On 3/25/2011 2:09 AM, Don wrote:
Walter makes all the language decisions. The rest of us have just been able to convince him on multiple occasions (but I think that even Andrei has not achieved 50% convince rate). Hint #1: if you want to convince Walter, produce some real world use cases. Hint #2: you've got more chance if you make a patch, but ONLY if you've satisfied hint 1.
Nobody would like it if I was easy to convince. There are multiple language change suggestions here *every day*. Day after day, that adds up to a frightening number.
To make a language change stick, it has to have a large benefit divided by its cost. Costs for even small changes can be huge - implementation, testing, documentation, tutorials, books, existing code base, distraction from other issues, instability, regressions, etc.
Priorities. If you spend any time on 'body', that's time taken away from fixing important bugs. The potential benefit is *tiny*. There are 1000 bugs that are more important.
Right. I'm currently working on the temp destruction problem. While not sexy or even particularly visible, it not working right has serious deleterious consequences.
