Show HN: Vitriol – distributed, serverless publishing platform | Hacker News Reader

Show HN: Vitriol – distributed, serverless publishing platform

vitriol.co

37 pointsvitriolum57 comments

I'm going to go out on a limb here... based on modestly more private conversations than this Hackernews thread, I'm inclined to read the name—Vitriol—and the fundamentally distributed architecture together as being an effective dogwhistle for already angry people further driven to extremism by the entirely appropriate blackholing of Gab.

It may be a neat concept technically, but I already don't like the direction this is taking in setting up the future of this product.

... which will kill it, or at least keep it a niche echo chamber for deranged hate and outrage addicts. If they don't want it to be a platform for rage porn vitriol is an awful name.

The word addict here is somewhat serious. One thing I've been thinking for a while is that a lot of these hate and outrage trolls on the net are addicted to the brain chemicals these emotions trigger.

> vitriol – cruel and bitter criticism.

Really strange name choice.

Especially for a publishing platform. It's almost like you're trying to attract the very worst kind of online commentary.

When it first entered English from Latin, the word meant ‘glass’ (https://www.etymonline.com/word/vitriol#etymonline_v_7846). But that’s not a sense any modern English speaker could be expected to recognize.

They might make the connection with “Vitrification” if they’re inclined in that direction. They might realize that Vitr- is a common root, and might even guess what it is.

Yeah, though it’s a stretch all the same. I’m not sure what the intention here is. The name is either a joke, just plain bad, or etymologically overoptimistic.

There is vitrine in French, https://en.wiktionary.org/wiki/vitrine#French, and it means a glass display cabinet in English, although I've never heard anybody use that word.

Have you tried the internet?

Visita Interiora Terrae Rectificando Invenies Occultum Lapidem

Well, the name at least makes it clear that author of the system knows what it will be used for. I've reached the conclusion after years of crypto quackery and free speech nutjobs that actually, publishing platforms need to moderate what they are used for and by who. Free for alls like this attract some of the work people imaginable.

Why not give users more filtering options, rather than censoring certain views? Let the users choose what type of content they want to see, rather choosing for them.

Because sometimes one doesn't want to host a platform for, say, Holocaust denial. Platforms can, and should, moderate speech pursuant to the values of the company, which _can_ mean anything from heavy moderation to none at all.

People who want to take part in "fringe" discussions, whether it goes agains the culture, or assholes (white supremacists) will end up on whatever platform they feel the most free and secure. But what happens when a group of antisemites discovers a Jewish community, and decides to comment on their posts, and harass them over PMs? The burden is placed on them to add filters for all of the users, or the content posted, which harms the growth of the group; a new member may visit the Jewish community, see it full of hate, and decide never to join. Through inaction, your platform has traded the safety and security of one group for another, and in my opinion, it's those who seek to harm others that should be removed, not the other way around.

I worked at Reddit for three years (left two years ago), and it's taken me a long time to realize this.

I suppose it depends what kind of platform you are hosting. With something like Vitriol, where you are mostly posting content independent of other people, it doesn't seem like there is much possibility for abuse. But with something like Reddit I can see how moderation is more important, since you can direct message people.

Exactly. It's not a community, it's more a tool for publishing content that other people consciously decide to follow. The point being that this content can't be shut down by a central organization, but will disappear if nobody cares about it.

> The burden is placed on them to add filters for all of the users, or the content posted, which harms the growth of the group; a new member may visit the Jewish community, see it full of hate, and decide never to join

This is how it should be. Communities (not their platforms) should be responsible for policing themselves. They should not be dependent on the platform to decides what types of speech are not appropriate in their community.

A community of historians should be the ones deciding if a post in their community questioning facts about holocaust is a holocaust denial misinformation or actual scholarly research. This is not a responsibility we should be asking our platforms to take off our shoulders.

> Through inaction, your platform has traded the safety and security of one group for another, and in my opinion, it's those who seek to harm others that should be removed, not the other way around.

Nobody should be removed by the platform, instead the platform should provide people with the tools to protect themselves from all kinds of harassment on that platform, whether it is harassment of people based on their religion, ethnicity, sexual orientation or political beliefs.

Platforms choose to censor to protect their PR image, not to protect you. Why would you trust them?

Because platforms are businesses and it’s sometimes cheaper to just ban the groups that are bad for business.

Because platforms are businesses, and businesses are run by humans, some of whom manage to preserve their humanity, and don’t want to see their work being used by lowlifes spreading mindless hate.

I got can error message on loading the user intro: "There has been an unrecoverable error connecting to IPFS. Please retry again later."

Maybe not quite ready for prime time.

This also means that it will be impossible for third parties to limit access to your articles.

?? I don't understand. Somebody, somewhere is hosting the reading UI at vitriol.co, right? That's the chokepoint to attack.

Political issues aside, it feels like this has to be client-side software in order to achieve what it sets out to - more like scuttlebutt.

You can download the code from here: https://gitlab.com/vitriolum/vitriol-web and serve it from whenever you want, even localhost.

How can I verify whether this is actually working as advertised? For all I know this is just a blog running on a central server with terrible URLs.

And what happens if I create a piece of content, close the window, and then share the URL with somebody? Will they not be able to see it?

You can download the code from here: https://gitlab.com/vitriolum/vitriol-web and serve it from whenever you want, even localhost.

I really like the idea. The load of the intro page took around 35 seconds. Could Erlang/Elixir/LFE be used instead of JS/React? I am not familiar enough with IPFS to know if this is a valid question.

Lucky you. My experience ended with

> There has been an unrecoverable error connecting to IPFS. Please retry again later.

IPFS bootstrappers out there are still a bit unstable. Sorry about that - please retry.

IPFS is a protocol, I don't think one can ascribe 30+ second latency to language choice.

I should have been more clear: not language choice, but since it is distributed computing, I figured, perhaps incorrectly, that using one of these OTP/Beam languages would help with fault tolerance, and more efficient distribution across the peer-to-peer system. I am not a distributed computing expert, so my question was to gain an understanding, and your remark will spark me to look into this. Thanks.

Just curious why did you name the platform Vitriol?

I like the idea, but came here to raise a similar point:

Vitriol - a distributed abuse platform

the snarks practically write themselves

Visita Interiora Terrae Rectificandoque Invenies Occultum Lapidem

The name could be a bit more polished.

It comes from "Visita Interiora Terrae Rectificando Invenies Occultum Lapidem", an alchemical motto. No reference to hate intended.

Whatever the intent, how it will be perceived is effectively out of your control. Its current name is a headwind, actively working against its potential for success. If the project is important to you, you must rename it.

https://www.merriam-webster.com/thesaurus/vitriol

I wish we had a few years without tech buzzwords. Web 2.0, virtualization, devops, cloud, containers, microservices, serverless... I'd so, so much like to focus on something existing and make it work well. Maybe we wouldn't have 20+ crappy chat systems, but one good federated xmpp or matrix.

</offtopic>

That's such good point it should be it's own topic even.

I thought about just that yesterday, too: In individual development, we need stimuli, but we also need a LOT of sleep, and being protected from stimuli. Too little or too much stimuli stunts development, roughly speaking from the armchair. So why wouldn't the same go for our development and use of technology, both as individuals and as a society? At what point is a too constant treadmill a detriment?

And isn't it interesting that the idea of taking a break for 2 years, fixing bugs and learning the tools we currently have, writing docs, would probably be met with fears of "stopping progress" or "falling behind"? Maybe that's like the idea of never sleeping so you can finish university at age 14, and maybe we're finding out that it doesn't work on a large scale, either.

I was also thinking about stuff like Directory Opus, which started on the Amiga and is still being improved. I don't want Windows anymore when Windows 7 "runs out", but the thought of Linux without DO makes me sad... a program so good, I will probably run Win7 in a VM just for it. Old Opera was also pure craftsmanship and skill, and empowering users, and I miss that "spirit" so much. It still exists, but it should be the norm IMO, we should accept no less.

Browsing new releases, freeware or open source or commercial, that should be purely awesome, not something where you have to be constantly on guard and worry about "fragmentation", or people not knowing about prior achievements and insights. I want to just celebrate and thank software authors. I miss not being cynical in the least, actually.

But again, that's not against Vitriol, which I only very superficially glanced at, I'll give it a more thorough look later tonight and if I have any suggestions I'll post in this thread.

That's all well and good, but the examples you gave are problems of scale and economics, not invention. XMPP works as well as you want it to, and clients are on all the platforms, but Apple users want to use iMessage. That's not a problem that any kind of invention can solve.

The buzzwords themselves are references to incremental improvements in the state of the tech ecosystem that allow more powerful solutions for all. An app we all have on our devices right now, Slack, was once buzzword soup. Now it's an actual solution used by businesses the world over. It's not perfect, but it works better for our collective needs than IRC ever did.

What's going to replace the Slack's of today's tech world? I dunno, but I bet they'll be built on buzzwords.

We do not all have Slack on our devices.

I see many many non-technical teams that use Slack. Slack's made inroads in the business messaging market that used to be the sole domain of enterprise vendors like Microsoft, Cisco, and Oracle.

How many of those are buzzwords and how many of them are concrete descriptions of technology? Virtualization, containers, and microservices at least have commonly-accepted definitions and examples of those technologies are widely used, standardized, and work well at scale.

VMs, now containers, and what they imply for technology are more important than yet another godforsaken chat protocol - YMMV.

It can be both a buzzword and a concrete description of technology. It makes the jump into buzzword when people who don't really understand what the technology is start espousing it's benefits, and using the word to signal to each other that they're in the know.

If you've ever heard a middle manager at a big company talk about the tech stack their teams use then you should understand.

love the concept and the name, but it isn't working (IPFS error)

Sorry, IPFS bootstrappers are still a bit unstable. Should work now.

Why aren't you storing the text content of the articles as text files? (If you were, they could be accessed from other places, not only through the Vitriol interface.)

They're stored as HTML, because of the WYSIWYG nature of the editor. I guess Markdown was the other obvious choice, but editing MD is not so immediate on mobile.

I still can't browse them as HTML files.

So there is one thing that I always think when I see something like this. So if an article is replicated and stored on every user who reads that article once wouldn't that mean after some point a user won't be able to read any more posts because they have run out of space? Also, isn't it just collectively more expensive than just paying for a centralized service?

In the same note, where would I start if I knew nothing about p2p but wanted to build something like this?

How is it distributed and serverless if I'm pointing my browser to vitriol.co and getting stuff from there?

You can download the code from here: https://gitlab.com/vitriolum/vitriol-web and serve it from whenever you want, even localhost.

No, because I won't have my username and all that.

From gitlab:

> ... your article will be accessible even if you're offline, as long as someone is reading (sharing) it.

How does this differ from IPFS?

It is IPFS based on the error that comes up

AKA Twitter

More like Medium, but distributed.

Why gitlab ?

doesn't work

it's great

Reply on news.ycombinator.com