On 21 March 2014 23:36, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Simon Riggs <simon@2ndQuadrant.com> writes:
>> On 21 March 2014 20:58, Noah Misch <noah@leadboat.com> wrote:
>>> It's not the behavior I would choose for a new product, but I can't see
>>> benefits sufficient to overturn previous decisions to keep it.
>
>> Speechless
>
> The key argument for not "fixing" this is that it would break existing
> pg_dump files. That's a pretty hard argument to overcome, unfortunately,
> even if you're willing to blow off the possibility that client
> applications might contain similar shortcuts. We still do our best to
> read dump files from the 7.0 era (see ConvertTriggerToFK() for one example
> of going above and beyond for that); and every so often we do hear of
> people trying to get data out of such ancient servers. So even if you
> went and fixed pg_dump tomorrow, it'd probably be ten or fifteen years
> before people would let you stop reading dumps from existing versions.
Noah had already convinced me, but thank you for explaining the issue
behind that.
-- Simon Riggs http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training & Services