Bug 99093

Summary: IndexedDB: IDL types defined in spec should be visible to scripts
Product: WebKit Reporter: Joshua Bell <jsbell>
Component: New BugsAssignee: Joshua Bell <jsbell>
Status: RESOLVED FIXED    
Severity: Normal CC: abarth, alecflett, dgrogan, ojan, webkit.review.bot
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
Patch
none
Patch for landing none

Joshua Bell
Reported 2012-10-11 13:18:21 PDT
IndexedDB: IDL types defined in spec should be visible to scripts
Attachments
Patch (6.81 KB, patch)
2012-10-11 13:19 PDT, Joshua Bell 		no flags
DetailsFormatted DiffDiff
Patch for landing (7.73 KB, patch)
2012-10-11 13:40 PDT, Joshua Bell 		no flags
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Joshua Bell
Comment 1 2012-10-11 13:19:59 PDT
Created attachment 168266 [details] Patch
Joshua Bell
Comment 2 2012-10-11 13:21:12 PDT
alecflett@, dgrogan@ - please take a look. With this we pass http://w3c-test.org/webapps/IndexedDB/tests/submissions/Opera/idbtransaction.htm
David Grogan
Comment 3 2012-10-11 13:30:46 PDT
LGTM I don't know why this is necessary, but I believe you that it is. Seems harmless. removeVendorPrefixes() in shared.js has a similar check (shouldBeTrueQuiet...). You might want to update that instead of or in addition to the layout test you have here.
Adam Barth
Comment 4 2012-10-11 13:32:47 PDT
> I don't know why this is necessary, but I believe you that it is. Seems harmless. You don't know why exposing these properties is necessary from an API point of view or why we need to declare them explicitly in the IDL from an implementation point of view?
David Grogan
Comment 5 2012-10-11 13:33:39 PDT
(In reply to comment #4) > > I don't know why this is necessary, but I believe you that it is. Seems harmless. > > You don't know why exposing these properties is necessary from an API point of view or why we need to declare them explicitly in the IDL from an implementation point of view? The first.
Adam Barth
Comment 6 2012-10-11 13:36:54 PDT
Typically its so web pages can manipulate the prototypes of these objects, for example to shim in future versions of the API.
David Grogan
Comment 7 2012-10-11 13:38:52 PDT
(In reply to comment #6) > Typically its so web pages can manipulate the prototypes of these objects, for example to shim in future versions of the API. Got it, that makes sense. Josh also mentioned that this could allow for easier feature detection.
Joshua Bell
Comment 8 2012-10-11 13:40:34 PDT
Created attachment 168268 [details] Patch for landing
WebKit Review Bot
Comment 9 2012-10-11 14:37:11 PDT
Comment on attachment 168268 [details] Patch for landing Clearing flags on attachment: 168268 Committed r131096: <http://trac.webkit.org/changeset/131096>
WebKit Review Bot
Comment 10 2012-10-11 14:37:14 PDT
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.