On Wednesday, May 28, 2014 7:41:53 PM UTC-5, Terry Reedy wrote:
> Claim: "Python 3 languishes in disuse."
> Fact: in 2013, there were around 14 million downloads of
> windows installers for each of 2.7.x and 3.3.x. 3.3 is
> over twice as popular as 3.2 (to be expected).
Terry, you cannot simply take the download numbers as a
"one-to-one" ratio representing the "actual" usage of
anything.
Consider:
* HOW MANY DOWNLOADED AND NEVER USED?
Might be difficult to believe, but i am one of those people.
Even though i don't "actively" write code for Python3, i
still keep a current version on my machine just in case i
need to test a code snippet for a NOOB -- i don't remember
the last time i even ran Python 3, must have been a year or
more!
* HOW MANY DOWNLOADED, TRIED, CURSED, AND NEVER USED
AGAIN?
* HOW MANY DOWNLOADED, TRIED, CURSED, AND ONLY USE
BECAUSE THEY ARE FORCED?
* HOW MANY DOWNLOADED, TRIED, WERE "AMUSED", BUT NOT
ENOUGH TO ACTUALLY WRITE 3.X CODE?
* HOW MANY DOWNLOADED, TRIED, LIKED, AND ARE "ACTIVELY"
WRITING 3.X CODE?
I would venture to say that at least half the downloads fall
into the first three categories, and the other half are split
between the last two categories -- equally split, WHO KNOWS!
Sheesh! Metrics measured on download statistics are about as
reliable as that ridiculous TIOBE drivel, hey, watching
Python climb the index might give GvR a raging semi, but
only a moron would take it for truth!
PIPING HOT CUP OF WISHFUL THINKING ANYONE?
--
https://mail.python.org/mailman/listinfo/python-list
