Half Baked Ideas from Holla Forums "idea men"

Sounds suspiciously like gopher, or at least highly compatible with it.

Do you mean HTML alternative? HTTP is just a transfer protocol, it doesn't have anything to do with terminals.
Replacing HTTP might still be a sensible decision for this, though, because HTTP is not persistent.

Yes that. My prototype uses curl to download for now, but that's just a quick botch while I work on the viewer.
Lynx and it's variants work nicely for Internet in a terminal, but it's obvious that sites weren't made with it in mind. I want something tailored to the terminal experience instead.

That's HTTP, then. It's probably best to stick to HTTP unless you need a persistent connection and opening a new TCP stream for every interaction won't do.
Lynx used to be a very popular web browser, so it's not like the world wide web is fundamentally incompatible with terminals. Your problem might be with specific design choices websites make, which could be fixed by changing those websites, or possibly just with lynx's user interface because it's one of the earliest web browsers and not very refined.
If you can find a way to fix websites, use a proxy of some kind to fix them live instead of throwing out the entire world wide web. Think of something like this:
tedunangst.com/flak/post/miniwebproxy (note that the TLS certificate is self-signed and won't be accepted by your browser because he's an edgy fuck who insists on using HTTPS without an established certificate authority; this is intentional)

Distributed Hypertext Vector FidoNet

You mean like a BBS? A lot of them used one-key menus, but some let you type in commands instead (example: Sapphire BBS, by Pinnacle Software). Generally they could accomodate anything from dumb terminal to ANSI graphics & control codes.

^ AmiExpress BBS video here. Couldn't upload video here, chan says bad codec or something.
youtube.com/watch?v=Lxx9UsopV9k

If you want to create something successful, go with Photos/Videos & Messages.
The list goes on. And why? Because 98% of all people in the world just use computers/internet for that.

You're going to need an edge otherwise no one will use your application.

What about a messenger app that only allows emojis and no text?

Teh next step is Fu-Fme.

Everything has already been invented

This is the half-baked thread.

Emojli and Emojicate is a thing

...

A desktop environment for *nix systems that is an exact clone of classic Mac OS.

A desktop environment for *nix systems that is an exact clone of Windows 7.

A desktop environment for *nix systems that is an exact clone of fluid design Windows 10

A desktop environment for *nix systems that is an exact clone of TempleOS.

That's why peanut butter exists.

Facebook integration for Holla Forums

Logo generator

lol'd. Wrong tool for the job, my friend.

not an argument

Here's one: GPUs excel at data parallelism. Efficient task parallelism on a GPU is hard and probably not worth it.

the shit part with a mult-core z80 isn't the hardware, its the software.

i think the easiest method would be clustering.
one master CPU delegating some of his load to others using a shared bus.

Twitter bot that

Great taste, user.

this would be easy as fuck, get to it
then post source

Good luck hiring anybody.

Get it's whole penis inside your mouth?

Modern """lefties""" are not proles, they are middle class rich spoiled kids stuck in rebellion phase for too long.

...

t. middle class rich spoiled kid

command line youtube search. returns a list of videos for a search term you pas, optionally also displaying title, uploader, upload date, length, or whatever other information is attached to each video on a youtube.com/results?search_query= page.
not textmode GUI, like mps-yt.
if it can be done as just a curl+grep+sed script it should.
typical usage:
$ yt -Ttud -n 2 -q 'lets play minecraft ep 55534334'|tac #(with default n, or larger values of it, youll probably want the first results at the bottom)
55:40
LetsPlay: Let's Play Minecraft - Episode 274 - Sky Factory Part 16 [2 months]
The boys continue making their Sky Factory even better, all while discussing their favorite fake metal bands. Join FIRST for ...
[LINK TO VIDEO]
[LINK TO THUMBNAIL]

30:12
Rooster Teeth: Let's Play Minecraft - Episode 41 - No Petting Zoo | Rooster Teeth [4 years]
Geoff has challenged Jack, Michael, Gavin, Ray and Ryan to catch Minecraft's most fearsome wildlife. Winner gets the Tower of ...
[LINK TO VIDEO
[LINK TO THUMBNAIL]
$
so, to use it easily, you put "mpv $(xclip -o)" or however exactly that goes in bash/sh/etc in a shell function or a script, copy the link you want, then run that with a program runner, a hotkey, whatever you want. if you want to see the thumbnail first, some funciton/script that downloads the image and opens it in your image viewer or something like that.

Wow I had no idea the "what about the memes" man and the charlottsville driver were lefties.

mpsyt not good enough for you?

No. It takes over the screen with a pisspoor textmode imitation of desktop GUI so it can present more or less the same kind of list I described. If I wanted a "youtube client" Id just download...
smtube or minitube or what my package manager lists on a search for that.
I generally cant stand TUI shit the likes of ncmpcpp, ranger, etc. Gimped versions of desktop shit for hipsters who mostly only think its better because its keyboard driven.

You can use mpsyt from the command line.

no you can't.
invoking it with a search term doesnt count.
as far as I can see downloading it again just now, it takes over the screen anyway, which is my problem with it. a screen based visual interface for getting a list of search results is retarded.

You should probably pet it first.

Who said anything about IPFS or disk space? The bot would store everything on other websites.

i've written a script before to submit to archive.is. you don't even have to take a screenshot yourself which makes your life much easier as a programmer. archive.is makes one automatically when you submit the link. just make a get request to the screenshot page and download it after you submit.

alright. I'll see if I can write something like that tonight.

Archive.is only offers a partial image, though. The idea is to get the entire article into a series of screenshots so that people can read the article directly from Twitter, and it's split because you have to wrestle Twitter to get it to show you the original image and not a compressed, illegible piece of shit. The archive is only there for verification. (Plus some sites like Vice have anti-archiving scripts and you need to get technical to see them.)

I think it would be easy to cobble together a shell script for this using cutycapt, imagemagick and twidge.

A semi-imageboard where every post (not just the images) is hosted through IPFS and the list of post hashes is maintained in a blockchain. Posts are added to the blockchain when they are approved by the authentication server which does nothing but verify with captchas (unless this can be decentralized too?). "Bans" would be possible by anyone who wishes to moderate by making a list of hashes in the blockchain to ignore. Anyone can make such a list for people to subscribe to and filter out whatever they want like CP or spam. Since anyone can do this, if the most popular moderation group becomes corrupt and hides wrongthink then everyone can just subscribe to a competing group instead.

Posts and boards would be the same thing, replies work in a tree structure so the top level posts would probably all be something like "videogames" and everyone that wants to talk about videogames would reply to that (if they want, this can't be enforced). You also might have generals within the videogame post like "quake" and anyone that specifically wants to talk about that would reply to it. Each post is bumped by any reply within it, so if you post in the quake thread then the entire videogame board is bumped to the top of the list. Since all the posts are permanent, you could attach something like an entire movie to a post which would make the site also work as a piracy website. Even if the authentication server is taken down, someone can just make a new one and continue the blockchain from where it left off.

tl;dr piracy site and imageboard that can't be censored or taken down using IPFS

Is it possible, Holla Forums? I know nothing about how websites work other than making shitty html pages.

look faggot, just go install freenet

Pajeet detected

...

This is something I started working on but gave up because I'm too busy with my day job as a programmer.

tl;dr a turnkey home server that can be used to serve simple web apps. This would make it easier to create and distribute apps designed for mobile devices since web apps bypass app store restrictions. The goal is to create an ecosystem where there is a low barrier to entry for creating a web app, and where users have control of their own data since they run the server themselves.

markdownshare.com/view/b86fc9ed-79b8-4b05-9620-125b7df60aa8

Whatever replaces this
aminoapps.com/

One Cloud™® instance at Amazon or Google would be cheaper for your average pajeet to run his botnet webapp because of system redundancy and fast infrastructure.

Another reddid clone no one uses?

No, a lot of """people""" use it.
Next gen tumblr+reddit+discord shitshow

Those were app numbers, this is total users
Reddit: 234M Tumblr: 373M
4chan: At least 27M Holla Forums: where are the stats?
Well in this case amino is small with only 650K

Lisp-style markup language to replace HTML.
(title "Hello world")(meta 'charset "UTF-8")(style "/static/style.sess");; Asks to run in browser beforehand, signed(script "/static/main.cl" :lang 'common-lisp)(body(header (section :class 'header-left Welcome to my websight) (section :class 'header-right (list :ordered nil (item (link :target "/" Home)) (item (link :target "/about" Narcissism)))))(section :class 'main (paragraph :class '(description big) (emphasis SEML) is a markup language with Lisp syntax meant to replace HTML. (list :ordered nil (item Easy to use) (item Clean syntax) (item Well-defined structure))) (heading :tier 1 Editor wars) (list :ordered t (item Neovim) (item Emacs) (item Vim) (item ...) (item :num 9999 Notepad))))
Also replacement for CSS.
(rule :element 'heading :tier 1 (add 'font-size (px 10)))(rule :element 'header (set 'layout "flex") (set 'width (pc 90)) (set 'align "center") (set 'border (black) (all-sides (px 10)))(rule :class 'header-right (set 'gravity "right"))...

Why would you simply not replace HTML with something else then? Now it is just a wrapper around HTML, why not _replace_ HTML with something more intuitive if you are gonna go through the trouble. Same for CSS.

Actually come to think of it. I should do something to replace HTTP, etc and you need a special browser to do that (closed source by me so that no evil company can control it) and then everyone switch to it so chrome IE etc dont work in this new hidden protocol

You guys know markup languages are descendants of lisp, right? Don't get me wrong, I'd prefer a lisp instead of HTML, but you can convert it to lisp fairly trivially. You're just converting between:
arguments
and
(function arguments)

A lisp for defining web documents would need to bring something new to the table other than syntax.

Too bad you gained all that weight.

you could probably just string up RX to TX of every processor , going into a circle of course and solve the data handling through software

Yes, this is intended for end-users. However as I said in the doc, at least the initial focus would be on non-social apps (think myfitnesspal, note taking/reminder apps etc.). The goal is to make it so easy to write apps like this (that are hosted by the user's own hardware) that there would be better open source versions available than the commercial versions.

See

Somehow we are not talking about the same thing?

My half baked idea is web with two protocols.
1st protocol is for transferring content and it uses content based addressing. You request file by sending hash of that file to server and it sends you data.
2nd protocol is for providing human readable addresses and permanent links to files that change. Basically, you hit server with request like "8ch.net/tech/" and server returns latest hash to you which you fetch using 1st protocol.

I thought was a reply to
but actually it wasn't.

In response to , the goal of isn't to recreate the public cloud, which I agree is in almost every technical sense better than a home server. But the goal is to create a server that end-users can directly setup for themselves so they physically control and own their own data, and so that there is no walled garden that limits app developers either.

Just started writing the script a little while ago. Should have it one shithub tomorrow

Here are some stats :
stats.4ch.net/Holla Forums/
How ever the really interesting stat (page-views) is currently not working.
I remember some convo made by jim a year or so ago where he got deeper in the numbers, like bandwidth used and so on.

Have you heard of Urbit or Blockstack? They're neat projects with similar ideas.

github.com/SpaceBudokan/WhyTee

Alright, there you go. There's your glorious command line youtube script. Install it, instll the google api, and put your API key in on line 26, and enjoy. Yeah it's shit tier, I cranked it out pretty quick. Feel free to make pull requests.

I know I should have made the logo first, but give me a break. I was in a hurry.

I knew about Orbit and I agree it's very similar. The main difference is that I would try to re-use the entire web stack (DNS, getting TLS certificate via letsencrypt, etc.) on the basis that the web is by design very open already it's just hard to actually create a server. And I would try to re-use the main software stacks (linux, Python or other server libraries).

In contrast Urbit want to reinvent at least the entire software stack (and I believe a lot of the network stack). This aspect of Urbit I'm not qualified to judge, maybe this is a good idea. My proposal is very minimal and conservative in part because I don't understand the web very well. But I don't think Urbit will ever succeed because in addition to reinventing everything, they also do everything in an odd esoteric way, perhaps intentionally. I don't think this approach can work because while it might screen out annoying people who would be a drag on the community, it also screens out most of the people who might contribute.

Blockstack looks cool, my project could potentially play the role of the "storage" layer which blockstack doesn't provide itself. Blockstack doesn't solve the problem of app store walled gardens. One of the main points of my proposal is that if a user runs their own server they can always access it from the browser no matter what other restrictions their device has.

A centralized file host, based outside of the US with a clean and methodical layout based on the subject matter starting with general files such as movies,art, schematics etc. then getting to specifics such as 3D printing schematics, CAD, layouts, then beyond that till a specified file is found in addition to specifics being user based so users upload, and moderate themselves if something doesnt belong is a designated folder mark it to be moved to the appropriate area. only problem I can think with the idea is the amount of storage and copy right

Definitely. One processor runs the kernel and distributes processes/threads.

Did you ever come back or were you just drive by shitposting?

I keked so hard.
Good job user.

even though it doesn't work here. But whatever. It brings no error, just no output at all.

So.IPFS?

The proles hate commies

I stopped checking this thread. Thats pretty cool user, thank you.

they really fellated the kool-aid man, didn't they?

55:40
mpv ytdl://ytsearch:"best fnaf amv"
you're welcome

Reimplementing the web of trust: easy to understand, easy to use, sane defaults, decentral, automatic trustworthyness estimation of key pairs using linear equations. I've implemented such an algorithm for gpg and it works nicely. It's the gpg part that sucks because it has no usable library and is incredibly slow if you have more than a few hundred keys in your keyring.

Reddit: 234M Tumblr: 373M
4chan: At least 27M Holla Forums: 17M
Well in this case amino is small with only 650K
Can someone just nuke aminoapp.com/ already?

Already exists github.com/weavejester/hiccup
An example of it in use reagent-project.github.io

Why not use a language that already has tons of well tested parsers in all kinds of languages?

That's just a bullshit translation, and that was my point. Converting between a lisp and HTML without doing something truly unique is pointless. All markup does is move the function inside the parentheses.

The problem with HTML isn't it's syntax. It's feature creep, CSS, and javascript. We have these ridiculously bloated webpages running scripting client side. That's the real problem.

To be fair it's not just a translator because of it's homoiconicity. Would be cool to hear what you think would be truly unique.

Responsible encryption.

A cli application for cataloging recipes.
I started looking at curses bindings for python.

the mkdir, ed, and cat tools of UNIX(tm) solve this problem

An app that kills musical.ly or replaces Vine

You're right I forgot about that. Now the problem seems to be that it can be used client side. If it was just for server side scripting I'd be all over it. I am vehemently against client side scripting. It's one of the biggest problems with the modern web. If doing it server side is too much for your server, you need to rethink what you're doing.

I agree. Luckily it can be used server-side. In my experience it easily beats html templating like the one used in django. Separation of technology != separation of concerns