W3C home > Mailing lists > Public > public-web-bluetooth-log@w3.org > April 2015

Re: [web-bluetooth] Switch to ReSpec's contiguous IDL mode.

From: Jeffrey Yasskin via GitHub <sysbot+gh@w3.org>
Date: Wed, 15 Apr 2015 06:14:51 +0000
To: public-web-bluetooth-log@w3.org
Message-ID: <issue_comment.created-93214907-1429078490-sysbot+gh@w3.org>







<sup>**[index.html, line 323 
\[r1\]](https://reviewable.io:443/reviews/webbluetoothcg/web-bluetooth/102#-JmuAI9RYJZizOk6E8fe)**
 ([raw 
file](https://github.com/webbluetoothcg/web-bluetooth/blob/56d3238dea4fcfe4475af1e45c7ad7b1ba2a097b/index.html#L323)):</sup>
And you were just complaining that the patch is too big. ;) Done with 
the `<a>`, although not uploaded yet. I wonder if these should be 
Notes for the same reason I'm using the notes below.

---

<sup>**[index.html, line 327 
\[r1\]](https://reviewable.io:443/reviews/webbluetoothcg/web-bluetooth/102#-JmuGB0Fw8f05-mcq2dB)**
 ([raw 
file](https://github.com/webbluetoothcg/web-bluetooth/blob/56d3238dea4fcfe4475af1e45c7ad7b1ba2a097b/index.html#L327)):</sup>
I can do sections with headers in this version too. I actually have 
more control over the titles of the sections, too.

However, the paragraphs that just name fields is more consistent with 
other specs: https://fetch.spec.whatwg.org/#dom-request-headers

Actually, the Fetch spec doesn't even describe the fields of its 
dictionaries, which lends more credence to the idea that they should 
be notes at most.

---

<sup>**[index.html, line 907 
\[r1\]](https://reviewable.io:443/reviews/webbluetoothcg/web-bluetooth/102#-JmuOJXiqd8U1eDK6O1r)**
 ([raw 
file](https://github.com/webbluetoothcg/web-bluetooth/blob/56d3238dea4fcfe4475af1e45c7ad7b1ba2a097b/index.html#L907)):</sup>
They're not normative, in that they don't say "name MUST return Xyz". 
Instead, attributes MUST return the last value they're set to, and the
 rest of the spec talks about setting them where necessary. I haven't 
yet added the "MUST return the last value they're set to" text for 
most of them, and instanceID in particular is an outlier in that it 
has way too much MUST that's already covered by how it's set.

This is somewhat inconsistent with other specs, but I found one that 
does it this way ... that I'm trying to re-find so I can imitate it in
 the "MUST return the last value they're set to" text. I think it 
would be difficult to invert this and give user-friendly descriptions 
of the attributes that are also normative.

---

<sup>**[index.html, line 981 
\[r1\]](https://reviewable.io:443/reviews/webbluetoothcg/web-bluetooth/102#-JmuQ1zXyPDpiq96G0Hx)**
 ([raw 
file](https://github.com/webbluetoothcg/web-bluetooth/blob/56d3238dea4fcfe4475af1e45c7ad7b1ba2a097b/index.html#L981)):</sup>
https://fetch.spec.whatwg.org/#dom-response-redirect uses the awkward 
construction "The static redirect(url, status) method, when invoked, 
must run these steps:". Everything in that spec starts with "The", and
 talks about the method doing things instead of the UA. 
http://www.w3.org/TR/service-workers/#navigator-service-worker-getRegistration
 uses the same construct, but skips the "The".

@marcoscaceres, any preferences?

---


---

Comments from the [review on 
Reviewable.io](https://reviewable.io:443/reviews/webbluetoothcg/web-bluetooth/102)
<!-- Sent from Reviewable.io -->


-- 
GitHub Notif of comment by jyasskin
See 
https://github.com/WebBluetoothCG/web-bluetooth/pull/102#issuecomment-93214907
Received on Wednesday, 15 April 2015 06:15:05 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 15 April 2015 06:15:06 UTC