‘TIOBE’ does appear to be based on "verifiable (or at least falsifiable) figures"...
No one's been able to reproduce them. (They confused a web browser with a language implementation for several months!) They're based on bad assumptions. No one's done TIOBE better because it's meaningless to do the wrong things better. See TIOBE or not TIOBE – “Lies, damned lies, and statistics” for one explanation.
Also, unlike science, they don't release raw data sets.
| [reply] [Watch: Dir/Any] |
Cheers for that. :)
I've never looked into TIOBE because I'm not particularly interested in it, but it sounds like it might be just what it sounds like it might be...a very crude metric, and obviously the input of "skilled engineers world-wide, courses and third party vendors" amounts to "because they use search engines too" and not that this criteria receives any special weight beyond the special weight of a bait and switch sales pitch by TIOBE.
Perhaps the "hayday" that JavaFan refers to needs a clearer definition. If it means, occupying the avante garde spot for dynamically typed, GC'd, runtime interpreted languages, maybe so, but that's because subsequently there became a bandwagon for such beasts, and the possibility of a singularity disappeared. Once the road's been paved, the glory of trailblazing is over.
But in any case: viva la renaissance ;)
| [reply] [Watch: Dir/Any] |
Even if you believe in tiobe, this graph does not support JavaFan's taunt that "the haydays of Perl where in the mid to late 90s" [sic]. That graph indicates that "Perl's tiobe haydays" were 2003-2005. At least that's what I see. So I doubt JavaFan had tiobe in mind when making his "haydays" wisecrack.
| [reply] [Watch: Dir/Any] |