Re: ALTER TYPE 0: Introduction; test cases

On Sun, Jan 16, 2011 at 12:07 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> On Sat, Jan 15, 2011 at 10:25 AM, Noah Misch <noah@leadboat.com> wrote:
>>> Do you value test coverage so little?
>
>> If you're asking whether I think real-world usability is more
>> important than test coverage, then yes.
>
> Quite honestly, I'd be inclined to rip out most of the DEBUG messages I
> see in that regression test altogether.  They are useless, and so is the
> regression test itself.  An appropriate regression test would involve
> something more like checking that the relfilenode changed and then
> checking that the contained data is still sane.

From my point of view, the value of those messages is that if someone
is altering or clustering a large table, they might like to get a
series of messages: rewriting the table, rebuilding this index,
rebuilding that index, rewriting the toast table index, .... as a
crude sort of progress indication.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company