3 ms·
It's simply the standard "Power Law" distribution kicking in with wide error bars. Note that #8 SQL is already under 1.5%. The error bars look like roughly 0.
by bsder 3y ago
It's simply the standard "Power Law" distribution kicking in with wide error bars.
Note that #8 SQL is already under 1.5%. The error bars look like roughly 0.3%-0.4%. Given that, the ordering of everything below the top 10 has a very high random chance to it.
- hn_throwaway_99 3y agoI doubt it. The TIOBE index not only feels quite "wrong" to me, but it also differs greatly from other surveys I've seen on programming language popularity. I don't think they announced their methodology (or at least I didn't see any details), but my bet is it's more likely that their methodology is simply giving bad results because they're weighting some pieces of data too heavily (or conversely not enough) when calculating their ratings.