> On 05 Jul 2015, at 05:09, Adam Kocoloski <[email protected]> wrote:
> 
>> On Jul 2, 2015, at 6:23 AM, Jan Lehnardt <[email protected]> wrote:
>> 
>> 
>>> On 02 Jul 2015, at 12:19, Hermann Norpois <[email protected]> wrote:
>>> 
>>> Hello,
>>> 
>>> how does couchdb parse the json-documents? Which tool is used? Is it a
>>> known library or is it a couchdb built in function?
>> 
>> We use https://github.com/apache/couchdb-jiffy a C-implementation of a JSON
>> parser.
>> 
>> Although there is now a discussion to move back to mochijson2, which is
>> native Erlang: https://issues.apache.org/jira/browse/COUCHDB-2732
>> 
>> Best
>> Jan
>> —
> 
> (moving to dev@)
> 
> Hiya Jan, we could be discussing a jiffy -> mochijson2 switch, but that’s not 
> the topic of COUCHDB-2732. The issue there is the locking required by the NIF 
> we use for JSON _collation_, which is something Filipe wrote for us back in 
> COUCHDB-1186.
> 
> On the topic of JSON collation, the findings by the IBM team are that pure 
> Erlang collation outperforms the existing NIF, but that a NIF taking 
> advantage of thread local storage is better still. Cheers,

I stand corrected, thanks Adam! :)


Jan
--

> 
> Adam

-- 
Professional Support for Apache CouchDB:
http://www.neighbourhood.ie/couchdb-support/

Reply via email to