Hacker News Reader

Comment by wl | Hacker News Reader

There's two distinct Nitter use cases I need to replace.

1. Someone drops a link to Twitter. Twitter hides threads and throws items in some weird non-chronological order—assuming I don't get a login wall. I need an unfucked UI.

2. There are some content I can't get anywhere else that I follow through RSS. I wish these people would move elsewhere, but if they haven't by now, they probably won't ever.

I may just run a local instance with an account created for the purpose if that remains viable, but associating all that with a single login/IP address is something I'd like to avoid.

View on HN

1) Ask people to send screenshots of the tweets instead of links. I think this has naturally been happening a lot more over the years anyway and people will just get used to do that.

2) I guess those people will slowly realize they have lost most of their audience anyway.

Right? (#2), I can't be so far in the minority to just not be willing to bother with Twitter. There is really very little content that will compel me to have a bad time consuming it. Is it 1 out of 10? 2 out of 10? More? Even if it's just 10% that's still significant. Just my $0.02 of course.

Kinda sad that we have to send screenshots since it hurts accessibility, but I guess that's less of an issue with local OCR on most devices these days.

Imagine the conversation just a few short decades ago...

"We perfected OCR! Blind people can read!"

Yay! Awesome.

"Yeah, and soon we can route around damage when the capitalism growth mandate breaks hyperlinks!"

Yaa— wait what??

You should ask people to send a textual copy before a screenshot, OCRs are not nearly as conveniently accessible and effective to justify requiring them

In addition to a chronological feed, Nitter provided pagination and link de-obfuscation. Would be nice to get that back somehow.

Right, all the "just don't use it" comments miss the point. I used Nitter specifically because I don't use twitter, so I could see the contents of a link that was being posted or discussed. I suppose a solution could be "ignore a large swath of posts and links and discussions" which is basically what I do, but sometimes it's nice to have the option to look at them. Same as if you don't use MS Word, someone might still occasionally send you a word document and it's nice to have a way to open it without having to install Word.

+1. Need something like archive.today/.is for Twitter so you can rip and archive the content that might not live elsewhere. Grab it, stick it in Wayback Machine, return a Wayback url.

+1 on Reddit as well.

Reddit doesn't have login walls yet but it has way too much information stored within their walls to not have a backup / non-social-media way of extracting it. It's infeasible to have Reddit blocked because it's UI is intended to be addictive like all social media but also be able to extract information from it.

For the moment, old.reddit.com is sill useful.

For the moment.

It’s dying a slow death through neglect. Image posts don’t work correctly, image comments don’t show up, and the dirent comment links generated from www don’t work on old

Still better than logging in, using the new web app or downloading the app.

It's not like 95% of the content is any good anyway. You have to dig deep into a niche to really get any value, and the last few years, less and less. Of course, I haven't logged in for 3-4 years so maybe I'm missing something. Doubt it.

There's also a thing lately where link targets posted on New Reddit (I assume, since mine aren't doing this) are all lower cased, while the link text is correct as you typed. This breaks the links for some sites. In addition to the issues with underscores getting extra slash-escapes.

> Reddit doesn't have login walls yet

There isn't one on the root directory, but Reddit has plenty of login walls.

On New Reddit. I haven't seen any on Old Reddit yet.

There's not a difference. If a subreddit requires you to be logged in, then it requires you to be logged in.

There are 18+ age walls that just force you to login, often in unnecessary places.

Plus mobile sometimes refuses to show some things.

old.reddit still works though.

Some subreddits will impose the 18+ wall because reddit hasn't specifically vetted/approved of the subreddit. So it will be something totally not adult, just a small sub with important information you're looking for and you can't view it anonymously from a browser.

These can be avoided by using old.reddit.com instead of www. I think.

I think this is only true if you're using mobile. If you're on a desktop you can get through just fine. Usually.

https://academictorrents.com/details/7c0645c94321311bb05bd87...

Reddit also now blocks all Mullvad connections that I'm aware of. It's kinda ironic seeing all the scamy YouTube ads promoting using VPNs to watch Netflix in another country when that's never worked and other companies tend to be hostile towards VPN users.

The EFF just recently wrote an article with instructions on how to persevere & archive your own tweets on the Wayback Machine, but it involves exporting your own backup and uploading it to them. Since the API is completely cut off from Twitter, there is no official way to backup other people's accounts.

But archive.today uses scraping and all sorts of tricky methods to bypass paywalls. I honestly don't understand why Nitter can't just stay logged out and rotate IPs. Although I'm sure that gets pricey when other people are accessing it constantly.

https://www.eff.org/deeplinks/2024/01/save-your-twitter-acco...

If the scraping model is impaired due to aggressive countermeasures, end game are browser extensions that scrape as users view the site and ship scraped data back to a processor, similar to recap the law (uses an extension to scrape the PACER legal database and ship digital artifacts to the Internet Archive). Care will need to be taken around potentially sensitive data that could be shipped if users are logged in.

https://free.law/recap

Oh, that's a very cool project! How successful has it been? If it wasn't for Sci-hub that would be a great idea for the scientific publishing world as well.

Very successful. Millions of court filings extracted and indexed, and a crucial component in driving down PACER costs.

https://news.ycombinator.com/item?id=24086570

> I'm the director of Free Law Project. For the case mentioned in the article we actually did a full expert testimony figuring out roughly how much per page it'd cost to run PACER using AWS GovCloud and a handful of other assumptions. It was...half a ten thousandth of a penny per page

https://www.courtlistener.com/docket/4214664/52/15/national-...

https://news.ycombinator.com/item?id=24085158

> Government’s PACER Fees Are Too High, Federal Circuit Says

https://news.bloomberglaw.com/white-collar-and-criminal-law/...

This model also works well for deep web content archiving.

There was a gaming message board where someone wrote a browser extension that would back up all topics someone visited in the background while they were reading them. It became important for archiving as much content from those forums as possible as the forum was in the process of shutting down.

Why don't you make a twitter account and install a duplicate web browser that you only use to open twitter links (and other crap that you don't want polluting your normal browser)?

I don’t want the company to have any information about me.

That's why I suggested you use a burner browser.

You also need a burner phone number. Otherwise it will be difficult to get an account.

I didn't know you needed a phone number for twitter. Yeah, then it's difficult.

These systems are completely losing propositions. My data is worth $$$$, my time is worth $$$, and I have to spend $$-$$$ in order to safely interop with them?

Twitter/FB don't understand just what a terrible value proposition and complete waste of time they represent. I'm happy Ediot Mashbrains has been bleeding money for the damage he's caused.

I also don't want to give them any traffic.

Spite, personally.

For 2, you might want to send them an email asking them to post said content elsewhere.

Given that it's Twitter, there's not a ton of text or anything in what you're missing. If it's a blog post, you could see about them enabling RSS on it? If it's video content or something, they could put it on patreon with you paying them for it?

> I need an unfucked UI.

Preach. I need a UI where I don't have to click "Read more" to only find out there is one word missing. Then required to swipe back and repeat the process because someone used all 240 characters in several tweets.

This is literally what happens when your devs don't dog food. What an insane thing to do and with such an easy fix. It is just baffling. There's so many little things like this but I can't understand how this isn't just a few lines of code somewhere.

Might archive.something eventually work on twitter? (To allow quotes and fixed views for use within HN discussions?)

I’d also vote for 3) the user interface is fast, responsive, and not total garbage inflicted with man boy ego whim randomly mutating a decade of questionable product management decisions

For 1: replace twitter.com with vxtwitter.com.

Reply on news.ycombinator.com