randomhackernews.com
I made this for myself, but figured others may find some interest in it.
Please realize that it isn't an exact representation of the front page at any particular time. For that you'd want a point, not an interval, and certainly not a 24 hour interval, which is what /front?day=yyyy-mm-dd is showing.
What's it called when you make a single photograph by overexposing a series of time-lapsed frames into one? It's like that. I found it trickier than I expected to write the code to do this, but I think the result serves its purpose: to give people a way to find frontpage stories they may have missed.
We do this via a combination of votes and frontpage time, but only back to 2014-11-11 which is when we started logging the lists of stories on the front page. Those are the 'frames' in the above analogy. Before 2014-11-11, we only rank by votes because it's all the data we have.
If you want to look at actual snapshots from the past, archive.org has a lot of them, e.g. https://web.archive.org/web/20160620131548/https://news.ycom.... Perhaps OP could add random links to those as well!
Long exposure
That’s what you have if you mechanically keep the shutter open for a prolonged time, exposing either film or a sensor to light.
But from what dang was saying, he seemed to describe emulating that effect in software on discretely captured frames.
Sounds almost closer to onion skinning, and in particular the part of this article https://en.wikipedia.org/wiki/Onion_skinning where they say:
> This effect can also be used to create motion blurs, as seen in The Matrix when characters dodge bullets.
IIRC, the front page before that was filled with stories about _why (they're split between https://news.ycombinator.com/front?day=2009-08-19 and https://news.ycombinator.com/front?day=2009-08-20 now), and people were complaining about that. I happened to have an Erlang article I'd already been meaning to post, so I posted it and it triggered a deluge, followed by complaints:
Ask NH: How annoying is Erlang day? - https://news.ycombinator.com/item?id=776789 - Aug 2009 (0 comments)
Ask HN: Why is all the news about Erlang today? - https://news.ycombinator.com/item?id=776622 - Aug 2009 (4 comments)
Are we celebrating Erlang's birthday or just that Erlang hackers cracked HN? - https://news.ycombinator.com/item?id=776519 - Aug 2009 (4 comments)
Poll: What do you think of Erlang on Hacker News? - https://news.ycombinator.com/item?id=776427 - Aug 2009 (3 comments)
Enough Erlang - https://news.ycombinator.com/item?id=776316 - Aug 2009 (9 comments)
Purge Complete: No more Erlang, _why, or Zed - https://news.ycombinator.com/item?id=776196 - Aug 2009 (25 comments)
I hate the Erlang days around here..... - https://news.ycombinator.com/item?id=776116 - Aug 2009 (9 comments)
Ask HN: Are we intentionally voting on Erlang articles? - https://news.ycombinator.com/item?id=776047 - Aug 2009 (6 comments)
... followed by pg, after a long day at Demo Day, manually killing all those threads at dinner using his iPhone. Nobody knew about that latter bit.
Ask HN: How did the Erlang articles disappear on HN yesterday? - https://news.ycombinator.com/item?id=778319 - Aug 2009 (11 comments)
I always wondered why I couldn't find any trace of this in the archives, as I distinctly remember the scolding I got! I've gone and unkilled them now, which should restore roughly what it looked like.
Unfortunately, jumping across timelines reminds us of digital decay, a very real thing, when you try to visit some old stuff.
javascript:(()=>{d=new Date(1171843200000+Math.random()*(Date.now()-1171843200000));window.location.href=`https://news.ycombinator.com/front?day=${d.getFullYear()}-${('0'+(d.getMonth()+1)).slice(-2)}-${('0'+d.getDay()).slice(-2)}`;})()+
> [US president] calls for municipal broadband
+
> A Quick Comparison of [...] vs. Rust
+
> Show HN: I built an iOS remake of that old [...] game [...]
+
> [...] Tops [...] as World’s #1 [...] Platform
+
> Student [...] AI [Competition] [Year]
+
> Bitcoin crashes over [X]% in 24 hours, under $[Y]
+
> Frameworks are for hacks, libraries for seeds
(Of course for some of the above it would likely be trivial if you just copy-paste them in a search engine...)
The first post I found (from 2012) was supposed to be the letter that Apple published on its website apologising for how terrible the new Apple maps app was after its divorce from Google. But Apple has removed that from their website now, so it was just a 404 - an auto fetch from the Internet archive would have been very handy.
> Why Startups Don't Condense in Europe
> DabbleDB -- web-based DBMS coded in Smalltalk
dead comment: > Fedora is far better than ubuntu
Link rot is always disappointing, but I actually expected it to be far worse for links from 17 years ago.
And then https://news.ycombinator.com/front?day=2007-02-19, as you say, which was the public launch. Let me dig up the link...here it is: https://news.ycombinator.com/announcingnews.html
And then https://news.ycombinator.com/hackernews.html 6 months later.
I posted links to Startup News pages the other day: https://news.ycombinator.com/item?id=39237746.
We should make the top left say "Startup News" for /front?day=2007-08-13 and earlier!
[0] https://web.archive.org/web/20070222125635/http://ycombinato...
[1] https://hn.algolia.com/?dateEnd=1707004800&dateRange=custom&...
( :p, but also RIP TT )
2009/July
It's not hacker, it's not news, it's not innovative, it's simply boring, and would be cool if a 10 year old did this. Sorry for this rant, but I'm reading on and off HN over ~10 years and contributing under various user names, and I read comments like mine over and over again, and each time agreeing more.
Maybe everyone who definitely will downvote my post should read the first guideline "What to submit":
On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.
A random() over the time HN exists, with a link to the generated date, which doesn't even handle back correctly is not something good hackers find interesting.
On the other hand, there are still lots of on-topic, curiosity inducing topics around here. Just getting scarcer every year. Oh my. Good night :)
If anyone is interested in any of these: https://news.ycombinator.com/front?day=2011-09-09
It's fascinating, too, to see some predicted solutions to problems discussed at the time that came true but were different from what was expected. For instance, we finally got an alternative to JavaScript in the browser recently, but it's not CLR, JVM, or NaCL like many imagined; instead, we got WebAssembly, which is pretty good but has basically failed as a true alternative to JavaScript for now.
Chrome: https://chrome.google.com/webstore/detail/hacker-news-previe...
Firefox: https://addons.mozilla.org/en-GB/firefox/addon/hacker-news-p...
"Yes Jim, my project did in fact hit the front page - take a look."
Seems like it would be such a useful data source for the rise and fall of various tech trends, industry sentiments, and so on.
The pulse of the Valley, even...
1. HN Lucky Dip
2. HN Roll
Does anybody want to give it a try?
May I ask why were you surprised? Which other website has this feature?
It's not pleasing to go back to use it again.
This feels backward to me. I don't do user research, but is that a real pain point? Just... close the old tab?
- I'm tech savvy enough and want to open in the same page: you are preventing me from doing it easily, while it's easy for you to open it in a new tab if you want it this way.
- I'm not tech savvy. I barely know to use a browser. The link opened in a new tab and I didn't notice. The back button I know to use is greyed out. I'm confused.
Said differently: you can always open normal links in new tabs. I do it all the time, to the point I almost never understand someone requesting a link to open in a new tab: help yourself, it's so easy. Don't rely on the page author to make the call for you because they can't please everyone. However, target=_blank links are harder to open in place, while it's reasonable to want and expect a link to behave normally. I get a bit mad when for some reason a new tab was opened and I didn't want to nor I expected it. Just leave me alone, you know. If I don't want to come back to the previous page, or if I want to do it using the familiar back button (or gesture), you should not force me to close the previous tab and break my history buttons.
Moreover, nothing usually tells you that a link is going to behave differently as you expected (principle of least surprise), and even if it does, I shouldn't have to install an extension or fiddle with the web dev tools to work around this and get the default and expected behavior, that's terrible UX for obvious reasons.
Deviations from the default behavior are sometimes better / necessary but you always risk causing surprise and thus confusion, which you should also avoid for good UX. Even someone who want to open the link in a new tab probably expects a link to open in place. It's usually impossible to know in advance if users will prefer opening a particular link in a new tab, especially since different users will have different preferences in the general case. As a user who will sometimes prefer a new tab, it's also uncomfortable to rely on the author of the page to make the call so you may as well just help yourself.
So, for good UX, the normal behavior is supposed to be the best option, unless in specific cases like those discussed in [1], basically when the confusion or frustration caused by the normal behavior is greater than for the new tab behavior, like loss of work or disrupted playback.
I don't know if there are proper studies about links specifically, but I'd expect existing research to prove the principle of least surprise, generally accepted as basic UX principle. I'm a bit lazy to find such work, on a phone during a small insomnia, I hope you'll forgive me. A "(opens in a new tab)" tag partially fixes the issue but doesn't fully address it.
Thanks for making that!
- How do you store data to last more than 50 years (diamonds and titanium were mentioned)
- The announcement that bitcoin reached 1 USD, a great party story (TL;DR I tried buying 100, had issues w/ payment, got pissed, bought 200 instead, lost wallet a few months later, would have sold it super early tho)
Your website should send a 307 Temporary Redirect to HN, so when I go back, I return to where I was, instead of being sent to another random day.
If it is done on the client, look into `window.location.replace`: https://developer.mozilla.org/en-US/docs/Web/API/Location/re...
Don't Microsoft's support pages [1] do this as well? I just checked with a random support page [2] and it looks like the middle redirection url is the same. Is that a different thing?
[1] support.microsoft.com
[2] https://support.microsoft.com/en-us/windows/fix-file-explore...
(Firefox if that matters. Currently on Linux, has happened on Windows)
Edit: I just checked on Ungoogled Chromium and it looks like this (the microsoft support) isn't an issue. It's only on Firefox. That's bizarre.