Item 10^6 | Hacker News Reader

Item 10^6

news.ycombinator.com

107 pointspg45 comments

Open articleView on HN

Thanks for building this site, Paul.

I was secretly hoping pg would hack things so that 10^6 was just skipped ;)

I considered it, but item ids are so deep in the code that it would have been asking for trouble.

A one-line change to new-item-id is asking for trouble?

A one-line change to something which probably has assumptions all over the rest of the code is asking for trouble, yes.

I actually read the code, and that's not the case.

OK, while we're on the topic of useless numerology:

How old are you in HN years? (ID of your most recent post / ID of your first post)

me: 1000140 / 174632 = 5.7 HN years

edw519 (current leader after pg): 999538 / 9357 = 107 HNyears

RIP, nickb: 685017 / 2649 = 259 HNyears

mechanical_fish: 999416 / 22714 = 44 HNyears

is there somewhere else where this convention is used?

if the Id of the most recent post is an exponentially growing function of time then the time between any two posts should be proportional to the log of the ratio of their IDs.. so

shouldn't it be log(ID most recent / ID first post) ?

I made it up.

But for exponentially growing ID, log(ID most recent / ID first post) is proportional to age of account & that's boring -- the nonlinearity is what makes the measure 'fun' (I use the term loosely). It agrees with some notion of the 'good old days' when Startup News was a fishbowl we could all look in on, and early HN participants were Founders not only of startups but also an online community.

I agree though, there are probably more interesting measures one could come up with.

me: 998366 / 243 = 4108 HN years

There may be earlier ones - I couldn't retrieve comments before about Jun 2007, so I went by my earliest submission:

http://news.ycombinator.com/item?id=243

I used searchyc to get earliest comment ids.

Does that go back all the way? I can only find entries back to Sept 2007, but seems like I was on before that.

Or perhaps my memory is going. They say the mind is the first thing to go. Or at least I think they do.

Well, http://searchyc.com/pg?sort=by_date_r has pg's first post (ID = 1) and what I assume is his first comment (ID = 17).

I lurked for a while before participating, maybe you did the same?

Incidentally that makes pg 1000205 / 17 = 59,000 HNyears old -- and that seems about right, I bet the average reader here cares ~10,000x more about what pg writes than what I write (pg HNyears / davi HNyears = ~10K). :)

When was it confirmed that nickb had passed?

I was referring to the account, not the person. Sorry for the ambiguity.

Only some trolls think Nickb is dead. The best solid evidence for him not being dead is that even if he doesn't answer his cell phone, he seems to be paying the bill. Though I suppose that's not very good evidence with automatic payment being so common.

Every so often in the news you read about some elderly person in one of the developed countries found dead in their apartment, having been dead for months. Their pensions keep getting direct-deposited and their bills keep getting paid automatically so no one suspects anything for awhile.

The most extreme case I've heard of: http://news.bbc.co.uk/1/hi/scotland/edinburgh_and_east/81326...

This tangent is not a pleasant one, but I couldn't resist making a citation.

20 years old again! Woo hoo!

1000225/1809 = 552.91

3.7 HN years. Probably after this post a little bit more.

997143 / 600 = 1 661.905

"... How old are you in HN years? (ID of your most recent post / ID of your first post) ..."

999899/425 = 2352 HNY

latest: http://news.ycombinator.com/item?id=999899 first: http://news.ycombinator.com/item?id=425

At least the comment was original.

would have been nicer if the post was a little bit more programming related.

Since it's somewhat an anniversary, can you tell us what kind of traffic HN is getting? users wise, pageviews wise. Or is that a trade secret?

We now get around 40k unique ips and 500k pageviews on week days. Here's the traffic as of last night:

http://ycombinator.com/images/hntraffic-16dec09.png

The reason pageviews peaked in Aug is that we got better then at shutting down over-aggressive crawlers. The big spike in uniques was when the HN thread about _why became the de facto condolence page for the whole web.

those numbers are pretty good, although the ratio is lower than I expected. 12 page views per user, per month? I view more than that while I eat my breakfast.

Those numbers are per day, not per month.

Congrats on the growth. I think you're going through some growth technology problems as well?. I find the responsiveness of HN has changed for the worse. I notice the main page doesn't respond sometimes and trying to check out the submissions for top leaders / heavy users yields no results. Just an observation.

Now we look to the 2^20 day:

http://news.ycombinator.com/item?id=1048576

Still a loooooong way to go... but the item IDs aren't limited to 32 bit ints are they? ;)

Are you kidding? HN is written in Arc, not C++.

Oh no, we've brought religion into the mix!

Once you start looking for "interesting" ids you can find all sorts of reasons to find them. 2^20, the next prime, Fibonacci numbers, etc.

Perhaps for any given point in a increasing number series there is an infinite number of interesting numbers just around the corner.

Sigh.

It's the sieve of Giraffes

> It's the sieve of Giraffes

Or the http://en.wikipedia.org/wiki/Interesting_number_paradox

Every number is interesting somehow: http://en.wikipedia.org/wiki/Taxicab_number - and the quote: "Every positive integer is one of Ramanujan's personal friends."—J. E. Littlewood, upon hearing of the taxicab incident.

From Littlewood's (wonderful) "Mathematician's Miscellany":

I read in the proof-sheets of Hardy on Ramanujan: "As someone said, each of the positive integers was one of his personal friends." My reaction was, "I wonder who said that; I wish I had." In the next proof-sheets I read (what now stands), "It was Littlewood who said ..."

(What had happened was that Hardy had received the remark in silence and with poker face, and I wrote it off as a dud. I later taxed Hardy with this habit; on which he replied: "Well, what is one to do, is one always to be saying 'damned good'?" To which the answer is "yes".)

Until now I've always taken these paragraphs as saying that it wasn't really Littlewood who made the remark about Ramanujan, but on rereading it I've changed my mind: I think the point is that he did say it, and was annoyed that Hardy didn't credit him, and made a little joke about it which Hardy interpreted correctly. Maybe I'm overinterpreting. Littlewood's second paragraph doesn't quite make sense to me (i.e., what it proffers as an explanation doesn't seem to explain very well) on either interpretation. If he didn't say it, then (having seen that second set of proof sheets) he and Hardy effectively conspired to credit him for something he hadn't said, which would be quite a breach of academic etiquette...

I've never heard "sieve of giraffes" before, and I'm coming up empty in searches for it. Can you explain?

Sure thing.

It's a word-play off three concepts. The First is the famous filter for finding primes, the Sieve of Eratosthenes. Sieve of Eratosthenes

Then there is the concept of "Interesting Numbers" http://en.wikipedia.org/wiki/Interesting_number_paradox (every number is interesting).

Finally RiderOfGiraffes' nickname, since they've demonstrated to me that once you start looking at those little commentid fields it does interesting things to my obsessive-compulsive tendencies.

Sorry so obtuse. My head's been down in code all day long. If I could post in binary it might be easier :)

Well, I figured out the first and third parts. So the the Sieve of Giraffes returns every number. I guess the next step is to include it in the Java library.

Can you estimate how quickly 2e6 will occur at the current growth rate? I wouldn't be surprised if it happened in 2010.

Based on fitting a simple logarithmic curve, plus some fiddling, I'm thinking 2e6 will happen in late September, 2010.

sadly i did not have time to test my script and there was a bug or i would have for sure had it :(

You can take solace in the fact that you just had your 1K karma day :-)

You can't use item numbers as an index of activity.

http://news.ycombinator.com/item?id=87846

1000000 GET

Reply on news.ycombinator.com