16312 2012-03-11 05:18:26 +0000 use of undocumented 'switch' convention for rule formulation is ambiguous 2013-11-11 02:01:57 +0000 1 1 1 Unclassified WebAppsWG XHR unspecified All Windows 3.1 RESOLVED FIXED http://dvcs.w3.org/hg/xhr/raw-file/8d4e9ccfdbd4/Overview.html#infrastructure-for-the-send()-method P2 normal --- 1 glenn annevk art.barstow hsteen ian mike public-webapps public-webapps-bugzilla oldest_to_newest 65320 0 glenn 2012-03-11 05:18:26 +0000 some, but not all of the rule formulations in XHR are based on an undocumented 'switch' convention, that, without further explanation, allows ambiguous interpretation: for example, in 4.7.7 appears the following: ↪ Once the whole response entity body has been received ↪ If there is no response entity body and the state is LOADING ↪ If there is no response entity body and the synchronous flag is set Switch to the DONE state. if I rephrase this as: ↪ If A ↪ If B ↪ If C Then X then which of the following is meant by the above convention? #1 If A && B && C then X then break from switch #2 If A && B && C then X then fall through to next rule #3 If A || B || C then X then break from switch #4 If A || B || C then X then fall through to next rule any of these are possible interpretations given the lack of an explanation of this convention [i note that this convention also appears in HTML5 drafts and is also not documented there either] 66073 1 annevk 2012-03-26 17:27:12 +0000 Maybe get it changed first in HTML? 66086 2 glenn 2012-03-26 22:23:13 +0000 (In reply to comment #1) > Maybe get it changed first in HTML? that would work for me (i.e., define in HTML then put a reference in XHR to that definition); btw, which of #1 through #4 is the intended interpretation? i still do now know how it is intended to be interpreted... 66105 3 annevk 2012-03-27 07:37:02 +0000 It's #3, which is consistent with how <dl> is defined, but that is probably not quite enough. 66106 4 annevk 2012-03-27 07:37:26 +0000 Ian, any ideas? 95951 5 hsteen 2013-11-07 14:31:44 +0000 I think this is now a pretty common convention and sufficiently clear that it needs no further explanation. If it needs explanation, that explanation doesn't belong in XHR as it's a convention shared by lots of specs. 95956 6 annevk 2013-11-07 14:39:18 +0000 So I don't like you closing these bugs. Should we get a separate WHATWG XMLHttpRequest component? There's an outstanding question in comment 4 and there's indeed room for confusion, even though we use the pattern a lot. 95973 7 hsteen 2013-11-07 20:51:51 +0000 No problem - I won't resolve more bugs (unless by prior agreement with you). I think the basic point still stands - why should this convention be defined in the XHR spec when used by lots of others too? - so perhaps we should move this bug to another component or something. (Do we have a "meta-spec" about spec markup?) 95981 8 glenn 2013-11-07 21:41:50 +0000 (In reply to Hallvord R. M. Steen from comment #7) > No problem - I won't resolve more bugs (unless by prior agreement with you). > I think the basic point still stands - why should this convention be defined > in the XHR spec when used by lots of others too? - so perhaps we should move > this bug to another component or something. (Do we have a "meta-spec" about > spec markup?) I agree that it would be better to document this convention in the HTML spec and refer to that definition from XHR. 96029 9 hsteen 2013-11-08 10:16:40 +0000 Isn't the last part of http://www.whatwg.org/specs/web-apps/current-work/multipage/introduction.html#typographic-conventions sufficiently clear? Shall we add a new paragraph to section 3 Terminology saying "This specification uses the same typographic conventions as [HTML5]."? That should be sufficient to close this bug - if Anne agrees this time ;) 96073 10 annevk 2013-11-09 01:51:00 +0000 Glenn, is that clear enough? 96084 11 glenn 2013-11-09 14:47:24 +0000 (In reply to Anne from comment #10) > Glenn, is that clear enough? Yes, Thanks. 96087 12 art.barstow 2013-11-10 04:01:02 +0000 So, there's agreement to s/WebAppsWG/HTMLWG/ and s/XHR/HTML5/? 96088 13 glenn 2013-11-10 06:44:05 +0000 (In reply to Art Barstow from comment #12) > So, there's agreement to s/WebAppsWG/HTMLWG/ and s/XHR/HTML5/? I will open a new bug against HTML5, rather than relabeling this bug. The action on this bug is to reference HTML5 conventions. 96096 14 annevk 2013-11-11 02:01:57 +0000 https://github.com/whatwg/xhr/commit/740a409e37321b411bee782fae28224559863e64