CVS log for html5/postmsg/Overview.html

Up to [Public] / html5 / postmsg

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

fix several broken links

1. Add initMessageEvent() method to the MessageEvent interface IDL;
2. Some minor fixs to minimize the diffs to the WHATWG version.

Big editorial cleanup. No normative changes. (whatwg r8634)

Instead of MessageEvent.channel, use MessageEvent.target.name. (whatwg r8589)

Print version of the spec (thanks IZh!); protocol-neutral version of the spec so that scripts and styles work on http://www.whatwg.org/ (whatwg r8578)

Fix some missing mentions of 'fillRule' in the canvas domintro (whatwg r8542)

Remove support for receiving an error message when a port's other side is destroyed by the OS, due to lack of implementation interest. (whatwg r8452)

Ok, go back to exposing MessageEvent.source in one form everywhere, relying on WebIDL for the rest of the logic here (whatwg r8444)

Try doing this a different way (MessageEvent.source exposure) (whatwg r8430)

Update the spec to say what's exposed to workers: take 1 (whatwg r8425)

Various editorial tweaks. (whatwg r8405)

Add an event index which acts as the definition of the event names, for editorial purposes (event names are really just strings, so they don't have, or need, a normative definition per se). (whatwg r8385)

Make sure subsequent owners of an ill-fated port's friend can know about that fate (whatwg r8342)

Updates for r8297, fixing <option> to treat 'dirtiness' correctly, and r8336, fixing 'error' events sent to MessagePort objects to not race messages sent from those ports (or, worse, the event that the port is delivered on...). (whatwg r8341)

Add a way to catch the other side of a port having a catastrophic death. (whatwg r8336)

People apparently don't like it when the spec requires things that are impossible to implement, go figure. (In this case, synchronously detecting that one of the MessagePorts being Transferred in the MessagePort message is actually the target of the message. You can't necessarily know this synchronously, since if the port has been shunted around between workers, you might only discover who the final target actually is after the message has itself bounced between threads for a while.) (whatwg r8306)

New BroadcastChannel feature to send messages to other tabs in the same origin, so you don't have to abuse the onstorage event, and don't need to use a heavy shared worker for something otherwise trivial. (whatwg r8274)

Move the spec from a stack of incumbent scripts to a stack of script settings object. This should in theory have no concrete effects (though I may have changed some of the origin used for Web Workers started from document.domain-affected scripts that were called from other scripts with different original origins). (whatwg r8262)

A MessageEvent object constructed using MessageEventInit doesn't get the actual array that was passed, that would be weird. It gets a copy. (whatwg r8260)

The bulk of this is editorial: refactoring how scripts are defined so that all the common stuff is in a shared 'settings object' rather than being duplicated per script. But this also cleans up how postMessage() interacts with the event loop and a few other things I've since forgotten. (whatwg r8247)

The giant clean-up of 2013. Improved internal consistency, adjusted section order to make more sense, started a move away from abusing title='' for cross-references, fixed minor typos, cleaned up some dangling TODOs, very minor typo fixes, dropped some obsolete comments, un-commented-out minor things that had previously been commented out for reasons that no longer apply, etc. No normative changes. (whatwg r8229)

Make a note of the ills of PortCollection (whatwg r8216)

Reorganise the stuff about port message queues to be clearer. (whatwg r8154)

Disabled port message queues shouldn't have their messages dispatched when they're still unshipped, nor should they stall the entire message queueing mechanism. (whatwg r8153)

Make sure you can create a MessageEvent object with a MessagePort as a source. (whatwg r8148)

Cleanup (whatwg r8118)

Try to tighten up text around event handlers (whatwg r8091)

Integrate ArrayBuffer structured cloning (whatwg r7972)

Try to clean up text where I referred to scripts that called a method, etc. Introduces the term 'incumbent script'. (whatwg r7954)

Cleanup (whatwg r7919)

Let's try some more explanatory text (whatwg r7892)

Remove redundant requirements. (whatwg r7852)

WebVTT is now being edited by Silvia Pfeiffer. If you want to follow further development, please subscribe to public-texttracks@w3.org, the WebVTT community group's mailing list. (whatwg r7834)

Fix example. (whatwg r7818)

Use more WebIDL, reduce amount of prose. Also, some typo fixes. (whatwg r7808)

Lots of minor editorial fixes thanks to Michael Dyck. (whatwg r7803)

Cleanup (whatwg r7695)

Cleanup (whatwg r7694)

Merge WorkerUtils into WorkerGlobalScope (whatwg r7675)

Make MessagePort objects synchronised until they are posted through another port. (whatwg r7669)

Embrace the Encodings specification. (whatwg r7647)

Update references to DOM spec so they call it its real name. (Boy am I late to _this_ party.) This update will also make <iframe seamless> do event retargetting once DOM is updated to support that (see bug 18780 - marked this bug 'editorial' since this change doesn't work without the DOM stuff). Also, a lot of clean-up to affected areas. (whatwg r7562)

Cleanup sweep. Should make diffs of upcoming patches easier to read. This edit contains no normative changes as far as I'm aware, and I believe the only textual changes of any kind are in the history section. (whatwg r7530)

Try to be more normatively tight about 'fire' and 'dispatch'. (whatwg r7506)

Fix some typos or copypasta. (whatwg r7503)

Provide a way for authors to broadcast to many ports without memory leaks. (whatwg r7502)

Add sandbox=allow-pointer-lock, and some nearby cleanup. (whatwg r7485)

Fix link.

Typos, thanks to foolip! (whatwg r7465)

Define when events are trusted and when they're not. Hopefully I got all the places that create and dispatch events... (whatwg r7437)

These specs are no longer being maintained at the W3C, so update links to point to where they are now being maintained. (whatwg r7361)

Updates the the vCard section, and related cleanup. (whatwg r7276)

Explain start() with an example. (whatwg r7171)

sync with WebIDL for event handlers (whatwg r7146)

Fix some typos or copypasta. (whatwg r7125)

Update 'connect' events in shared workers to include the port on event.source, as well as event.ports[0], to make the event more intuitive. (whatwg r7114)

Move slightly more towards using the word 'type' rather than 'name' for the concept of an event name (whatwg r7099)

Add new escapes to WebVTT for bidi authors (whatwg r7063)

Only apply styles for [dir] if the attribute value is actually valid (ltr, rtl, auto). (whatwg r7043)

Add ellipse support to canvas. (whatwg r7025)

Add more explanatory text about postMessage()'s first argument. (whatwg r6903)

typo (whatwg r6852)

Prepare a WebVTT standard. Remove the WebVTT text from the HTML standard; refer to the WebVTT spec instead. Drop PeerConnection (moved to W3C WebRTC WG). Clean up some internal annotations (sorry, this affects a number of topics but is purely editorial). Fix some WebIDL issues flagged by recent updates to the WebIDL checker. (whatwg r6837)

New topics: DOM APIs, and Security. (Note that by default, people are _not_ subscribed to new topics.) Also, some tweaks to the previous annotations. (whatwg r6836)

refs (whatwg r6747)

Mention some DOS risks with window.postMessage(). (whatwg r6743)

typo (whatwg r6738)

Correctness (useMap reflects 'usemap' not 'useMap'). (whatwg r6702)

Update postMessage() to do the transfers after the cloning so that if cloning fails, the transferables don't get killed. (whatwg r6697)

Clarify that you can transfer a port more than once by avoiding confusing terminology that suggested otherwise. (see also r6653) (whatwg r6696)

Revert last checkin. My testing was bogus. (whatwg r6691)

Make postMessage() accept the second and third arguments in either order. Compat with WebKit and hopefully solves the difficulty people have about remembering which goes first. (whatwg r6690)

Clarify that you can transfer a port more than once by avoiding confusing terminology that suggested otherwise. (whatwg r6653)

include information about the capabilities model (whatwg r6647)

Shake up the publication pipeline a bit. Merge the main two WHATWG specs. Update styles. (whatwg r6619)

Exceptions DOM Core/WebIDL fix. (whatwg r6602)

Event handler IDL attribute compatibility WebIDL fix. (whatwg r6599)

Update Event interfaces to use constructors instead of init*Event methods; remove some typedefs. (whatwg r6551)

Update WebIDL blocks to use new syntax. (whatwg r6502)

prettyify (whatwg r6461)

style update

Move the Web to a world where things that implement EventTarget just inherit from it. I, for one, now welcome our new EventTarget overlords. (whatwg r6378)

styling

Fix the handling of infinite loops in the various microdata algorithms. (whatwg r6277)

Revamp how audioTracks and videoTracks work on HTMLMediaElement and MediaStream. (whatwg r6275)

add infrastructure to postMessage() to support ArrayBuffer cloning (whatwg r6273)

Update IDL to recent WebIDL changes, thanks to heycam. (whatwg r6220)

Make structured clone handle getters. (whatwg r6185)

Don't overpromise in security sections... (whatwg r6169)

Fix WebIDL-related issues uncovered by the recent updates to the webidl checker (whatwg r6152)

w3c lc preparation

Mention the case of getContext('webgl') returning null due to hardware limitations. (whatwg r6130)

remove the external scripts from the w3c specs

fix pub issues

pipeline update (whatwg r6003)

full regen because the publishing pipeline seems to have gotten out of sync (whatwg r5987)

typo

block more spam

Add a note about cases where scripts won't run. Also, update some references (unrelated to the bug below). (whatwg r5856)

xref (whatwg r5832)

forced refresh

Revamp the way the boilerplate stuff is done so that Tab can help me out doing publications. (whatwg r5719)

remove whatwg.org form dependency

change shortname to webmessaging

Added red block to Implementor warning

updates for pubrules

first version to make FPWD ready

push a manual update

update the form to be less spammed

Add <video audio=muted>. (whatwg r5636)

wd pub preparation (whatwg r5618)

Add the remaining MathML entities -- the ones that expand to two characters. (whatwg r5557)

Reintroduce <ol type> as conforming. (whatwg r5478)

couple of minor errors (whatwg r5415)

Try to clarify an example (whatwg r5278)

drop support for '/' origins in postMessage(), and require that paths be ignored. (whatwg r5277)

try adding registration templates to the spec (whatwg r5273)

new class attribute in header

Add WEBIDL ref to HTML modules

Mention that postMessage() won't work when the target hasn't loaded yet. (whatwg r5217)

Make a WebSRT-only view for Jonas. (whatwg r5192)

update various boilerplate text to be more accurate and more precisely follow pubrules (whatwg r5105)

oops, got the urls for the fonts wrong... (whatwg r5103)

play with @fontface

Remove a paragraph from the W3C version by request of the chairs. Add a list of differences between the versions of the spec to the WHATWG version now that there's more than a few. Add a paragraph to the W3C version mentioning this. Make it easier to keep track of what is in what version of the spec when editing. (whatwg r5101)

change how boilerplate is added and update some boilerplate text with the new mechanism (whatwg r5099)

build system update (whatwg r5098)

Remove the experimental timed tracks stuff from the W3C copy, by request of the chairs.

boilerplate updates based on w3c agreement

id update for request from anne (whatwg r5091)

regen for build scripts (whatwg r5090)

regen (whatwg r5089)

update bugs db links (whatwg r5086)

Update component annotations (whatwg r5085)

Captions - Stage 8: MIME registration for text/websrt and minor edits around that. (whatwg r5078)

make the open bug list more visible (new link from maciej) (whatwg r5064)

regen with updated scripts (no changes expected other than dates)

Make microdata not depend on HTML5 for now to decouple them process-wise. (whatwg r5037)

Merge the two postMessage algorithsm back into one. (whatwg r5002)

fix domintro which had gotten out of sync with reality (whatwg r5001)

updating header per chair request (whatwg r4963)

typo (whatwg r4961)

spec gen updates (whatwg r4958)

Apparently position:relative on table-cell is undefined. Let's rely on the slightly less undefined top:auto;left:auto. (whatwg r4916)

Add the glyphs to the named character references table (whatwg r4915)

stablise ids for security sections (whatwg r4900)

make the domintro warning boxes less like .issue boxes (whatwg r4895)

Move the 'ports' argument to the end of postMessage() instead of in the middle. (whatwg r4870)

Fix meaning of 'live' throughout. Hide a Window object that had leaked out instead of turning into a WindowProxy object. Fix a note that had rotted. (whatwg r4857)

w3c continues to insist on using html4 apparently.

Fix some pubrules problems. (only affects the W3C variants of the specs) (whatwg r4819)

Update all copyrights. (whatwg r4818)

First draft of possible change to the WebSocket handshake. Also, make the protocol and API be consistently called WebSocket. (whatwg r4814)

typo

pubrules updates

minor wording change for status section

prepare to xref Event (whatwg r4799)

Make 'Referer' work correctly for scripts in shared workers. (For some definition of 'correctly' -- it uses the URL of the document that actually created the script. Arguably it should use the URL of the script itself. However, this doesn't change that, it just makes it not leak the URL of documents that that document's browsing context is navigated to.) (whatwg r4789)

Emphasise the non-normative status of the domintro boxes. (whatwg r4787)

Oops, included the extension text in the main html5 spec.

browsing contexts don't have origins... these should have been referring to the active document. (whatwg r4774)

front matter updates for publication this week

Change postMessage()'s targetOrigin argument to not actually resolve but to just treat '/' specially. This will mean that targetOrigin can no longer be in the form '//example.com/' and can no longer be the empty string. (whatwg r4720)

Add a note to the 'fetch' algorithm about redirects. I'm sure it'll need tweaking later, since browsers ignore HTTP here for compatibility with deployed content. Also, tweak the W3C postmsg draft's boilerplate. (whatwg r4692)

Fix the mess that was included in the last checkin. (whatwg r4690)

Editorial changes to 2D and Microdata specs for FPWD. Clean-up of the references sections. Fix some problems caused by dropping language versioning at the WHATWG. (whatwg r4646)

missed the dfn styles

Make the microdata spec stand slightly more alone. (whatwg r4634)

Allow skipping sniffing steps. (whatwg r4556)

Merge core and vocab specs back together. (whatwg r4553)

Put back <details> into the W3C copy of the spec. (whatwg r4552)

Split out the Web Browsers section and related bits. (whatwg r4551)

split out communications section (whatwg r4548)