On 01/04/2016 12:34 PM, Daniel Colascione wrote:
> On 01/04/2016 12:32 PM, John Wiegley wrote:
>> OK, I've heard the concerns and input from all parties, and my decision is
>> that we will include the stack overflow recovery logic, as it is now
>> implemented, while keeping our ears open for any problems this causes to
>> users.
>>
>> Daniel, Eli, Paul, thank you for spending the time to be vocal and deliberate
>> in enumerating your concerns. To Daniel specifically: I appreciate your
>> experience in this area, and that you are not speaking from your imagination,
>> but I'd like to give the current recovery approach a try before calling it a
>> non-starter, or branching out into more complex solutions.
>>
>> We *will* have the freedom to reverse this decision in a future release if
>> things become worse rather than better. But there is no more benefit in
>> debating this future.
> 
> Thanks for considering the issue. I still think this decision does not
> bode well at all for Emacs robustness, and I'm terribly disappointed.

I take it you're not interested in other improvements to crash recovery?
That's a shame. I was going to do some work there. As it stands, I'm
forced to conclude that we don't give a damn about making Emacs robust.