Closed GoogleCodeExporter closed 8 years ago
Oh, and this patch requires the patch I submitted in bug #14. I found the
problem
explained in bug #14 while working on this feature.
Original comment by schi...@gmail.com
on 12 Jun 2008 at 3:51
isn't validation planned for the CouchDB server itself. Then a second validation
would add confusion I think.
Original comment by gdam...@gmail.com
on 10 Sep 2008 at 11:37
Also, validation isn't really the job of the DB access library. I think it's
more
pertinent if you either a) talk to the guys over at CouchDB about this, or b)
write a
layer on top of CouchDB/Python which affords things like schema and validation
(which
defeat the point of using CouchDB entirely).
Original comment by crack...@gmail.com
on 17 Dec 2008 at 12:03
[deleted comment]
It would be fine if this was done in CouchDB itself, but there was no sign of
that
feature when I submitted this patch.
I strongly disagree that adding validation "defeats the point of CouchDB
entirely".
Sure CouchDB lacks a formal schema and is malleable, but that doesn't mean
users will
never have a plan for what their data looks like at any given moment. The
schema-free
bit just lets people change their mind without major hassles. If having any
form or
schema was bad, then you could argue that this entire access library is
defeating the
purpose of CouchDB since it has things like typed fields.
Original comment by schi...@gmail.com
on 17 Dec 2008 at 12:38
Original comment by djc.ochtman
on 14 Dec 2009 at 10:30
This seems related to something I would like to see in couchdb: namespaces.
Although I still see no reason to include any kind of enforcement or validation
in the
db itself.
See: https://issues.apache.org/jira/browse/COUCHDB-714
Original comment by DeviantR...@gmail.com
on 21 May 2010 at 4:43
A validate_doc_update function in your design do is the way to do this sort of
validation, http://wiki.apache.org/couchdb/Document_Update_Validation. That
will work for non-schema (now called mapping) documents too.
Original comment by matt.goo...@gmail.com
on 8 Jul 2010 at 8:46
Original issue reported on code.google.com by
schi...@gmail.com
on 12 Jun 2008 at 3:43Attachments: