On Wednesday, 6 September 2017 at 10:45:48 UTC, Basile B. wrote:
> And now ?
I'll check later today.

On Thursday, 31 August 2017 at 16:37:35 UTC, SrMordred wrote:
> On Thursday, 31 August 2017 at 14:57:28 UTC, bitwise wrote:
>> https://www.tiobe.com/tiobe-index/d/
>
> What happened in 2009?

My guess is constant random methodology changes. I was tracking TIOBE index each month from 2011 till 2016. I remember they announced changes in methodology in title page approximately once per 3-4 months. For example, changing the base from 100% to sum of percentages of all languages(<100%) increased reported % of each language. Taking this into account means that changes in particular month tells nothing. The trend is, however, positive: in 2014-2017 years D stands higher than in 2011-2014 (if you have faith in TIOBE averages). I see sometimes positive discussions about D at completely unexpected local tech sites.

Although D's position becomes higher in TIOBE, I don't see progress in other statistics, for example in github.

On Wednesday, 6 September 2017 at 10:45:48 UTC, Basile B. wrote:
> And now ?
Just tried. Last version (3 update 4) works well.

On Wednesday, 6 September 2017 at 11:14:00 UTC, Maksim Fomin wrote:
> On Thursday, 31 August 2017 at 16:37:35 UTC, SrMordred wrote:
>> On Thursday, 31 August 2017 at 14:57:28 UTC, bitwise wrote:
>>> https://www.tiobe.com/tiobe-index/d/
>>
>> What happened in 2009?
>
> My guess is constant random methodology changes. I was tracking TIOBE index each month from 2011 till 2016. I remember they announced changes in methodology in title page approximately once per 3-4 months. For example, changing the base from 100% to sum of percentages of all languages(<100%) increased reported % of each language. Taking this into account means that changes in particular month tells nothing. The trend is, however, positive: in 2014-2017 years D stands higher than in 2011-2014 (if you have faith in TIOBE averages). I see sometimes positive discussions about D at completely unexpected local tech sites.
>

That was my thinking too. A real dip as large as what's shown on the graph for 2009 would probably take 3-4 large companies that use D randomly shutting down in perfect unison - highly unlikely. I think that if the glitch were removed, what would remain would be a nice steady upward slope.