I have always loved programming - its like Lego without gravity.

Basic on my ZX81 graduating to assembler and Turbo Pascal during my teens.

Developed phone OS software - engineer, architect, product manager - but got made irrelevant by the iPhone and redundant by Android.

These days I mostly work with data, big data and fitting big data onto small boxes.

Reddit vs Hacker News vs Twitter

  • Deeply technical stories might get up to about 5K visitors from  HackerNews (“HN”) and 10K from reddit/r/programming (“proggit”)
  • Deeply controversial ’Cogs Bad’ -type posts still get under 10K from HN, but can get 40K+ from proggit
  • Hacker News is a much stabler community; if you have posted stories in succession, the same 5K+ visitors coming back each time
  • Reddit is a flash crowd; they claim 350K+ subscribers, but that number doesn’t mean much; very few clickers on any given story are repeat
  • If a story gets on the front page of both at the same time, you get might get more people from HN than proggit
  • But the story dies a lot sooner on HackerNews; proggit can keep people streaming in for a whole day
  • Being very top spot on HN is a dramatic visit booster; you can watch in Google Analytics Real Time the sudden influx of visitors when you get sorted to top; and then a trickle again as you drop to second place; Google Analytics Real Time is gamification for blogging
  • Google Analytics doesn’t count people using noscript (when there are so many technical avenues they could pursue instead)

Very very few people who read a blog post go back to the link aggregation site and vote on it:

  • About 100 to 1 visitors to points seems about right; if a story has 100 points, its likely 10K people read it.  That’s right; two orders of magnitude!
  • West coasters up-vote less than Europeans; maybe they are busier?  Or my content is more pleasing to Europeans?
  • The per-page time is actually quite high; often over 5 minutes on average for a lot of my stories even when its a big crowd.  I suspect some subtle misreading of Google Analytics on my part, or else people might actually be reading the whole thing carefully!?
  • I speculate that most HN readers are not actually logged in

Most compliments are never given:

  • If someone likes a post, they probably say nothing
  • A comment on the blog is going to be very polite and constructive
  • A comment on HN is going to be mature and reasoned; often expanding or exploring technical issues raised
  • People comment on proggit to complain about a story
  • Never bother mentioning Haskell.  That crowd has this circular argument going ;)

Its lovely to read the tweets:

  • A tweet is the biggest compliment of all; 140 chars to summarise the post really makes it obvious that the tweeter groks it; and its almost never negative
  • Nobody actually follows the links in tweets though; click-through is often in the low digits per tweet

This is like search-engine-optimisation; now go shower!

  • The ideal proggit time is Sunday daytime in the states; it gets to the top for Europe’s Monday, and is still up there for stateside Monday
  • The ideal HN time is a weekday European afternoon; discerning European votes can get the post onto the front page in time for a stateside morning
  • Proggit, stories get stuck in spam filters.  Then it often gets mercilessly down-voted by people not leaving comments and so it dies
  • On HN, stories start in the ’newest’ fire-hose; the click rate here is in the low double digits, but these early-clickers are far more likely to go back and up-vote, so a story can gather the 3 or 5 votes it needs to get on the front-page.  Once on the front page, its onwards and upwards

Naturally, I turn to technology to tame coder social ills: self-organising link aggregation.

Meta:

  • as seen on HN
  • as seen on reddit (not proggit yet)
  • need I repeat that thing about how sad it is that nobody up-votes things they like reading? ;)

jump to ↓



performance
Faster searches with non-prefix fields in composite indices
Compressing MySQL databases
What highscalability.com says about Scaling my Server
Scaling my Server: follow-up
old classics
The kid's computer
Making the History of Worlds Religions map
If you defend those involved in the OpenGL ES specification, you are an idiot
Stackoverflow unwinding?
general
Why Swift?
Python annotations and type checking
pycon 2014 Sweden: the bad bits
Table-based Template Translation in C++
recreation
games programming
Perlin Noise
Perlin Noise
Drawing RTS maps fast
WillCity update
ludum-dare
Ludum Dare #35 Mosaic
LudumDare 33 wallpapers
SSIM vs MSE for Mosaics
Ludum Dare 30 results are in!