Feedback on Átila's Vision for D

Dukc ajieskola at
Fri Oct 18 21:37:22 UTC 2019

On Friday, 18 October 2019 at 15:03:18 UTC, Chris wrote:
> On Friday, 18 October 2019 at 14:20:00 UTC, bachmeier wrote:
>> He wants to prevent this community from having a way to 
>> evangelize.
> "evangelize", I'd say that's the problem. It's a religious 
> belief / community, is that what you're saying? Yes, I do have 
> issues with that. A PL should convince based on its quality not 
> on making people "feel good" about it, because reality will 
> come back to bite you. How long has it been since ARM 
> (Android/iOS) has been neglected? I first asked about it in 
> 2015.

Perhaps that quote of yours explains the reason for why your 
criticism tends to be less than popular. If I interpret that 
right, you are saying D should sell itself much lower than it 
does when advertising?

Remember, nothing is so bad that one can say it can't rightfully 
brag about something. Basically any language -even Assembly or 
COBOL- has some upsides. Are we still on the same page? Now, you 
might be saying that when advertising D brags about stuff it does 
not have, and that needs correction for sake of honesty. But I 
think many feel that your reactions overshoot a lot from what 
honesty mandates. When one criticizes, there are basically three 
things to remember:

1: Getting all the criticism on target. I can hardly imagine 
anyone accusing a language of dishonesty if it talks about most 
important missing stuff, but leaves general self-hate out.

2: Being reasonable with criticism. If someone advertises D being 
portable, and somebody else feels that's unjustified because our 
runtime isn't implemented for wasm/js, I think its enough that 
the person kindly clarifies what "portable" means for the 
possible newcomer, if he/she thinks that's necessary. After all, 
does the newcomer miss some important info if we leave the "how 
dare you"-style corrections out?

3: Considering if the criticism helps anyone. Fairly much 
everybody AFAIK agrees that autodecoding would be a good 
riddance. But at the same time, the breakage would be HUGE. If 
someone complains about it, even reasonably per se, it will not 
help anything, as all the people already know about the problem. 
That's why scorning maintainers about autodecoding reminds of 
trolling -there is no other apparent reason to do it. Of course, 
if you can suggest a migration path out of it that nobody has 
thought of, that's an entirely different matter.

More information about the Digitalmars-d mailing list