Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Inside a viral website

notfunatparties.substack.com

Post reply on HN
Print view View original

281–290 of 305 posts

Re: Inside a viral website

#281

Post by tomglynch »

quicksnap wrote: post #168
l00sed wrote: post #157

I'd love if someone could point me in the direction of WHY this Next.js setup "just works". This was a great article, and the author described the sensation of getting traffic very well. I've been on the other side of the coin where my apache server crashed or 50* error'd out. It would be interesting to know more about what makes this stack or this server architecture "just work" under the hug of death.

I've been working with Next for a couple of weeks, here's my take: - Next.js aims for "pit of success" in building sites with React that are optimized for CDNs and small builds - Deploying said sites to Vercel takes advantage of the CDNs, so most of the users are getting cached data. So, for this site, I would assume: - There is no backend - All the HTML is cached and contains a static generation of the page(s) - The…

Would it being cached mean that when the ship became unstuck, and the website changed from 'Yes' to 'Sort Of' or to 'No', that people who were checking back would receive an incorrect cached version?

Re: Inside a viral website

#282

Post by noahtallen »

tomglynch wrote: post #281
quicksnap wrote: post #168

Earlier quoted context omitted.

I've been working with Next for a couple of weeks, here's my take: - Next.js aims for "pit of success" in building sites with React that are optimized for CDNs and small builds - Deploying said sites to Vercel takes advantage of the CDNs, so most of the users are getting cached data. So, for this site, I would assume: - There is no backend - All the HTML is cached and contains a static generation of the page(s) - The…

Would it being cached mean that when the ship became unstuck, and the website changed from 'Yes' to 'Sort Of' or to 'No', that people who were checking back would receive an incorrect cached version?

The new copy is pushed to the CDN pretty quickly. I don’t know much about vercel. But the benefit of CDNs is that there is a 1:N relationship between you and the edge servers, and then an N:M relationship from the edge servers to the users.

So there are a lot more servers much closer to people using your site. Pushing a new copy from “you” to each edge server is pretty straightforward. I imagine Vercel does it pretty quickly, but sometimes the edge server could poll for changes maybe once a minute, which is pretty cheap to do.

Re: Inside a viral website

#283

Post by selcuka »

ryantgtg wrote: post #264
selcuka wrote: post #262

Earlier quoted context omitted.

What we need now is a donation service aggregator. Any donations for this idea?

Take my dollar! Seems like linktr.ee does a good job of aggregating links. So, the "donate" link could link to that, which then displays a bunch of options. However, I believe that a familiar icon - like the patreon icon - helps people make decisions.

My idea is the other way around, it does not require the content providers to change the way they want to get paid.

I was thinking of an aggregator for the donator, i.e. something that aggregates your multiple accounts on multiple donation services.

Re: Inside a viral website

#284

Post by chrispickels »

kbenson wrote: post #258
cstrat wrote: post #253

Earlier quoted context omitted.

I also pay for tarkov-market :D It is almost a must if you're a casual player. It is too hard to keep up with the barters, flea market rates and even just the quests and items for quests.

Yep. :) I signed up a couple months ago for historical market data because I'm a sucker for that, but the added features since then have really cemented it as a good decision. I haven't played in a few weeks though, from a combination of the friend I usually play with being less interested (and not wanting to level/quest much without him) and being busier myself. That's sad, because it really is pretty amazing once y…

Any ideas how he is collecting the data? They are extremely on the ball about banning tradebots in game.

Re: Inside a viral website

#285

Post by T-hawk »

retsibsi wrote: post #279
T-hawk wrote: post #233

Earlier quoted context omitted.

Reddit gold makes tons of sense. People can pay money to amplify the opinions they want amplified. Reddit actually monetized internet arguing.

Does gold affect the ranking of a comment (i.e. where it is displayed on the page)? I guess I never really thought about it, but I assumed not. If it's all about the icon displayed next to the comment header, I suppose that's a kind of amplification -- but it seems like less of a zero-sum game, because it's not forcing another opinion out of view.

Yes, it does. I think most of the various silly awards they have nowadays don't do it, but at least the original Reddit Gold award before all that other stuff does elevate a post or comment.

Re: Inside a viral website

#286

Post by kbenson »

chrispickels wrote: post #284
kbenson wrote: post #258

Earlier quoted context omitted.

Yep. :) I signed up a couple months ago for historical market data because I'm a sucker for that, but the added features since then have really cemented it as a good decision. I haven't played in a few weeks though, from a combination of the friend I usually play with being less interested (and not wanting to level/quest much without him) and being busier myself. That's sad, because it really is pretty amazing once y…

Any ideas how he is collecting the data? They are extremely on the ball about banning tradebots in game.

AFAIK he's using RatScanner[1], or something like it. I'm pretty sure the admin of RatScanner is in the Discord quite a bit, but that could be a coincidence (but I think it's some form of image processing lib no matter what, they include a screen cap of the latest market info for an item on the page when you pay).

I'm not sure I would consider what they do effective in banning trade bots (the captcha seems ridiculously easily beaten), I think the Found In Raid object flag is probably the really effective thing keeping most bot usage down, since it attacks the actual incentives of using a bot (it's still useful to auto-buy when markets dip or someone under lists an item, but that's about it I think, but I may be missing a use).

That said, if all you're doing is scanning the market results and not buying anything, it's probably much easier to not trigger any problems with the devs. The market data is really useful for those that like that sort of thing, but it's not really that much of an advantage, just a way for those into the economic portion of the game to go deeper.

1: https://github.com/RatScanner

Re: Inside a viral website

#287

Post by cstrat »

chrispickels wrote: post #284
kbenson wrote: post #258

Earlier quoted context omitted.

Yep. :) I signed up a couple months ago for historical market data because I'm a sucker for that, but the added features since then have really cemented it as a good decision. I haven't played in a few weeks though, from a combination of the friend I usually play with being less interested (and not wanting to level/quest much without him) and being busier myself. That's sad, because it really is pretty amazing once y…

Any ideas how he is collecting the data? They are extremely on the ball about banning tradebots in game.

It’s all done just by image processing. Once premium you can see the screenshots of the items on market used to provide the price estimate. They must have a couple accounts running continually going through the market checking prices. That’s why some items are 2-3 hours delayed.

Re: Inside a viral website

#289

Post by nicbou »

bkanber wrote: post #86
TeMPOraL wrote: post #55

One thing I don't understand about monetizing on the Internet. I see the author considered ads, tried an affiliate scheme and even did an NFT. But what about plain old, honest to God payment? Couldn't they just put a paragraph of text on the site, "Hey folks, I'm worried about hosting costs, if you find this page useful, please [Donate]", with a button redirecting to Paypal or Stripe or whatever, collect the money an…

For the last 10 years or so I've been running a free-to-use Morse Code online radio thing. It has cost me about $1000 to host so far, and has a couple of hundred DAUs. The users are not shy about feature requests, and they're not shy about complaining that I haven't implemented their feature requests. So I made a pledge to the userbase that if they collectively donate $1000 to the site, I'll start working on v2. I po…

I can confirm. I get 1-2 donation for every 50,000 visitors. It doesn't even cover for the time I spend answering their emails. There's a donation prompt at the end of my email answers. Those are people who start their emails by saying how valuable the service was to them.

Donations are about 1% of my income. They are a complete waste of time.

Re: Inside a viral website

#290

Post by nicbou »

worik wrote: post #207
bkanber wrote: post #86

Earlier quoted context omitted.

For the last 10 years or so I've been running a free-to-use Morse Code online radio thing. It has cost me about $1000 to host so far, and has a couple of hundred DAUs. The users are not shy about feature requests, and they're not shy about complaining that I haven't implemented their feature requests. So I made a pledge to the userbase that if they collectively donate $1000 to the site, I'll start working on v2. I po…

"Asking for donations doesn't work" Depends. I have paid for things I could have gotten for free, and raised money for political organisations (not in the USA, so we could not sell laws the way the Americans do). The best way to get money is to ask. There is a lot of research into the best ways to ask, and what gives best results. But just asking thoughtlessly (not accusing you, no idea how you asked) does not work

That doesn't prove that donations work. It only proves that you sometimes donate. Most people don't donate to most projects.
Post reply on HN