[webkit-dev] About WebIDL supplemental interfaces
Hello Webkitters! The latest WD for WebIDL defines a supplemental interface of an interface A as any element of the acyclic graph of interfaces that is connected to A through implements statements or through the inheritance chain of another supplemental interface: http://www.w3.org/TR/WebIDL/#dfn-supplemental-interface However, WebKitIDL defines Supplemental more like a partial interface: http://trac.webkit.org/wiki/WebKitIDL#Supplemental I'm asking this because CSS Regions spec defines the Region interface as an WebIDL supplemental interface that needs to be implemented by any object that can be a region (Element, Pseudo elements etc.). Is there a way in WebKit to write an interface that implements another as specified by WebIDL? Thanks, Andrei. ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev
Re: [webkit-dev] About WebIDL supplemental interfaces
Hi Andrei I'm asking this because CSS Regions spec defines the Region interface as an WebIDL supplemental interface that needs to be implemented by any object that can be a region (Element, Pseudo elements etc.). Is there a way in WebKit to write an interface that implements another as specified by WebIDL? Currently, no. (Correct me if I'm wrong.) However, WebKitIDL defines Supplemental more like a partial interface: http://trac.webkit.org/wiki/WebKitIDL#Supplemental The current WebKitIDL implementation is on the way. Eventually we want to remove the [Supplemental] IDL attribute and support the A implements B syntax as defined in the WebIDL. Do you have any alternative solution to make CSS Regions work without the A implements B syntax? If no, maybe it's time to consider supporting the A implements B syntax (, which would require a substantial amount of work though:-). On Thu, Jul 19, 2012 at 5:08 PM, Andrei Bucur abu...@adobe.com wrote: Hello Webkitters! The latest WD for WebIDL defines a supplemental interface of an interface A as any element of the acyclic graph of interfaces that is connected to A through implements statements or through the inheritance chain of another supplemental interface: http://www.w3.org/TR/WebIDL/#dfn-supplemental-interface However, WebKitIDL defines Supplemental more like a partial interface: http://trac.webkit.org/wiki/WebKitIDL#Supplemental I'm asking this because CSS Regions spec defines the Region interface as an WebIDL supplemental interface that needs to be implemented by any object that can be a region (Element, Pseudo elements etc.). Is there a way in WebKit to write an interface that implements another as specified by WebIDL? Thanks, Andrei. ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev -- Kentaro Hara, Tokyo, Japan (http://haraken.info) ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev
Re: [webkit-dev] About WebIDL supplemental interfaces
Hey Kentaro, On 7/19/12 6:20 PM, Kentaro Hara hara...@chromium.org wrote: Hi Andrei I'm asking this because CSS Regions spec defines the Region interface as an WebIDL supplemental interface that needs to be implemented by any object that can be a region (Element, Pseudo elements etc.). Is there a way in WebKit to write an interface that implements another as specified by WebIDL? Currently, no. (Correct me if I'm wrong.) However, WebKitIDL defines Supplemental more like a partial interface: http://trac.webkit.org/wiki/WebKitIDL#Supplemental The current WebKitIDL implementation is on the way. Eventually we want to remove the [Supplemental] IDL attribute and support the A implements B syntax as defined in the WebIDL. Do you have any alternative solution to make CSS Regions work without the A implements B syntax? If no, maybe it's time to consider supporting the A implements B syntax (, which would require a substantial amount of work though:-). The only way I can see this implemented without the new syntax is by writing some crazy hacks, probably very similar with how EventTarget works. And even then I'm not sure everything will work out well. Right now, hacking things in does not seem to worth the effort, especially if it's going to be removed once the implements syntax gets in. I'm open to suggestions though :). However, it would be nice to have the new syntax on board sooner than later. Is there a roadmap for adding this in WebKit? I'm sure you are very busy so having a timeline would allow other hackers help move the implementation forward :). Thanks, Andrei. ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev
Re: [webkit-dev] About WebIDL supplemental interfaces
On Thu, Jul 19, 2012 at 9:29 AM, Andrei Bucur abu...@adobe.com wrote: Hey Kentaro, On 7/19/12 6:20 PM, Kentaro Hara hara...@chromium.org wrote: Hi Andrei I'm asking this because CSS Regions spec defines the Region interface as an WebIDL supplemental interface that needs to be implemented by any object that can be a region (Element, Pseudo elements etc.). Is there a way in WebKit to write an interface that implements another as specified by WebIDL? Currently, no. (Correct me if I'm wrong.) However, WebKitIDL defines Supplemental more like a partial interface: http://trac.webkit.org/wiki/WebKitIDL#Supplemental The current WebKitIDL implementation is on the way. Eventually we want to remove the [Supplemental] IDL attribute and support the A implements B syntax as defined in the WebIDL. Do you have any alternative solution to make CSS Regions work without the A implements B syntax? If no, maybe it's time to consider supporting the A implements B syntax (, which would require a substantial amount of work though:-). The only way I can see this implemented without the new syntax is by writing some crazy hacks, probably very similar with how EventTarget works. And even then I'm not sure everything will work out well. Right now, hacking things in does not seem to worth the effort, especially if it's going to be removed once the implements syntax gets in. I'm open to suggestions though :). Can you say what specifically in the CSS regions spec is giving you trouble? Adam However, it would be nice to have the new syntax on board sooner than later. Is there a roadmap for adding this in WebKit? I'm sure you are very busy so having a timeline would allow other hackers help move the implementation forward :). Thanks, Andrei. ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev
Re: [webkit-dev] About WebIDL supplemental interfaces
Hello Adam, Sure: http://www.w3.org/TR/2012/WD-css3-regions-20120503/#the-region-interface The spec does not explicitly states that the Region should be supplemental, however after raising some issues on www-style ( http://lists.w3.org/Archives/Public/www-style/2012Jul/0251.html ) and talking them through (offline) it seemed the best way to solve them was making Region supplemental for Element and any other object that could become region. Right now, I'm feeling out the situation to see how feasible this is to be implemented in WebKit. Andrei. From: Adam Barth aba...@webkit.orgmailto:aba...@webkit.org Date: Thursday, July 19, 2012 7:31 PM To: Andrei Bucur abu...@adobe.commailto:abu...@adobe.com Cc: Kentaro Hara hara...@chromium.orgmailto:hara...@chromium.org, webkit-dev@lists.webkit.orgmailto:webkit-dev@lists.webkit.org webkit-dev@lists.webkit.orgmailto:webkit-dev@lists.webkit.org Subject: Re: [webkit-dev] About WebIDL supplemental interfaces Can you say what specifically in the CSS regions spec is giving you trouble? Adam ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev
Re: [webkit-dev] About WebIDL supplemental interfaces
What else can become a region besides an element? If there aren't too many interfaces, we can just copy/paste this stanza into each IDL, like we do for EventTarget. If there are a lot of them, then I agree that you'll probably want a fancier Supplemental feature. One way to approach that is to make the [Supplemental] attribute take a list of interfaces. That should be straightforward to implement given our current implementation of Supplemental. Adam On Thu, Jul 19, 2012 at 9:58 AM, Andrei Bucur abu...@adobe.com wrote: Hello Adam, Sure: http://www.w3.org/TR/2012/WD-css3-regions-20120503/#the-region-interface The spec does not explicitly states that the Region should be supplemental, however after raising some issues on www-style ( http://lists.w3.org/Archives/Public/www-style/2012Jul/0251.html ) and talking them through (offline) it seemed the best way to solve them was making Region supplemental for Element and any other object that could become region. Right now, I'm feeling out the situation to see how feasible this is to be implemented in WebKit. Andrei. From: Adam Barth aba...@webkit.org Date: Thursday, July 19, 2012 7:31 PM To: Andrei Bucur abu...@adobe.com Cc: Kentaro Hara hara...@chromium.org, webkit-dev@lists.webkit.org webkit-dev@lists.webkit.org Subject: Re: [webkit-dev] About WebIDL supplemental interfaces Can you say what specifically in the CSS regions spec is giving you trouble? Adam ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev
Re: [webkit-dev] About WebIDL supplemental interfaces
Adam, Currently, an Element is the only thing represented in the object model that can become a CSS Region. Pseudo-elements (::before and ::after) can become CSS Regions, but AFAIK there isn't yet a representation of those in the OM. I'm hoping this changes in the future (I'm working on a spec that addresses this), so that could be a second copy. I can construct a region chain now that includes pseudo-element Regions, and the NamedFlow interface is supposed to return a sequence of Regions from its getRegions() method. So ideally we'd have a way of returning a Region interface for those pseudo-elements that have been added to the region chain. Alan On 7/19/12 10:50 AM, Adam Barth aba...@webkit.org wrote: What else can become a region besides an element? If there aren't too many interfaces, we can just copy/paste this stanza into each IDL, like we do for EventTarget. If there are a lot of them, then I agree that you'll probably want a fancier Supplemental feature. One way to approach that is to make the [Supplemental] attribute take a list of interfaces. That should be straightforward to implement given our current implementation of Supplemental. Adam On Thu, Jul 19, 2012 at 9:58 AM, Andrei Bucur abu...@adobe.com wrote: Hello Adam, Sure: http://www.w3.org/TR/2012/WD-css3-regions-20120503/#the-region-interface The spec does not explicitly states that the Region should be supplemental, however after raising some issues on www-style ( http://lists.w3.org/Archives/Public/www-style/2012Jul/0251.html ) and talking them through (offline) it seemed the best way to solve them was making Region supplemental for Element and any other object that could become region. Right now, I'm feeling out the situation to see how feasible this is to be implemented in WebKit. Andrei. From: Adam Barth aba...@webkit.org Date: Thursday, July 19, 2012 7:31 PM To: Andrei Bucur abu...@adobe.com Cc: Kentaro Hara hara...@chromium.org, webkit-dev@lists.webkit.org webkit-dev@lists.webkit.org Subject: Re: [webkit-dev] About WebIDL supplemental interfaces Can you say what specifically in the CSS regions spec is giving you trouble? Adam ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev
Re: [webkit-dev] About WebIDL supplemental interfaces
It sounds like you don't have a problem today, but you might have a problem in the future. I'd recommend going with the simple solution today and worrying about future problems in the future. Adam On Thu, Jul 19, 2012 at 11:04 AM, Alan Stearns stea...@adobe.com wrote: Adam, Currently, an Element is the only thing represented in the object model that can become a CSS Region. Pseudo-elements (::before and ::after) can become CSS Regions, but AFAIK there isn't yet a representation of those in the OM. I'm hoping this changes in the future (I'm working on a spec that addresses this), so that could be a second copy. I can construct a region chain now that includes pseudo-element Regions, and the NamedFlow interface is supposed to return a sequence of Regions from its getRegions() method. So ideally we'd have a way of returning a Region interface for those pseudo-elements that have been added to the region chain. Alan On 7/19/12 10:50 AM, Adam Barth aba...@webkit.org wrote: What else can become a region besides an element? If there aren't too many interfaces, we can just copy/paste this stanza into each IDL, like we do for EventTarget. If there are a lot of them, then I agree that you'll probably want a fancier Supplemental feature. One way to approach that is to make the [Supplemental] attribute take a list of interfaces. That should be straightforward to implement given our current implementation of Supplemental. Adam On Thu, Jul 19, 2012 at 9:58 AM, Andrei Bucur abu...@adobe.com wrote: Hello Adam, Sure: http://www.w3.org/TR/2012/WD-css3-regions-20120503/#the-region-interface The spec does not explicitly states that the Region should be supplemental, however after raising some issues on www-style ( http://lists.w3.org/Archives/Public/www-style/2012Jul/0251.html ) and talking them through (offline) it seemed the best way to solve them was making Region supplemental for Element and any other object that could become region. Right now, I'm feeling out the situation to see how feasible this is to be implemented in WebKit. Andrei. From: Adam Barth aba...@webkit.org Date: Thursday, July 19, 2012 7:31 PM To: Andrei Bucur abu...@adobe.com Cc: Kentaro Hara hara...@chromium.org, webkit-dev@lists.webkit.org webkit-dev@lists.webkit.org Subject: Re: [webkit-dev] About WebIDL supplemental interfaces Can you say what specifically in the CSS regions spec is giving you trouble? Adam ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev ___ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo/webkit-dev