Hacker Newsnew | past | comments | ask | show | jobs | submit | more fuckcensorship's commentslogin

Based on the sources you linked, summer interns only worked on the guitar tuner and metronome functionality. I see no indication in the second blog post you linked that the same interns (or interns at all for that matter) had anything to do with the development of the color picker, encoders, etc.


As I mature as a web developer, I become increasingly fond of simple, functional, and minimal JavaScript sites like nearlyfreespeech.net[1]. It always loads quickly, it’s easy to find what you’re looking for, and I find the simplistic styling of the site visually appealing.

[1]: https://www.nearlyfreespeech.net/


I think you'd love https://plaintextsports.com


It's missing all of the sports, lol


Should be called someplaintextsports :D


Flat Text

its not a popular opinion, but java is not neccesary to handle a webpage. if i see an unfamiliar webpage i will cURL it instead of http GET-ing it.

at this point in the game im familiar enough with code that it reads easy as, often easier than a blog.

webpages are served out to be read by people, if a webpage demonstrate primary purpose beyond/other than this, it is webkrupht.

[in adndm]

before the question if it should come to mind; Yes i am about 2secs away from giving up on browser entirely and just going back to basic CLI surfing.


> java is not neccesary to handle a webpage

I assume you mean JavaScript, which has no relationship to Java.

Unfortunately it is necessary depending on what you need the page to do. It's badly overused, it's abused by AdTech, and it's often unnecessary, but it is part of the web and some things just can't be done without it.

Forms are a good example. They're one of the core features of websites, and they're much more useful with JavaScript (live validation, dynamic show/hide of certain fields, saving progress, etc.)


yes i did mean javascript, and its not all gloom, after all linux can interpret ascii text as executable code if you let it, those were fun times when that was a prevalent issue.

interactive/reactive pages are good things when you want them.

one of the most useful things for me working on a plaintext basis is winlink

https://en.wikipedia.org/wiki/Winlink

its best to attach an flat ascii file due to the way it will probably be used, by analogue radio systems in the group.

its a ham radio thing, and is as close to whats left of the original BBS scene.

nowadays certain LoRa WAn antics will open a portal as well.

https://en.wikipedia.org/wiki/LoRa


I’ve tried using sites like Nitter and Teddit but always give up due to the sporadically crippling performance issues. I’ve tried using different instances but any improvements seemed to be temporary. My guess is that most of the performance issues are from Twitter/Reddit. Is this true or am I just having back luck with the instances I choose?


If you are using random instances run by hobbyist it will happen. Such instances come and go every week.

Try to use proper instances from projects like kavin.rocks, pussthecat, esmailelbob.xyz, bus-hit.me, vern.cc, tokhomi.xyz etc.

For eg., try https://nitter.kavin.rocks. It is one of the fastest instance and also hosts its CDN on Cloudflare.


In addition to this suggestion, another viable route is to self-host those applications you rely on and don't expose them to the world (so as to reduce load/attack surface). Using a VPN can allow you to access the applications privately/remotely.

e.g. I self-host the applications I rely on such as Teddit, Nitter, Bibliogram and Cloudtube and then use Wireguard to always remain connected to the network they are accessible on. I have also implemented identity-aware SSO[1] so I can expose those applications remotely to specific individuals.

[1] https://github.com/buzzfeed/sso


This is exactly what I was looking for. Thank you!


in libredirect, they added a feature where you can test the latency of each instances


The problem with public instances is that they're constantly overloaded. Your best bet is to run your own.


I started running my own this week because the public instances are inconsistent.

Performance is better overall, but some things are still sluggish (teddit) because you're usually the only fetching data and nothing is cached.

Paired with Tailscale makes for a good combination too.


Teddit specifically has this unfortunate behavior where every referenced subresource is loaded before a response is returned: https://codeberg.org/teddit/teddit/issues/248

Until that is resolved, libreddit (similarly self-hostable) performs much better.


Which is what I figured is the point, since it's libre. On the other hand doesn't that defeat the purpose of the privacy thing? If every request for ~one user's worth of traffic is coming from one proxy you can sort of identify one person that way.


> On the other hand doesn't that defeat the purpose of the privacy thing?

It depends - a lot of the privacy invasions go beyond just the IP address and also include browser fingerprints and on-page activity collected by malicious Javascript - these third-party frontends defeat that.


All (most?) of them also trivially supports using a proxy so you can selectively route them through Tor or some other proxy or VPN.


Heh, then who hosts that proxy? And might you want to self-host that if it gets overloaded? TOR of course makes sense here though.


My experience is that Nitter is faster than the official Twitter frontend



Not what I expected.

But why?


Are the steps different for regular Coke?


The point is that ordering a single item like this should be fast.


But is that really the main use case? Do people often go to McDonald's to order just one item?

If I wanted a Diet Coke, I'm more likely to go to a convenience store. The operation would be simpler: I pick up exactly the item I want from the shelf, pay for it, and leave. I don't need to wait on anybody to bring me anything. Especially for something as generic as a Diet Coke.

If I am at McDonald's, I'm expecting to order McDonald's food. As a designer, I wouldn't optimize for the case of somebody ordering just one item that they could get elsewhere.

That doesn't excuse arbitrary up-selling and obfuscated interfaces, but it doesn't surprise me that ordering a single item (especially a non-food item) isn't as fast as it could be.


Improving AirTag battery life is the first thought that comes to mind for me.


I don't see how that's possible. The Airtag pings the Find My network periodically anyways, sending battery life along with that data cannot possibly be that detrimental to battery life.


> cannot possibly be that detrimental to battery life

Know anything about batteries? There are other ways of determining remaining battery capacity, but most popularly it is determined by testing voltage, which itself can be split into two methods: testing voltage under load and testing voltage not under load. Testing voltage under load takes energy. Testing voltage not under load will also take energy (if the tester is built in to the device, which it is) but not as much, though testing not under load is not nearly as reliably accurate as testing voltage under load.

So, in fact, not only possibly, but definitely, checking battery life is always detrimental to battery life because it takes power to check voltage under load to produce battery life data, and that power comes from the battery's remaining capacity.


Then poll the battery at a lower frequency. Make it a cronjob that only updates every week, or a method that gets called after broadcasting device position. If Apple really wanted to empower their users, they would have added this as a low-power mode switch in the Airtag settings, but... nope.

I get where you're coming from, but Airtags have worked perfectly fine in the wild thus far. Obviously the battery testing isn't that detrimental, and since Apple hasn't made an official statement confirming that's the case, I think I have to chalk this up to apologism. Apple took away another feature, and the community can't help but find excuses. This is the whole "we need the notch for our 1080p webcam" hustle all over again...


> I get where you're coming from

It's ok, but obviously not. I was merely correcting your mistake, perhaps pedantically, and really that's all. I was not, in fact, giving you license to build a straw man factory.


Parent comment specifically mentions recording for game development so 1080p at 10fps probably isn't going to cut it.


Update it to be 20~25 fps then, still won't take ridiculous amount of space.


Unlike movies, the content on screen recordings usually remains roughly the same for minutes before switching scene. So it can achieve significantly higher compression ratio even at high fps


I wrote more about it here: https://news.ycombinator.com/item?id=32223240

But the TLDR, ffmpeg can remove duplicate frames:

    ffmpeg -i in.mkv -map 0:v -vf mpdecimate,setpts=N/FRAME_RATE/TB out.mp4


YouTube video it redirects to: https://youtu.be/1hfk8kh75ic


Another interesting thing is then when you go to the channel page of the uploader's video, they're subscribed to an account named iluomo, and if you go to that account's channels page, it keeps loading the channel itself over and over again


That channel iluomo is the 40th channel ever to be created on youtube if you believe the about page. However the channel was created on Apr 30, 2005 which could track for the 40th channel. I suspect this 666 and 69 channel redirect, and yop/iluomo, are legacy site exploits/easter eggs from the actual original developers of Youtube. yop/iluomo also had various other easter eggs through the years, like this one. https://www.reddit.com/r/TimeworksSubmissions/comments/oidgv...

Anyone else got anything interesting to add about this? edit:exploit is a better word for this, as easter egg implies something else entirely


In these kinds of situations, I prefer to just dispute the transaction through my bank, get a temporary credit for the full amount (with Capital One at least), and let my bank investigate it. Most of the time this will be the end of it but in some cases you will need to provide additional documents/info to settle the dispute.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: