23
Looking for some help breaking in a new torrent client.
(lemmy.dbzer0.com)
submitted
5 months ago* (last edited 5 months ago)
by
jatone@lemmy.dbzer0.com
to
c/piracy@lemmy.dbzer0.com
tl;dr (even this is long ug): need halp from my fellow anarchists (~5-10) who happen to also share content with the world! Read the requirements section and then post a response. I figured this was the best place to reach out, I checked in with @db0@lemmy.dbzer0.com before posting, and we didnt have a better idea for a comm so here I am. =)
For those who volunteer and help, I'll be giving a much more detailed overview of how the system will work and what each part will do and why.
Some of the details about how the system will work that I'll be giving those selected, I'd appreciate being kept confidential. I want you informed but I also would appreciate the ability to surprise and delight everyone when things are released and to build the system at my pace without worry especially in the initial year while I get the base features working.
Requirements
- you've been a member lemmy.dbzer0.com community for 3+ months. (this isnt strict but i will be checking)
- willing to keep the system details confidential. I'll be explicit about which and why.
- reply with 'I'm Interested' and include some information about you.
- you dont need to include your technical / work information like I did. That information is relevant from me because I'm building the software and its directly applicable.
- by responding with an assertion of your interest you're asserting you fit the bill. your post is a vibe check, be weird, lord knows I fucking am.
- you run linux, It'll support macosx/windows eventually just not there yet.
- confident in self hosting / linux administration. while I plan on stream lining everything for the general usage its early yet.
- are familiar with torrenting ecosystems.
- able to safely setup and run a torrent client for whatever you're using it for.
- comfortable with buggy software. im literally building the ship while sailing.
- comfortable with slow update processes come august when my sabbatical ends.
- comfortable with software that will require a full state resets as it evolves. I'll obviously try to minimizes these but there is a whole storage layer aspect I'll need to get to at some point that will require a backup and restore.
Why I'm doing this
- primarily the rise of fascism is going to require non-techy and user friendly anonymous media distribution.
- I believe most (all?) the media platforms on the market are exploitative instead of communal and they suffer as a result.
- I've been extremely bored technically lately.
- I need (emotionally) to be able to do the crazy ideas I have in a number of areas in ML/distributed systems/security.
- I'm tired of my technical know how being used to build things I dislike/disapprove of.
- no company/investors are willing to pay me to do this, since they highly dislike not having control of an ecosystem, see the exploitative comments above.
- I want people and content producers to have a space to share together that's mutually beneficial: producers give up their content to the system and get paid on an ongoing basis. Consumers provide the distribution (free tier), and capital (subscription which enables at cost storage feature) that support everything. The worker collective will be able to fund off the margins.
a little about me:
- I'm a principal engineer specializing in non-linear dynamic systems. (systems that function off random processes in lay terms, its essentially the math that underpins the intersection of crypto/AI/distributed computing)
- I primarily work in startups helping companies go from ideation to real world working systems.
- I dabble in biochemistry and medicine. (currently picking up a BSN for humanitarian work).
- I'm heavily neurodivergent, I was literally one of the case studies children in the 90s for research. take that as you will.
- I enjoy taking things apart to see how they tick. if its a system that does something interesting I literally devour information about it...
- I love niche communities, you'll find me in my home town palling around with the LGBTQ+, fetish groups and often in drag on relevant holidays or just to fuck with people. You won't see pictures of me in tights with a riced linux system though. ;)
What I'm building
- a media player that has torrent protocol builtin implements bep51 for content discovery, rss feeds.
- tracker agnostic. use whatever tracker you want.
- will be daily usable by EO july, but the more complicated stuff will be later.
- the daemon / UX code will be released under AGPL v3 for personal use post this private alpha/beta. No commercial use will be permitted with the single exception of the non-profit I'll be founding as an entity at some point. the non profit will be a worker collective.
- will have a flatpak/debian/pacman installer, in flatpak its containerized essentially by default but will support native package manager installs.
- is efficient of resource utilization. I want it to run efficiently on something like the raspberry pi. its not there yet obviously and some of the initial models won't.
- builtin user space VPN client, wireguard protocol, not implemented yet, hopefully by end of year. which means essentially it can setup a VPN inside its program memory space without affecting your entire system and will work with any wireguard VPN provider.
- personalized content filtering. everything here is controlled client side. nothing is uploaded anywhere. but it'll have builtin filters for content genres, NSFW, etc.
- ability to stream your archived to devices, including mobile, while away from home.
- anonymous recommendations from your peer network (people in the network you subscribe to essentially, but you wont know who it came from) providing virality of content discovery without allowing abuses by larger players in the ecosystem. nothing is sent off your device for this.
the hard part of what I plan on building (1-2yr for everything I think):
- everything below here is opt in. But the storage alone will make it worth while.
- the system Im imagining is designed so only the verified content publishers will be able to determine how much time you've spent watching their particular content. and only the metadata for these providers will be sent off your device to my systems by default. The system will only be able to de-anonymize this metadata during the monthly payout to those producers.
- primarily these things are to fund the content creators, if my idea works 100% of the subscription costs (less the damn fees for the processing) will be given to the content providers in the system. however that requires a minimum number of users before thats possible.
Opt-in subscription tiers
I have a bunch of ideas planned and they're mostly self funding. the subscriptions (long term) are primarily to do payouts to content creators. but initially until economics of scale kick in at 1k subscribers at the base tier will let me do this full time.
- no min seeding ratio in the the default DHT for the application.
- publishing tooling for creators to upload and distribute their content to swarms.
- tiers (probably range from $3-$12/month). There are forces outside my control on the floor here (processing fees, ug) but its around $1.
- E2E encrypted cloud storage for files at cost. Only you will be able to retrieve this content. no distribution from the applications systems. basically this is for backups and long term storage.
- automatic subbing/dubbing client side. any models used / built for this will be 100% open sourced.
Opt-in profit shared advertisements
Two things here, first this requires a user base to be effective, so I dont plan anything here for awhile. Second, I hate ads to, primarily because they are almost never relevant to me. On the other hand informational distribution is important. I'd love to see ads for the specific stuff I want to see. My plans here for the profit sharing will ideally be between you the person who saw the ad and publisher the content you watched that month, but i'm not leaving out a small portion going to the non-profit atm, especially initially.
- Im hoping to be able to hit 95-98% of the ads value will be credited to you. 2% to the content creators.
- you'll be able to directly tailor what you see.
- Im hoping to focus on concerts/festivals/local local community events primarily.
because you and these protests are quoting 3.5% as if its a fact. its a magnitude without a vector. the vector is the important aspect, not the magnitude. if your vector is 'run around city limits with pithy signage' (our current protests) I assure you, nothing will change even if you had 50% of the population doing it.
historically its when the vector hits around 3.5% that matters. if you have the wrong vector the magnitude is completely unrelated and you'll end up in a different location than you expected. Notice that around aspect of 3.5% because guess what: it actually isnt a value that matters much. its the vector that actually institutes change.
the population against the GOP and Trump already exceeds 3.5% they just dont know how to effectively protest and the people running these protests have nfc what they are doing when it comes to leading such activities. that was the point of the voting numbers.
my point is quoting that number as if its a magical threshold we need to reach without anything else is actually counter productive to actually getting change.
the problem with you being an idiot and spouting this nonsense is hilariously exemplified by your assertion that reading papers and books somehow gives you insights without actually putting that information into its proper context which you clearly havent, nor have the people running these protests.
if you want these protests to be effective you need to do a few things:
anything outside of those 3 aspects in just noise. I'll simply point you to luigi as a wonderful counter point to the 'nonviolence' aspect of that 3.5% number tossed around. His actions where highly supported by the population, used violence effectively, and it did effect change however small. if you look at basically every successful movement there is always violence involved. which people like the author try to ignore. take gandhi's movement against the british, widely lauded as non-violent, however if you take a wider perspective beyond that movement itself you'll learn there was violence involved. by other groups at the same time you can't decouple those two groups from the outcome. they both play roles in the result and thats exactly what the paper where the 3.5% number comes from. ignored the activities of the groups operating at the same time as the peaceful ones.
And then there is the question of what actually constitutes violence. did they consider property damage as violence? how about economic blockades? I consider both to be violence and effective agents of change.