On 12/23/20 11:19 AM, Axel Wagner wrote: > I have to call it out here though as I see statistic abuse on the news > every > day. Not to mention that asking the question encourages people to think of > something. > > Ignoring that encouragement in the question (and not remembering survey > structure). This would more accurately be described as 80% reported not > needing > any new features and 15% reported needing Generics! > > > That re-framing is itself abuse of statistics. Because (as rightly pointed out > by Ian) > >> Of course this isn't definitive, since there was no clear way for people they > say that do not want generics. > > Unless there was a "I don't need any new features" checkbox, exactly 0% of > respondents reported not needing any new features. > There not being such a checkbox might be considered a methodological flaw. But > luckily, the goal was neither to do science, nor to vote, so it doesn't hugely > matter. > > I think what how Ian phrased it, is unambiguously the most accurate > description: > 25% of the survey takers answered that question and of those, 79% mentioned > generics. Any inaccuracies come from reading more meaning into these numbers - > as you tried when saying "80% reported not needing any new features" or as > would > happen if Go *was* designed based on public polls and these numbers would be > used to say generics where desperately needed. But it isn't and the survey > isn't > the (only) reason to include generics and Ian was quite careful in pointing > out > the methodological flaws in trying to interpret the numbers in any of these > ways, so there is nothing to see here :)
I disagree on multiple counts. Most importantly, I misleadingly state "more accurately" and not accurately described in order to emphasize that the impression given by the statement may be the complete opposite to reality. Primarily. It misled me to the wrong conclusion at first. You can argue, that is my problem, but it isn't. It is framing, though I am sure not intentionally. An accurate statement might be, "it is the most requested new feature, when asked"? Of course that assumes new features are good, which is part of the debate raised in this thread. On a side note, generics may alleviate some confusion around interfaces and maybe type safety (I avoid interfaces in my code) and may prove to be a good or bad thing to happen to Go for me. Unfortunately, I am not sure a survey would help anyway. Personally, I have found the discussions confusing and haven't the time or likely the expertise for any sufficient analysis. Perhaps many have but I doubt that? I have never used Generics and potentially never will. Maybe I will learn to love them. The likelihood is a love/hate relationship AFAICT. Thankfully, Ian is very engaging on the details for those that can and seems to be very knowledgeable and considerate on the topic. I am glad I don't have those pressures, too. -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/a7ced0ab-a7de-c48c-3ce1-9424405cf50c%40gmail.com.
