Superhighway84 – a Usenet-inspired decentralized internet discussion system
xn--gckvb8fzb.comI wanted to ask why use this instead of the still existing Usenet. But the IPFS based hosting is a neat idea.
This solves the problem that Usenet has with ISP's dropping the alt hierarchy for example. Using IPFS also deals with the hurdle of actually hosting the News servers.
I wonder how it will deal with the inevitable spam and csam.
Still, I'm following this with interest.
Aether was interesting, but it's many tcp connections chocked my consumer grade modem to death, which made it impossible to keep my node online. It was a legit cool project. I wonder if the #radio group I started still exists on it...
A bit off topic, but with regards to CSAM: I don't understand the motives of people who share it online for free¹. Distributing child pornography risks serious consequences, and I don't see what the compensation is that makes people willing to take that risk for no monetary profit. Maybe if we understood that motivation, we could make it less attractive to do; preventing the motivation for the activity might be easier than preventing the activity.
¹ Which is among, of course, the many other things I don't understand about them.
Offhand I'd say there's at least 3 reasons. Notoriety - either pseudonymous or otherwise, bait - like a state actor, and as an attack or exploit against a site operator. This is assuming some definition of public where a casual user not searching for or desiring it can discover it.
So basically the same reasons as any other undesirable or "illegal" content being found publicly.
> I don't understand the motives of people who share it online for free
From what I have read, it's to share and / or (worse) collaborate. A distant example that I recall was a couple in US that sold their 12 year old child in marriage to an adult (the legal marriageable age, with parental consent, in MA and NH is 12 years old).
> I wonder how it will deal with the inevitable spam and csam.
I'm thinking a modern Usenet would need something like ad blocker lists - volunteer run filter(s) that you choose to subscribe to.
> I'm thinking a modern Usenet would need something like ad blocker lists - volunteer run filter(s) that you choose to subscribe to.
This would be most useful if they're maintained per-category, by the people who frequent those categories. So then each filter list has "moderators" who can add stuff to it, the people who frequent a category usually subscribe to its most popular filter list, but nobody is required to.
Then because it's community-moderated and non-centralized you don't get the same problems with you do with email spam filtering where people will try to get their competitors' non-spam senders added to the global spam filter list.
> community-moderated
So’s Reddit, though. I’ve become skeptical of the sort of people who volunteer to moderate.
Incentive is tough. You end up with a percentage of toxic moderators in people who are looking to fill a social need in their lives, burning themselves out "volunteering", or people looking for power or status and internet points. It's one of the cases where a steem like cryptocurrency could work to correct for the toxic incentives, but nobody's figured out how to make it work yet.
I’m of the (unpopular?) opinion that users ruined the internet, and not megacorps.
For example is google crappy now because they just don’t care or is it really all the “content marketing” that did it?
I suspect any forum that gains significant enough traction to attract the bad actors will turn to shit.
I think even in the cases where users hurt the experience, we tend to blame the higher org for not resolving the conflict. "There are people bullying on the platform" -> "Why doesn't Google get rid of the bullying behaviors?!" instead of "Why don't people stop bullying?"
I think we also have a tendency to take it to the next level, which is often government. If people are scamming me through Gmail, we can tend to not think it's the fault of the people scamming or of Gmail but of the government for not forcing Gmail to stop the scamming or the person from doing scamming directly.
So I appreciate you bringing the attention back down to the ground level. While I disagree in that it would be only the users, I believe they (we) contribute to the problems as well as Google and the government and many other entities.
Yeah I don’t mean to give megacorps a free pass either. I just think it’s in our nature to collectively ruin most things that start out great… and like you say, blame everyone else.
Yea I hear ya. The more I've observed it in my own life, the more I notice that a lot of the blame I give others is blame that I have been privately giving myself.
Blaming my family for an Xmas event where I may have been exposed to covid? Mostly just blaming myself for choosing to go against my own desire. So on and so forth for soooo many situations.
Aether kainda did that by requiring proof-of-work for posting. And having mod elections amd other stuff.
> So’s Reddit, though. I’ve become skeptical of the sort of people who volunteer to moderate.
This depends heavily on the sub.
And in this case you have an even better alternative to bad moderators. If they're bad on Reddit you can start your own sub but then you have to convince everyone there to move. With filter lists, you can start your own filter list for the same community and any subset of members can use it without having to fork the community itself. If you do a better job, yours will be more popular.
You can also potentially have filter lists by category. One is for true spam, i.e. commercial solicitation by for-profit entities and scammers. The other is for trolls and shitposters. You might think a moderator is overly aggressive in classifying things as trolls but still want the spam filtering, and then you can.
Agreed. Never giving power to anyone who wants power is usually the right choice.
I'd rather use a karma based moderation system, which works on the premise that the majority of people isn't made of trolls, spammers, scammers and the like.
reddit is cabal-moderated, or moderated in the same way as prison gangs moderate their communities
Me too!
Maybe people could just make and sign posts in an appropriate group that you can subscribe to/diff/edit as you please.
That is exactly how Usenet works right now sans the restriction to sign posts. But several folks do indeed sign their posts.
Something like a killfile. I'm old.
Green Card Lottery 1994 May Be The Last One! THE DEADLINE HAS BEEN ANNOUNCED. The Green Card Lottery is a completely legal program giving away a certain annual allotment of Green Cards to persons born in certain countries. The lottery program was scheduled to continue on a permanent basis. However, recently, Senator Alan J Simpson introduced a bill into the U. S. Congress which could end any future lotteries. THE 1994 LOTTERY IS SCHEDULED TO TAKE PLACE SOON, BUT IT MAY BE THE VERY LAST ONE.
And to think, now the spammers run the Internet. How things have changed...
Spammers always ran the internet. Where were you in 1998?
I was on the Internet that was not literally run by Google and Facebook.
At the time I was very young and the whole thing was a scam. I couldn't believe that there actually is a green card lottery. /Tangent
A lot of the comments here seem to be missing the point and presumably come from people that didn't actually install the program and try it out, or hell even read the README.
This thing is extremely minimalist. A little bit silly to be talking about adding Markdown, antispam or privacy support when it uses $EDITOR for viewing/authoring posts, and the DB is world writeable...
Personally I think this is a lot of fun, it's the first project that has actually motivated me to install IPFS (which I thought would be much harder). It is in the vein of something like SDF where the retro interface might attract the right kind of people. I hope the author keeps working on it and if I knew Go I might contribute :)
IPFS is one of those things that shows up on my radar every year or so recently, I drool over it from afar, then I download it and install it and try to use it for a real use case and... it doesnt work.
I've went thru what I thought were the default tutorials and... nothing. Made me feel like the docs were leaving something out.
On paper it would be a useful tool to add to my toolbox. Someday...
Let's say I have ten hard drives where no more than two are the same size, so at least 5 pairs or maybe 10 different sizes. For this, zfs cannot be used to expose a single mountpoint.
Perhaps with something like ipfs a single mountpoint could be exposed and allow replicated data to be retained?
The key word here is "replicated" - as in a certainty that the data exists on more than one physical storage device.
I'm spitballing and this isn't necessarily directed at the parent of this comment.
Just an idea: What if instead of writing a client that uses a new protocol we did that on the server? That is, implementing something like this project, but into a local news server, so that users can connect with whatever client they wish to localhost:119, while the server maintains the groups and messages pool in sync through IPFS or any similar mean.
Discoverability has always been the #1 issue for me, so if someone can figure out how to establish any kind of namespace for IPFS I'm willing to apply a few more braincells to it.
It's a little ironic that this application actually does not run on Windows[1] apparently due a bug in a library[2] unfixed since 2018, which is owned and maintained by a $83B publicly-listed company[3], last updated 14 days ago, but unwilling to fix that little cross-platform bug.
[1]: https://github.com/mrusme/superhighway84/issues/13
Windows Port : https://github.com/altilunium/superhighway84
Subject: Re: Re: Re: Re: hello world
Newsgroup: misc.linux
= = = = = =
Yes. Altilunium was me. After i saw the HN thread about superhighway84, i can't wait to try it by myself. Clone the repo, install the IPFS, "go build .", then i found that issue. I solved that issue by changing zap.NewProductionConfig() to zap.NewProduction(), then i "disabled" the whole config module, switching it temporarily by using hardcode configuration. Now, it's finally working. New notepad.exe window will be spawned whenever i opened a new article.
But i found a new problem. The TUI is broken whenever a new notepad.exe window is spawned. To solve it, i make a new thread by using goroutine to "fork" the current process, so it could spawn the notepad.exe safely.
Reading a thread : Ok
Replying a thread : Ok
Creating a new thread : ((I'm currently testing it right now)
On Tue Dec 28 22:41:47 2021 mrus@cbrspc7 wrote:
> Oh hey, that's awesome! :-D There's an open issue on that rn. Will happily accept PRs if you'd like to. Don't have a Windows computer around to test, though.
On Tue Dec 28 22:36:16 2021 nya wrote:
> Finally.. After tweaking the source code, now i can run it from Windows.. ^^
It's a shame that Usenet did not persist. There's nothing stopping anyone from using NNTP to create a new separate "forum" that is not connected to "Usenet".
But since Usenet lost its popularity there's very little NNTP client software available now.
A shame.
It is a bit moribund but there are some newsgroups that are alive. It's full of spam, trolls, and obnoxious contrarians, but you can post about whatever you'd like, receive replies for whomever, and never have to deal with upvote politics or comment pile-ons.
One newsgroup I recommend is `comp.sys.raspberry-pi`. It has a pretty high SNR and is quite active, especially in comparison to some other groups on Usenet.
> upvote politics
I hadn't heard this term before for upvote/downvote wars (and associated groupthink/monoculture), but I think it's a good one.
Yup. I’m particularly fond of the results of my comment when called vote counters “sentiment thermometers”.
BTW Usenet is still around, I still read it every week. Most groups are dead or pure bot spam, but there are pockets of actual activity.
Usenet is quite awesome conceptually. It's a much better solution than mailing lists for large group discussions, and obviously better than all proprietary walled garden forums.
Ah the aesthetics are absolutely perfect.
software minimalism x vaporwave
And it's not just empty nostalgia. A text interface should be lightning fast. Hit a button, next screen repaint reacts. The ultimate retort to all the 5mb websites full of dropped frames, drop shadows and jank.
Solidity developers are already out here counting every instruction. Maybe what's old is new.
I’m confused by the term “decentralized internet.” The internet has always been decentralized. When the internet was created it’s primary use case was for decentralized communication.
A better term might be “blockchain internet”?
I think the author is saying: "decentralized internet discussion system", not "decentralized internet".
Reddit and Hacker News are centralized. Usenet was decentralized, the usenet servers were run by various companies (many different ISPs across the globe) and they synced up with each other via a p2p protocol.
I now I wonder what happened to Retroshare, which had a similar approach and never catch on.
Reddit and HN aren't "centralized", they're just individual nodes on a decentralized network. By that definition the usenet groups of yore were also "centralized".
Your newsserver could go down or block you and yet you could go to some other newsserver and have access to the network. If Reddit or HN block you, you aren't chatting on those platforms again with those names.
Reddit or HN could block you and yet you still have access to the network, just not the part of it hosted on reddit or HN.
But with Usenet, you don't lose access to any content at all. It would be like if one reddit server bans you, but you can still use reddit and other people can still see your comments because other servers, owned by other people, also host reddit.
Usenet is a loose federation of services handing articles to each other. If one newshost decides that they don't have the money/time to keep the lights on, you can seamlessly use another newshost which will let you fetch and post articles. Likewise if a newshost believes you're not worth the trouble to allow posting, another newshost can let you post. It's very hard to get banned from every newshost. That said, individuals often decide whom to block as many newsreaders allow users to score content and remove content from their local feeds that they feel score low.
If Reddit decides that a certain user or even a certain subreddit shouldn't be a part of Reddit anymore, that user or subreddit is off Reddit. If a news server decides not to accept or relay articles destined for a certain group, chances are high that another newshost _will_ be okay with accepting/relaying those articles.
This feels very... style over substance? As if it was designed with a 'usenet was so hacker-cool' idea behind it.
Traditional usenet clients were the most efficient discussion software in terms of usability. Everything since then (web forums, Slack-like clients, Discord, HN, etc.) has been a step backwards. That’s very much substance in my book.
I’d argue in terms of usability, todays platforms are an improvement. You can just look at pure user counts and see they are more usable/accessible overall than usernet (or irc for that matter) ever was. Discord, for example, is easy enough to learn that millions of non-technical users have been able to hop on and start using it.
Perhaps you’re thinking of efficiency in terms of screen/white space or CLI workflows over GUI workflows.
> You can just look at pure user counts and see they are more usable/accessible overall
User counts are a poor comparison when you consider how much the pool of potential users has grown.
As points to reference, Microsoft Windows 95 sold 7 million copies in its first 5 weeks so call it ~73 million copies sold the first year.[0]
20 years later in 2015, Samsung, Apple, and Huawei combined sold ~73 million phones across 5 models.[1] Windows 10, released that same year, had Microsoft shooting for it installed on 1 billion devices within 3 years.[2]
2005 estimates 1 billion (or 16% of the global population) online. 2020 estimates 4.9 billion (or 63% of the global population) online.[3] A million users used to be a big deal not that long ago.[4]
Not that I disagree we've gotten better at some things but Outlook Express (newsgroups) and mIRC (irc) seemed plenty accessible to millions of non-technical users.
[0]: https://news.microsoft.com/announcement/launch-of-windows-95... - when searching for a number I saw first-year estimates at ~40 million
[1]: https://en.wikipedia.org/wiki/List_of_best-selling_mobile_ph...
[2]: They made it in 5, https://www.theverge.com/2020/3/16/21116762/microsoft-window...
[3]: https://www.itu.int/itu-d/reports/statistics/2021/11/15/inte...
[4]: https://en.wikipedia.org/wiki/The_Million_Dollar_Homepage ; https://www.youtube.com/watch?v=4e0n7vTLz1U
> Not that I disagree we've gotten better at some things but Outlook Express (newsgroups) and mIRC (irc) seemed plenty accessible to millions of non-technical users.
This was my experience as well decades ago. Unfortunately, people claim that Usenet and/or irc are to technical for the average user to figure out. I guess the skill level of the average user has gone down in the interim.
Clients like Free Agent had a three-pane design that preceded iTunes' by years and I wouldn't doubt was the inspiration for its layout. New users today would feel right at home with it.
Forte Agent was the only reason that I had a windows PC in the mid/late 90s. Yep an IXer posting with Agent may have made me the butt of the joke, but that was some of the most intuitive software ever written.
Improvement in user counts? I'll give you that. Improvement in the quality of discussion? Not so much. It isn't a coincindence that we're discussing this here and not somewhere on discord.
Remember the eternal september.
One important Usenet feature that enables efficient tracking of and participation in discussion is that read/unread status is tracked per posting. (See my related comment in the sibling subthread.) That kind-of requires keyboard navigation to make it practical (e.g. press Tab to jump to the next unread posting, which then automatically becomes read). But keyboard navigation by itself also tends to maximize usability for frequent/regular users, which I'd argue are the important demographics for discussion forums.
One issue is with mobile UI, where there's really no good way to make per-posting read/unread status practical with just touch (and/or voice) input. But I'd also argue that asynchronous discussions (as opposed to chats/IM) are best performed on desktop. On mobile they will always remain a compromise in usability and content quality.
When a community I'm part of moved from Usenet to a forum around 2007, I remember a few technical points that persuaded members, mainly easier moderation and post editing.
One troll in a community of 200 has an outsized impact.
One Usenet killer feature that was lost in web forums is the per-post read/unread status, which made long-running subthreads manageable, decoupling them from each other. In web forums you only have "up to page xy" read/unread status, which forces participants to follow threads linearly and thus decreases asynchronicity and makes parallel subthreads hard to follow.
While Usenet traditionally had the minimalist interface that software people enjoy/cargo cult, FidoNet echos and other people sending and receiving from Usenet would often have similarly stylized custom BBS style interfaces to send and receive Usenet articles.
If it's just OrbitDB, you could probably put together your own UX to view it.
The branding comes off as pretty self conscious, but it sounds pretty neat. I’ll give it a shot when I’ve got some time.
Marketing works. And anything that doesn't follow the popular aesthetic is probably made with real love.
The idea of building new Usenet-inspired discussion system is great. We can update it with new distributed tech, privacy, etc. However authors chose to limit it to "retro text-only inteface." In my opinion this is a handicap. At least basic hypertext a-la Slack markdown is expected. Otherwise it will appeal only to old users reminsicing about good old times.
A number of NNTP clients supported rendering bold, italic, or underlined text by just adding enclosing asterisks, slashes, or underscores.
Markdown was designed to be human-readable; there's no need to support it at the protocol level or the implementation level. That said, I don't see this appealing to old users, either. IPFS kind of ruins the appeal.
Markdown isn't Slack-esq. By you saying that it makes it seem like you want this to be like some millennial service for kids who grew up on Discord and Twitter.
This is such a silly thing to bikeshed over. The oldest millennials are 40. Tastes change with time.
Cool!
I'm off to read about it. ...the basic idea on the main page is super attractive.
That domain! I’m surprised it didn’t trigger any security filters.
That said, this looks really cool. I like it integrates with your own editor.
Japanese people should be allowed to own domain names (in this case, the domain is literally just their name: Marius).
So should those with accented characters, but those can easily be used for phishing. It would be arguably unfair to allow Chinese and Japanese characters but not allow accent marks. Perhaps only partially supporting international characters would be better than only supporting ASCII, but that's the status quo.
That said, Chrome supports them both. here's cañon.com (which is just a parked site): http://xn--caon-hqa.com/ There is also http://xn--ms-mia.com/ (más.com) which redirects.
I know Chrome has done some stuff to prevent problems and understand when people just make the easy choice and show the punycode though.
The letter "c" can also be used to "phish" people who want a domain name that should include the letter "o" and yet we don't say the letter "c" is disallowed :/. The premise of domain names being your source of security is what is broken: we should be helping end users by scoring domains based on some combination of page rank and a web of trust implicitly built out of their existing address books, so when they are sent to bankcfamerica.com it shows up as "likely a wrong URL" without them having to squint at the domain name to notice the "c" (at which point we can probably also replace domain names with a post-scarcity collision-tolerant naming system, which would be absolutely glorious).
HN shows it before de-punycoding (punydecoding?), which if you're not familiar with Punycode, does indeed look a bit "spooky."
Even after accessing the link from HN the domain looks exactly the same in my browser.
Even if it is punycoding, I’m not sure how one would distinguish the actual domain from the coded one (unless you know the specific format convention that punycode follows).
Had to look up punydecoding (おもしろいですね!). Any reason why it's applied to kana when HN displays it just fine in comments?
DNS is generally restricted to a subset of the ASCII character set. While this isn't a strict limitation of DNS, there are enough DNS servers, clients, and applications using DNS that break on non-ASCII sequences that it's a de-facto standard.
In place of storing Unicode inside DNS, Unicode sequences outside RFC 952 (ASCII alphanumerics, case-insensitive, along with '-' and '.' characters) are encoded into RFC 952 compatible hostnames using Punycode, and stored thus in DNS.
Here in HN comments, Unicode is just embedded as actual UTF-8, no strange DNS encoding needed. The hostname, however, is actually xn--gckvb8fzb.com, hence why it's displayed as such.
(your browser will automatically convert from whatever encoding to Punycode where appropriate, so a link like https://xn--gckvb8fzb.com (should) work correctly, but the actual hostname lookup performed on the wire is for xn--gckvb8fzb.com)
Edit: well, HN also converts my Unicode hyperlink into the punycode equivalent. Interesting. It preserves the original encoding when I go back to edit the comment, but displays the Punycode form everywhere else. This gives credence to the idea this is intentional to avoid homoglyph attacks, as csnover states.
The punycode representation of the domain name is the “true” representation, HN just doesn’t detect that and convert it from punycode to Unicode. This may be unintentional, or it may be to avoid homoglyph attacks.
Because HN sucks as a web site.