Three people out of four dislike SDL

[Chris via Digitalmars-d]

On Wednesday, 2 December 2015 at 12:40:59 UTC, Timon Gehr wrote:
>
> It's a question of process. When a poll should return results 
> meaningful for motivating plans of action, it is conducted in 
> roughly the following manner:
>
> 0. It is decided that a poll makes sense. (Ideally by means 
> previously agreed upon.)
> 1. The consequences of the different outcomes are specified.
> 2. Arguments for each option are made available. (Somewhat 
> optional, but there certainly shouldn't be any bias and/or 
> vitriol in the original announcement!)
> 3. Steps are taken to ensure everyone affected gets a 
> reasonable chance to vote. (Ideally, those who are not affected 
> should not be allowed to vote, but that is hard to enforce in 
> practice.)
> 4. The voting takes place, with a deadline defined well in 
> advance.
> 5. The consequences that were initially specified are applied 
> (after the poll has closed).
>
> To support those points, just re-check the results. Since this 
> thread has started (with a suggestion to take actions based on 
> the poll), the answers have basically evened out. It is 
> currently at 48% vs 52%:
>
> http://www.easypolls.net/poll.html?p=565587f4e4b0b3955a59fb67
>
> This effect wasn't as strong when I wrote my previous post, but 
> it does not surprise me at all. I myself have still not taken 
> the poll of the troll, and I am quite confident that there are 
> others with a similar stance.

I, for my part, didn't vote and do not intend to do so. We need a 
pragmatic solution, not a subjective opinion poll. A pragmatic 
solution, imo, would be to revert dub to default to JSON, and 
keep SDL as a second option (which will be convertible to JSON 
with the next version anyway). The less-known format should be 
the second option, not the default.