4 ms·

How can they say they're the fastest growing when Popcorn Time nearly doubled the unique contributors (95-ish to 40ish) in nearly the same time? Seems rather d

How can they say they're the fastest growing when Popcorn Time nearly doubled the unique contributors (95-ish to 40ish) in nearly the same time? Seems rather disingenuous to include PopcornTime clearly showing it's much younger than CodeCombat, showing at the same age of being open-source it has nearly double the unique contributors, then claiming it's growing slower than CodeCombat. Don't get me wrong, what CodeCombat is doing is great and all the well wishes to them, but I don't think they're interpreting the data very fairly.

stonogo 13y ago
The Fastest-Growing Open Source Project As Determined By Our Incredibly Specific and Sometimes Downright Strange Metrics
- nwinter 13y ago
  Unique contributors since open source launch is strange?
  - stonogo 13y ago
    "Unique contributors within inital 86-day window of project upload to Github" is not equivalent to "unique contributors since open-source launch."
    - schmatz 13y ago
      You're right, that's true. It's not the ideal metric, but it's one that is realistic to compute with limited time and resources. To analyze that statistic using a growth metric like "largest 30 day average contributor gain" across such a gigantic dataset, consisting of hundreds of millions of rows and tens of millions of unique repositories, requires resources and engineer time which few people possess. We certainly don't want to divert too much of either of those things away from development of CodeCombat just for a fun data science experiment!
      - stonogo 13y ago
        So you narrowed the problem set down to one that was sufficiently effortless, then uploaded a blog post declaring yourself world champions. I'm not sure that makes anything better. "Fastest-growing open-source project, as far as we care" is a perfectly fine headline. What you've got now is disingenuous and baselessly self-congratulatory.
schmatz 13y ago
Hi mkal_tsr, you have a valid point there! Popcorn Time is awesome, and if it continues (unlike the other Popcorn Time repository) it will eventually pass CodeCombat. However, it might also experience some event or plateau, like Pullup did, that will stop its growth. A really interesting way to analyze the GitHub Timeline would be to rank repositories based on sustained average growth rate. We chose a fast, (sort of) cheap, and computationally easy way to run these statistics, but there are a myriad of other ways to process the same dataset, which we'd definitely be interested in seeing.

stonogo 13y ago

The Fastest-Growing Open Source Project As Determined By Our Incredibly Specific and Sometimes Downright Strange Metrics

nwinter 13y ago
Unique contributors since open source launch is strange?
- stonogo 13y ago
  "Unique contributors within inital 86-day window of project upload to Github" is not equivalent to "unique contributors since open-source launch."
  - schmatz 13y ago
    You're right, that's true. It's not the ideal metric, but it's one that is realistic to compute with limited time and resources. To analyze that statistic using a growth metric like "largest 30 day average contributor gain" across such a gigantic dataset, consisting of hundreds of millions of rows and tens of millions of unique repositories, requires resources and engineer time which few people possess. We certainly don't want to divert too much of either of those things away from development of CodeCombat just for a fun data science experiment!
    - stonogo 13y ago
      So you narrowed the problem set down to one that was sufficiently effortless, then uploaded a blog post declaring yourself world champions. I'm not sure that makes anything better. "Fastest-growing open-source project, as far as we care" is a perfectly fine headline. What you've got now is disingenuous and baselessly self-congratulatory.

schmatz 13y ago

Hi mkal_tsr, you have a valid point there! Popcorn Time is awesome, and if it continues (unlike the other Popcorn Time repository) it will eventually pass CodeCombat. However, it might also experience some event or plateau, like Pullup did, that will stop its growth. A really interesting way to analyze the GitHub Timeline would be to rank repositories based on sustained average growth rate. We chose a fast, (sort of) cheap, and computationally easy way to run these statistics, but there are a myriad of other ways to process the same dataset, which we'd definitely be interested in seeing.