[racket] SSL and db/postgresql performance

On 05/11/2013 10:08 PM, Neil Van Dyke wrote:
> Ryan Culpepper wrote at 05/11/2013 06:34 PM:
>>
>> Yes. I just pushed a workaround.
>
> Thank you for this, Ryan!  Regarding the "FIXME" comment: would you say
> this fix is ready for production use?

Probably. That function is certainly wrong as a general-purpose port 
constructor, but it should be safe for the way that particular port gets 
used (no concurrent access, not used as a synchronizable event, etc). 
The test suite runs with one failure (related to disconnect after a 
custodian shutdown), which I believe is related to SSL ports and not to 
the buffering wrapper port.

Ryan