- From: Justin James <j_james@mindspring.com>
- Date: Thu, 18 Sep 2008 16:15:50 -0400
- To: "'Leif Halvard Silli'" <lhs@malform.no>
- Cc: "'Justin Anthony Knapp'" <justinkoavf@gmail.com>, <john@netpurgatory.com>, "'N-at-Work'" <info@n-at-work.net>, <public-html@w3.org>
Leif - My concern here is that I don't think that we should be delving into the realm of defining UA's UI behavior, and the _tab proposal requires exactly that. What I *do* like, however, is the idea that we may need a way to create another instance of the same browsing context, forking the process, as it were. Something like "_fork" or "_continuation" would be sufficient. I also feel that in scenarios where the Web developer/designer requires precise control over how a new window is created, the corresponding DOM functions are best suited for this task. J.Ja > -----Original Message----- > From: Leif Halvard Silli [mailto:lhs@malform.no] > Sent: Thursday, September 18, 2008 3:37 PM > To: Justin James > Cc: 'Justin Anthony Knapp'; john@netpurgatory.com; 'N-at-Work'; public- > html@w3.org > Subject: Re: No tabbed browsing in HTML 5? > > Nathalie, perhaps you could explain how you envisionaged the > effect of _tab? I'll offer my perspective here. > > Justin, Lynx could also have had a form of Tabs support. In an UA > without Tabs, it should probably just open a new Window, though it > possibly could create a some kind of grouping as well (see below). > > With "_blank" one creates a new browsing context, where the new > context has no "link" with the originating context. This, in fact, > seems to me to be exactly what we would expect from a "_tab" also. > Except that for "_tab" one would expect that the link would open > in a Tab and not in a Window. Thus, a tab is forming a context > anyhow, since the Tabs of a Window form a kind of group. > > Even Lynx and other UAs without Tab support in a GUI sense, could > develope a way to group several pages into one "usage context". > And until they support something like that, the "_tab" would have > the same effect as "_blank". > > One possible benefit of defining _tab could be that we could also > define a cleaner _blank, where _blank would *always* open a new > Window even if the browser does have support for Tabs. > > The thing with _blank is that many authors apparently expect that > it opens in a new Window, and rely on that behaviour. By that > expectation, they kind of expect the *opposite* of what the draft > says that _blank does. (Draft says that _blank creates a new > context, thus - that there is no "link" between them. ) > > When a new Window opens above the current Tab or Window, then all > it takes to get back to the originating Tab/Window, is that one > closes the Window. But if the page opened in a new Tab instead of > a in a new Window, the outcome of closing the Tab is uncertain. > Quite likely you land in another Tab than the originating one. > > Example: An Internet Banking service I use has a calendar in the > form of a date selector, which opens in a new window. Upon picking > a date, the window closes and the author then expects that the > user will be back in the Banking page again. But in fact, because > the calendar opened in a new Tab, and because I might have opned a > new parallell Tab side-by-side the Banking page before I chose the > Calendar, I might land in that nearby Tab instead of the Banking > page. Annoying! > > The question still is: When is target:_tab useful? A possible > answer could be: On start pages of many kinds. Any page where the > author expects the users to read the links of the page in > parallell tabs. Also, if you operate a page without a mouse, then > it might be convenient if it is simple to open the links in Tabs. > > Even so, I am still not certain that _tab would be much used > though. But it does seem to me to be logical to have one way to > open a page in a Tab and another way to open a page in a new > Window. The fact that we do not have a way to separate them today > has at least created problems for me in my Internet Banking surfing. > > Justin James 2008-09-18 18.44: > > > The behavior of _blank, _self, etc. is only defined in HTML as > referring to > > browsing contexts. However a particular UA chooses to implement that > is up > > to them. _blank, _self, etc. have meaning, even within a browser like > Lynx. > > _tab does not. It just so happens that the most popular use case for > _blank > > is to open a new tab/window, but that does not mean that this is > within the > > scope of the HTML specification. > > > > J.Ja > > > > From: public-html-request@w3.org [mailto:public-html-request@w3.org] > On > > Behalf Of Justin Anthony Knapp > > Sent: Thursday, September 18, 2008 11:43 AM > > To: john@netpurgatory.com > > Cc: N-at-Work; public-html@w3.org > > Subject: Re: No tabbed browsing in HTML 5? > > > > It seems that most browsers have an option to "open in new window" or > "open > > in new tab" and that is probably best handled as a end-user issue. I > > understand your rationale, though. > > > > And, as pointed out, someone will find a work-around: some script > > bookmarklet or Firefox extension that opens the _tabs in _windows and > > vice-versa. The notion of declaring parts of the browser is a bit > silly to > > me. Why not _new or _newdocument or somesuch to signify that this > link is > > probably not supposed to replace this document's display, but can be > > displayed however the end user and the permissions of his program > allow? > > > > -JAK > > On Thu, Sep 18, 2008 at 11:33, John C. Vernaleo > <john@netpurgatory.com> > > wrote: > > > > On Thu, 18 Sep 2008, N-at-Work wrote: > > So what do you think about ?_tab? as an additional keyword. > > > > http://www.w3.org/html/wg/html5/#target6 > > > > I for one really don't see how that is a legit thing for authors to > decide. > > Seems completely a user agent decision to me. (Not to mention > terminal > > based browsers like lynx and screen readers that would have no > reasonable > > way to support it.) > > > > I can barely remember how I used the web before tabs, but I still > wouldn't > > want to give control of that to the html author (and as an author I > wouldn't > > want that control either). If it were added to the spec, then > browsers > > would have to come up with a way to ignore them which seems like it > means > > extra work has to get done by everyone > > -- > leif halvard silli
Received on Thursday, 18 September 2008 20:17:03 UTC