couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chris Anderson <>
Subject Re: Using HTTP headers
Date Thu, 19 Feb 2009 03:49:27 GMT
On Wed, Feb 18, 2009 at 7:41 PM, Paul Davis <> wrote:
> My third thought is that we can be purists or pragmatists, but not
> both. We are already not being pure in a number of respects so I find
> it a bit disingenuous to claim purism on a given topic while blithely
> ignoring concerns on other topics.

I tend to think query-strings are more pragmatic than headers, even if
they take a little more work to code for.

I think it takes a more advanced developer to code against
headers-based APIs, partly just because they aren't exposed in the
browser. Query strings are in the url, so they can even be copied and
pasted around. The same holds for Accept headers vs .html or .json
file extensions in the URL.

However, I think there's an argument that the commit semantics
shouldn't be part of defining the resource in this case. Maybe
allowing full_commit=true only on the bulk docs query string, since
those are post requests anyway.

Just thinking out loud.


Chris Anderson

View raw message