Rosetta Code:Village Pump/Whoa! 10000 examples!: Difference between revisions
Rosetta Code:Village Pump/Whoa! 10000 examples! (view source)
Revision as of 18:35, 14 September 2018
, 5 years agoEdited quote to be more accurate, grammar, spelling fixes
m (updated the total number of Rosetta Code programming examples/entries.) |
Thundergnat (talk | contribs) m (Edited quote to be more accurate, grammar, spelling fixes) |
||
| (3 intermediate revisions by 3 users not shown) | |||
Line 18:
</strike>
<br>As of August 11,
== [[wp:Lies,_damned_lies,_and_statistics|Lies, damned lies and statistics]] ==
It is almost impossible to get an accurate algoritmically determined count of programming examples. The number cited above ('''60,282 as of 08/13/2018''') is inflated by including all categories related to a programming language, not just programming examples.
The numbers on the [[Rosetta_Code/Rank_languages_by_popularity/Full_list]] are obtained by directly querying the site "For each programming language, how many programming examples are there?" ('''59,366 as of 09/11/2018'''), so that is an "accurate" count insofar as it is the one that the site agrees with, but that suffers from it own inaccuracies.
There are several language groups which are counted as separate languages but only have a single entry "credited" to both languages. [[Icon]] / [[Unicon]] is perhaps the clearest example. Each has 587 (currently) tasks listed under their respective language page, but each has only a single example under each task credited to both. [[Mathematica]] / [[Wolfram_language]], various flavors of Basic and other family groups of languages also have some overlap, so some examples are counted twice, inflating the actual number.
Ok, If I look at the totals from [[Rosetta_Code/Count_examples/Full_list]], how about that? Well... yes and no. The code generating that page actually visits every task page and manually counts the number of task examples actually on each page, and is nominally the "most" accurate, ('''58,101 as of 09/11/2018''') but even it totally glosses over the fact that many languages have multiple solution examples for a particular task. This count only checks that a programming language '''has an entry''' on the task page.
''Technically'' it could be argued that each of the multiple solutions should count as a programming example because ''technically'' they are. Trying to accurately count them would be extremely challenging though. Things inside a <nowiki><lang whatever></lang></nowiki> block may be a complete example or may just be a fragment of a larger program, or may be something else entirely. Due to the variability in how each task author has formatted their various entries it is nearly impossible to (accurately) programmatically count them.
What does all of this mean? Not much, I suppose. It is just something I have pondered on a bit and decided to share my observations.--[[User:Thundergnat|Thundergnat]] ([[User talk:Thundergnat|talk]]) 17:43, 14 September 2018 (UTC)
:59K +/-2% or so? [[User:Paddy3118|Paddy3118]] ([[User talk:Paddy3118|talk]]) 18:20, 14 September 2018 (UTC)
| |||