PGP key in order to send you spam, it would reduce spam. Just reject anything that's plain text. It's not a panacea, but I don't think spam blocking is a good enough reason to skip out of encryption.
Re: (Score:2)
"We should take a step back here and decide what the problem really is with the web. ... These ideas might not all be good ideas, but they have a common theme. Poor design, commercialization, and centralization are all human-factors issues. Technological solutions to human behaviorsâ"like that alarm clock you keep ignoringâ"are generally doomed to failure. A Code of Conduct is a much better way to police bad contributors than a profanity filter."
Thanks for the insightful post. To take it one step
Re: (Score:2)
That is interesting, but not likely to change for one reason. Money is the ultimate tool of scarcity. Until that changes in some way scarcity will always be a goal, or weapon, or however you want to phrase it. I need money to survive, so if I have something I can use then I want to ensure its scarcity or I can't make money from it. In the old days, they wanted to solve communications problems so they created protocols (tcp, smtp, irc, ldap). Now that money is involved, the last thing people want to do is cr
Yes, money as a sign of poverty (Score:2)
As Iain Banks said in his Culture series: https://scifi.stackexchange.co... [stackexchange.com]
"Within the Culture money isn't used and most citizens have largely no conception of how money works beyond various truisms that are taught to young children about 'money implying poverty'. If you want something (within reason), you merely need ask for it. If it's being freely given you can have it. If that person says you can't have it, you can ask the nearest passing AI to make another one for you."
Or as I suggested around 2008:
htt [pdfernhout.net]
Re: (Score:2)
I was actually thinking about Banks and the Culture myself when I wrote my post. I guess that shared connection could be why your initial post caught my eye. I don't know if you're familiar with Ken MacLeod, who is a Scottish contemporary of Banks also writing space opera with political themes - although maybe more blatantly. His book 'The Stone Canal' comes to my mind at these times as well. It presents almost the counter theme, where the free market controls everything - police, courts, right through to n
Can't fucking wait... (Score:2)
Re: (Score:1)
Re: (Score:2)
What exactly are you waiting for? I still use Pine -- well, the Alpine fork.
When I went to university in the late 90s, Pine was one of the recommended clients, and I ran it on my 486 laptop with 8 MB of RAM and Windows 3.1. The same machine was also fine for web browsing with Netscape or Opera -- back then, we had this concept of a light client. This is why I'm somewhat interested in the small internet idea.
Don't forget the cookies (Score:1)
Solitaire over gopher (Score:2)
You can play Solitaire on gopher, so what more do you need? https://cosmicrealms.com/blog/2019/06/14/solitaire-over-gopher
Gemini was analyzed on LWM, last month (Score:2)
As it's past one month, in principle access to the full article is open, or really about to be : https://lwn.net/Articles/84544... [lwn.net]
Protocols not Products (Score:1)
from http://beneficialmanifesto.org... [beneficialmanifesto.org]