>>Simon Riggs wrote:
>>On 9 May 2012 19:17, Peter Eisentraut wrote:
>>
>>> I have been reviewing how our new libpq URL syntax compares
>>> against existing implementations of URL syntaxes in other drivers
>>> or higher-level access libraries. In the case of SQLAlchemy,
>>> there is an incompatibility regarding how Unix-domain sockets are
>>> specified.
>>
>> Is there an open standard that already defines this?
There are many. The most recent, as far as I know is RFC 3986, which
updates one previous RFC and obsoletes three others.
http://tools.ietf.org/html/rfc3986
We should also take the JDBC URL requirements into consideration.
One unpleasant aspect of this is that what JDBC calls a "URL" is
"jdbc:" followed by what could be a valid URI; but I don't see how
the *whole thing* (including the leading "jdbc:" qualifies as a URI
or URL). Unless someone has a better idea, I suggest that we make
what follows the "jdbc:" portion of the JDBC "URL" match what we use
for a URI for everything else.
>> If there is an existing standard we should follow it
+1
I don't know whether recent work on this has respected the standards.
I hope so.
-Kevin
