On 2011-11-04 17:36, Anne van Kesteren wrote:
On Fri, 04 Nov 2011 09:18:29 -0700, Julian Reschke
<[email protected]> wrote:
On 2011-11-04 17:10, Arthur Barstow wrote:
The group discussed this on October 31 [1]. The gist of the agreement is
that since the text that is now in the API spec used to be in the
protocol spec, the totality of a review of the two specs is effectively
the same. In this view, the change to the API spec is not substantive.
...
Doesn't compute. The text was *removed* from the protocol spec because
the WG found it to be misleading (suggesting WS URIs are different
from other URIs). It was *not* removed because we thought it belongs
somewhere else. Also, in case that wasn't clear, it was *replaced* by
different text.
Citation needed.
http://lists.w3.org/Archives/Public/public-webapps/2011OctDec/0640.html
I pointed you to the thread on the HyBi mailing list that caused the
change to happen (starting at
<https://www.ietf.org/mail-archive/web/hybi/current/msg07377.html>). The
WG discussed the topic, I was asked to propose text, and the editor
adopted it. As far as I recall, there was no further discussion. (that's
how IETF WGs normally operate)
I think we need to continue to move forward and to acknowledge several
implementations of the API spec have been deployed. As such, I tend to
think we may have already passed the point of diminishing returns
regarding minor tweaks to the spec and if there are bugs, in the spec,
please file bugs and we can address them during CR.
...
Right now, the spec "uses" an algorithm without actually referring to
it. A *minimal* fix is to make that a proper reference.
http://lists.w3.org/Archives/Public/public-webapps/2011OctDec/0661.html
"You open the two dependencies, do a full text search, and you are done."
If you think that's sufficient then please have that text added to the
spec, pointing out it applies to things that might be names of
algorithms and which are marked up in bold (right?).
Personally, I don't think this is acceptable. Editorial glitches like
this happen, and that's ok, but that doesn't mean it's ok to leave them
in a published spec when the problem is discovered beforehand.
Best regards, Julian
