I enjoyed the "managing expectations" section here. It reminded me that even as an average programmer contributing to the project directly (via PRs, triaging, or just testing / running an instance) is going to be more valuable than donating $10 or whatever to a project like this.
Would it be possible to set up some form of CDN. I.e. a service where you could submit torrent links so that "peertubers" could pool their resources somehow, or if you would like to support certain user by making sure that they get the bandwidth?
So I am thinking of nodes that wouldn't actually host the websites but rather only would accept bittorrent URIs and then seeding them. To make sure that the ecosystem doesn't fail.
What would the economics behind this look like? So in short what server providers provide the best deal for just bandwidth. I suppose that this mostly would require bandwidth. The second priority would be storage space of course.
So what seems to be a potential problem with peertube is that if you put up a node and then if you get popular you might run into a bandwidth cap with your VPS. The average user probably wouldn't be interested or have the know how to do this, but I believe it would suffice that if there is a certain percentage of people who have interest in keeping peertube online it might be doable.
> Would it be possible to set up some form of CDN.
Could be possible with torrent seedboxes ... For example, http://dash.sonicbit.net/ offers 20 GB for around USD 2 / month with HD streaming of torrent ... A couple of these can be used to create some kind of CDN.
That is a really good idea actually, didn't know that such services existed. Perhaps several peertubers could pool their resources and set up CDNs using seedboxes.
In all decentralized solutions I've seen, they always seem to effectively become roundabout centralized solutions. No matter how "decentralized" something is, cost is the great centralizer: you can always be excluded either by the cost of running your decentralized hub or by network effects of being blacklisted in free "decentralized" centralized servers
I think that the real decentralized solutions can only happen when the cost of "I'm going to build my own with blackjack and hookers" is free or at least until the price is embedable into "usual" costs, like if your phone could become your decentralized, always-available hub
So, 10 -20 years. Prices will have come down enough for every household to host ALL their own data in their router. And offer interested parties API access. And host distributed social platforms.
Running a server from your home violates the terms of service of many ISPs (they want you to upgrade to a more expensive business plan for that). I’m not sure that is going to change in 10–20 years. Nor do I think router storage space is going to get significantly larger in the next decade, at least not in terms of being able to host large videos.
It probably depends upon which ISP you have and how "big" or "small" they are. I know 3 families who run their own servers from the home internet plan. Though they use it mostly from small sites and not for anything big/commercial.
The choice to maintain 80 projects and not grow beyond 10 employees is puzzling to me. Those two decisions seem antithetical to each other. 80 projects is too many regardless
I'm not particularly productive really, but seeing a few people I interact with pumping out open source PRs, and doing releases, all while (say) being the CTO of a company boggles the mind
Has this line of thinking ever actually convinced even one person?
It's why FOSS can't make serious inroads in a lot of domains. Like 20 years ago I was convinced we'd all never be paying for Word, but here we are, 20 years later, still buying Word.
How many people work on YouTube, Vimeo, ... the platforms themselves, as opposed to promotion, monetization algorithms, marketing, etc? PeerTube actually works really well, and as long as communities come together to help with support/promotion/... I don't see why it couldn't "challenge any of the incumbents" even with only one developer on staff.
Would probably violate trademark law and icann would make you give it up. I guess if you go opennic or use a different alternative root they might not care about court-ordered domain forfeitures.
Can someone explain the point of federating PeerTube. Their site doesn't really explain it, and it seems like you still need to create an account on every instance you want to use (if you want to comment on/rate videos).
How is it different from lots of independent sites, with RSS feeds for each channel?
You only need one account on any instance and you can rate/comment on any video in the federated network. However there are instances that don't follow each other and for those you would need a separate account.
Actually, with v3, you can follow "any" peertube channel or account regardless of whether your instance does the same or not. Earlier this was only possible with a mastodon/pleroma account only. Now your peertube user account in instance "a" can follow any channel anywhere in the fediverse.
This is like you want to follow some channel from tilvids.com but dont want to follow the entire instance.
yes. you only need an account in one "instance" and you can comment or aggregate videos from any participating instance and comment on the videos as if you had a local account there. Social function is a big deal of activitypub
What would IPFS solve that is still a problem today? The only disadvantage of bittorrent is that it doesn't handle mutable content, but 1) videos never change and 2) any mutation of content implies a fixed identity, and that is what Peertube is here for.
The biggest problem to solve is diacoverability, a thin shell is not enough if you want to solve the general problem
nope lbry uses their own custom torrent-like protocol on top of their own fork of bitcoin with their own weird python lbry daemon which communicates with a javascript frontend which is pretty much completely centralized in the browser version. Even the desktop version relies on centralized features like comments system.
Yeah I think so. Instinctively I do believe that this would be the most natural development. However I thought that federated social networks where kind of cool so maybe this is a better idea.
I do believe I tried to sign up for LBRY but I don't know if it really took up.
Might it be that LBRY simply doesn't scale or that there is some network effect that causes it not to work unless there are some specific conditions.
Is there a good search engine for peertube content? I feel that's the big problem now. Browsing random instances, I found a lot of (bad) porn, some anarchist concerts, tech tutorials in languages I did not know.
I made a search for "music" in sepiasearch.org, assuming this should be broad enough to get something interesting. The top result has 9 views. I can filter by "popular".
Look, I understand that Youtube's recommendation algorithm is problematic, but I need a way to find relevant content. "Sort by views" should at least be an option.
It almost makes me want to build a crawler and search engine for these.
Thanks! You should try to get a reference in framasoft sites, it is really useful! I could not find it while googling for peertube search. You should try to get some references up.
I was excited about seeing an alternative to YouTube. I'm getting more and more worried about the content YouTube shows to my kids. But I was so disappointed in the various peers i went to and the content there. Surely there is good content that doesn't rely on monetizing it as YouTube does.
I realize getting kids' content is going to take effort that requires funding somehow, but why aren't there good technical discussions posted on PeerTube? It seems like a great place for controversial opinions about tech, even if that isn't banned on YouTube.
edit: a comment about PeerTube itself, be sure to check out probably its most interesting installation (TILvids), by which interesting means you'll be watching a lot of their content for the edutainment:
I would say that if those members are hosting on a PeerTube controlled instance, then perhaps it is a moderation issue. If those members are hosting their own instance that federates? Well, then I think the tool is working as designed. Similar to Mastodon, federation will probably need some controls for flagging external servers as sensitive so users/instance admins can choose to avoid that content if they wish. That'll slightly fragment the market and frustrate content indexing, but... ehh, I think some market fragmentation is preferable to one central authority that not everyone agrees with.
One of the fundamental goals of free software is to enable its use by all people, even and especially people you do not like. That's not a problem for PeerTube to solve, it is a problem for society at large to address within itself.
To frame this another way: think of the criminal organizations which spin up a web server somewhere and install Apache on it to host their illicit content and advertise their services. Is Apache somehow to blame? No, of course not, the criminals are. The same logic applies here; PeerTube, as a sharing platform, is technically able to share any data, including hate speech and other objectionable content. As an unbiased tool, that's a feature not a bug.
PeerTube isn't a platform, tho. It's software and a corresponding protocol. It's akin to asking Microsoft why their operating system allows people to view csa content.
1) peertube is not a platform, 2) nothing, hopefully. If you don't like it don't watch it. There are 0 wholesome reasons to have a hard on for dictating to people what they can or can't say.
I think the answer is that they will allow them to use the platform which is equally open to anybody wishing to debate what they are saying. I know it's hip to consider everybody who does not parrot your opinions as a "white supremacist" but if you really look at the people that are in the movement you will find that they are miniscule in numbers and, for the most part, pretty foolish and easily ignored.
The thing about free speech is that it is not there to protect popular opinion but rather unpopular opinions. If it was there to protect popular opinions there wouldn't really be a need for it would it?
I read this kind of complaint, but frequent most of the services concerned and have never seen a single white-supremacist. I see open discourse much the same as HN.
The few times I've visited any of the 'chans however, it's been quite different - but curiously I don't see the same people making these complaints mentioning them.
Email is purely point to point though, it cannot be used to discover new content.
The mailing lists exist, but they are separate entities, not related to email providers or email protocols.
In contrast, https://joinpeertube.org/ recommends videos on front page, and you can find offensive (to some) videos in 2 clicks.
Are you suggesting that a significant use for email could be communication between white supremacists? It's probably less than 0.0001% of all emails sent.
On the other hand, if PeerTube is used to host a lot of pro-facist video content made by racists that could be a problem for PeerTube. It's an existential threat to the technology. If governments and ISPs see it as predominantly used for that then they'll ban it. If more legitimate users see it being used for that then they'll stop peering video content and it'll become much less useful. PeerTube will live or die by the network effect it needs to be effective. Guarding against the "wrong" users, or at least marketing it so that more of the "right" users join, is essential. Nothing about that is a judgement on the content or the users; it's the simple reality of developing technology in a society full of different opinions.
You're going to crucify an emerging technology for being where edge groups congregate, even though that's that's been the MO of the Internet since BBS days?!
Some technologies (e.g. Secure Scuttlebutt) have the ability for subgroups to exist without aggressively recommending content to every user, as an inherent property of their design; I think "should Peertube adopt some aspects of these designs" is a valid question.
There should be freedom of speech, however that needs to be balanced with the ability for centralized services - or trust networks - to form and allowed to moderate, without a centralized single government entity dictating what may be said or not to avoid potential tyranny - nor to force all organizations/platforms to be required to share everything [minus whatever "hate" list is deemed inappropriate by the government].
I do agree with Twitter, AWS, deplatforming who they want: why would you purposefully allow a bad actor (or misguided people) to use your technology/weapons? If they're not sophisticated enough or want to piggyback on the technology of people who don't believe you should be allowed to freely incite violence, then those bad actors will learn a lesson - as they seemed to - and then they will have to rally and align with those who do fully support a lack of moderation, etc. There are other problems like a lack of data and network portability laws, and PeerTube to some degree is a bridge for that, but the laws still need to be in place as a canary - if they ever happen to get removed.
The broader issue is societal - and that our institutions have been weakened by a number of factors including regulatory capture by industrial complexes, tied with the duopoly - leading to policy that kills off and suffocates the majority of people instead of supporting them to be able to thrive. Racism, as an example, could be argued as a multi-generational health whereby narratives, role modelling, overlaid onto circumstances of excess suffering, excess stress, that prime the racism to continue to perpetuate. It's also a sign of a lack of deep, genuine community and interconnectedness; we have politicians and the mainstream media and the vast majority, if not all, global brands constantly trying to manipulate us - and the whole ad industry enabling little "mom and pop shops" from being able to join in on the manipulate to sell things to consumers, buying attention for little to no effort - undeserved, unearned attention.
It's possible that if Trump hadn't been deplatformed then he could have further incited violence, more easily, and quickly enough; Parler and other forums did at least temporarily organize at least the most avid. It's possible too - though based on the Capitol incursion security forces weren't prepared and/or lacking integrity - that security forces would have been prepared to counter whatever portion of disenfranchised (and misguided) Trump supporters would be ready to escalate further violence. But the important part to all of it is that it was a wakeup call - that Trump got voted in because who slid by the control mechanisms and corridors the duopoly had managed to evolve and strengthen over the last X decades - that there are 80+ million people who aren't happy and are prone to lies, propaganda coming from and being reenforced by these various complexes.
In short conclusion, Andrew Yang's core policies seem like the new foundation necessary to counter the majority, if not all, of the problems that lead to this current state of America; Presidential candidate, now running for Mayor of NYC.
I am very hopeful for Peertube and their project, since I'm currently partaking in hosting a large video website for otherwise content too controversial for mainstream platforms like Youtube. (Political content if you were wondering, but I won't indulge what kind)
This takes back control to creators in such a way that essentially only domain registrars will be able to dictate what content may stay alive on the internet. (And naturally that each to their own get to federate to their own liking) At the moment domain registrars have been staying on what is IMO the correct stance of content allowance, requiring more or less police warrants to take down content. In fact I'm hopeful enough for this project that I've considered spending some time to get stuck in the development enough that I could even contribute pull requests or helpful issues to the tracker.
I would like to see a world instead where all channels on Youtube were their own "Peertubes" instead. Youtube/Google have really let consumers and creators down as it is right now, which in a way is natural, since they only really have allegiance to their sponsors and advertisers.
I got to see this first hand recently. Tesla has a rabid fan base that try to take down any negative information. Somebody posted a video on YouTube criticizing Tesla’s full self driving system and using short clips from other channels that are clearly fair use. Users spammed DMCA takedown requests and YouTube removed the video. Other users published it to Vimeo, Veoh, and a few other sites but those all got taken down as well, basically killing the video’s spread. The only ones still up are PeerTube and a split version, although I’m not sure if it would disappear if that host chooses to remove it.
It just surprised me how effective this method of censorship was and I wonder how often it happens with more important topics.
Have they tried just putting the video on a VPS VM and a single static web page with a synopsis of the video? Unless there are tens of thousands of simultaneous viewers, one should not even need a CDN. Happy to test this to see if things have changed. If there are tens of thousands of simultaneous viewers, there are about 35 or so decent CDN vendors out there. Some of them are quite affordable.
Thanks for hosting it. The video is very well done and rises valid points, but it's going to be difficult for it to gain traction fighting against the rabid fanbase. Wonder what could be a solution for this type of censorship.
2010-2015 was the golden age of YouTube. Adpocalypse effectively killed it as a neutral platform and now even mildly controversial videos get demonetized and/or removed. I therefore consider Peertube as essiantial as Wikipedia and Openstreetmap. I hope it becomes mainstream alternative to YouTube.
Wikipedia often lets smears stay on their site - even if false and then locks down the page so the person isn't allowed to present their own side of story to defend themselves.
As for YouTube, similar example - There's a current trending HN thread on:
> YouTube suspends account for linking to a PhD research on WPA2 vulnerability
>For all I've seen "non-white non-male" is generally given priority in all kinds of areas, certainly in start-up interest. It's considered "hip" to associate with these "non-white non-male" groups, irregardless of their actual performance in whatever area they were prioritized. (To a certain extent at least, they can't be completely nonperforming) Being able to show that you have included these groups in whatever way your organization works, gives you advantages in press situations foremost, but also in other ways such as meeting arbitrary quotas set by higher-ups and HR dept.s in your organization.
I'm guessing...videos leaked from inside factory farms.
Blending of conflicting ideologies implies compromise or trade-off, which would be a popular front.
In a united front, the groups do not merge or compromise. Ideologies and ultimate intentions remain pure, and the groups only collaborate in the name of overlapping interests.
These days as long as you promise financial benefits for about 10-30% of population, anything goes. That's why you've got so many populist parties around Europe winning elections. People don't care much as long as they get paid, can buy food, pay rent, and spend rest of the day browsing internet or getting drunk in front of Netflix.
Couple this news with this: https://news.ycombinator.com/item?id=26804977 . We need not only decentralization and "openness". We need independence. People should be held accountable by lies and defamation they eventually spread but the possibility of publishing information or "personal media" should not be arbitrarily controlled by any government, company or market.
Edit: replaced "Punishment for crimes like hate speech should be enforced" by "People should be held accountable by lies and defamation they eventually spread"
Punishment for crimes like hate speech should be enforced
Can I assume you're not in America? In the USA, "hate speech" is not, and cannot be, a crime. The 1st Amendment forbids content-based limitations on speech.
The US Supreme Court has ruled in favour of restrictions on free speech before; the First Amendment is not absolute. It is not unthinkable that with changing society and changing judicial review, the USA too may one day find hate speech to be a crime.
It's also unclear to me why that, in turn, is a problem. There are several fuzzy areas in American free speech law as it stands, and arguably most people are already happy to allow (and encourage) regulation in other important areas of life, from medical/food regulation to warranty regulations, to electrical safety regulations, to frequency band regulations. The 'power of the arbiter' there is almost never brought up as a problem, despite the fact that this immense power can have very real consequences.
If a democratically-elected government can be trusted (by the populace who influence the laws) to regulate what you can sell and what you can transmit in the airwaves, to regulate the minimum amount you can pay people, to regulate the age at which someone can be employed, etc. - why shouldn't the same government, subject to the same safeguards against misregulation, also have the ability to regulate this portion of speech too?
This is more of a devil's advocate argument than anything, but I'm interested in your thoughts.
So you're saying that because we've already given the government too much power in areas that it was never meant to be involved we should just keep going?
In China, "hate speech" includes anything that is critical of the government. I'm very happy that the U.S. has a very high bar for free speech, otherwise we would have been in a very bad spot all throughout the previous presidential administration.
Yes. In india criticizing the government lands you in soup for "destroying the social fabric of the country" and "against the interests of the nation".
A death threat against a whole ethnicity is illegal and in no way considered protected speech
No, in America this is false. A true threat must be a credible, present danger to its target. Saying "all people of race X should be killed" is nothing like that.
same goes for libel and threats to individuals.
No. Libel is not illegal, and is still protected against prior restraint. It's only addressable through civil liability suits.
Yes, I'm not in the US. But I think I didn't express myself clearly. Actually, what I mean is more something along the lines: "people should be held accountable by what they spread". Things like defamation are punishable in the US, right?
Also, in the USA, truth is an absolute defense to a defamation claim. The burden of proof is on the plaintiff to prove that the speaker was wrong, and further, in most cases, that the speaker should have known that it was wrong.
People like to talk about "defamation", but in America the bar is so high that it's only rarely applicable.
> Punishment for crimes like hate speech should be enforced
> "personal media" should not be arbitrarily controlled
This threw an exception, because "hate speech" is an arbitrary type. It has no objective definition. It is defined by whoever happens to be defining it.
Not that i agree with the GP comment, but isn't that exactly the point of courts?
Eg death threats and all sorts of illegal wording are inaccurate in the literal, programming sense. Humans (courts/etc) interpret and ultimately rule if a set of vague concepts is in violation.
If it wasn't for this capability it would be hilariously easy to avoid almost all repercussion from written/spoken word by just adding a dash of vagueness to the subject. Throw in a pinch of speaking in code and you'll never be charged for any crime, yay!
I loosely agree with you, but lets not pretend that english is even remotely as well defined as a programming language in all cases.
I wasn't saying it does, i'm saying _we already have to do that_. Ie, this isn't a new problem. It's a well understood problem, because language is not concrete nor so well defined as to be always explicit.
Well i use the term courts loosely here, as it is besides my point - my point was simply that humans _already_ interpret word as allowed or not.
We do not allow all english that isn't strictly defined as bad, i thought. I can threaten you in ways that will get me in trouble but are not, by themselves, a threat. English is not that simple.
So my point is merely that it seems to be the precedent is already set for humans to moderate human english. It is not a programming language to be parsed and verified.
