Re: Work in Progress: draft 1.9a
Sparr via wm-spec-list <[email protected]> Thu, 24 Jan 2019 14:42:47 -0800
| Newsgroups | gmane.comp.gnome.wm-spec |
|---|---|
| Message-ID | <CANp3UNArs3HT+jWFEWMyYpyFNBi=ddFSixXBOdMRKTRRxzQ0yQ@mail.gmail.com> |
--===============1364018586441107814== Content-Type: multipart/alternative; boundary="0000000000009d1ebc05803beee7" --0000000000009d1ebc05803beee7 Content-Type: text/plain; charset="UTF-8" > On Sun, 04 Jul 1999, Marko Macek wrote: > > Matthias Ettrich wrote: > > > > > > > > It looks like we need at least three lists then : > > > > > > > > > > _NET_CLIENT_LIST creation/mapping order > > > > > _NET_CLIENT_LIST_FOCUS focus order, for external alt-tab implementation > > > > > _NET_CLIENT_LIST_STACKING stacking order (for tools like a pager). > > > > > > > > > > While STACKING looks important to me, FOCUS isn't. An external alt-tab > > > > > implementation can monitor both _NET_CLIENT_LIST and _NET_ACTIVE_WINDOW to keep > > > > > track of the focus chain. > > > > > > > > To implement Alt+Tab you need Z-ORDER. If we are going to do this we > > > > > > For CDE-style Alt+Tab you need the stacking order, yes. For the IMO more useful > > > KDE(and MS-Windows)-like Alt+Tab you need the focus chain. > > > > Since when does Windows use focus chain? AFAIK it has used Zorder since > > forever. > > (I admit you would need focus chain if activating the window doesn't > > raise it). > > On Windows, z-order and focus chain is the same. KDE does focus chain. > > Sorry for the confusion. > > Matthias First, an apology for resurrecting a 20 year old thread. I don't actually expect to make a difference here, but I am hopeful to collect any historic insights tempered by later information and experience. It seems to me that when _NET_CLIENT_LIST_STACKING was implemented and _NET_CLIENT_LIST_FOCUS was omitted from the standard, there was at least some expectation that the two lists would be the same in some, perhaps most, use cases. Today this expectation is broken by tiling window managers that allow some floating windows, or floating window managers that allow windows to be pinned to the background. Have there been any proposals considered since then that might be useful in this scenario? Google suggests no one ever mentioned _NET_CLIENT_LIST_FOCUS again after this thread. Clarence "Sparr" Risher --0000000000009d1ebc05803beee7 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr">> On Sun, 04 Jul 1999, Marko Macek wrote:<br>> > = Matthias Ettrich wrote:<br>> > > <br>> > > > > It l= ooks like we need at least three lists then :<br>> > > > > <= br>> > > > > =C2=A0 =C2=A0 _NET_CLIENT_LIST creation/mapping= order<br>> > > > > =C2=A0 =C2=A0 _NET_CLIENT_LIST_FOCUS foc= us order, for external alt-tab implementation<br>> > > > > = =C2=A0 =C2=A0 _NET_CLIENT_LIST_STACKING stacking order (for tools like a pa= ger).<br>> > > > > <br>> > > > > While STACKI= NG looks important to me, FOCUS isn't. An external alt-tab<br>> >= > > > implementation can monitor both _NET_CLIENT_LIST and _NET_A= CTIVE_WINDOW to keep<br>> > > > > track of the focus chain.<= br>> > > ><br>> > > > To implement Alt+Tab you need= Z-ORDER. If we are going to do this we<br>> > > <br>> > >= ; For CDE-style Alt+Tab you need the stacking order, yes. For the IMO more = useful<br>> > > KDE(and MS-Windows)-like Alt+Tab you need the focu= s chain.<br>> > <br>> > Since when does Windows use focus chain= ? AFAIK it has used Zorder since<br>> > forever.<br>> > (I admi= t you would need focus chain if activating the window doesn't<br>> &= gt; raise it).<br>> <br>> On Windows, z-order and focus chain is the = same. KDE does focus chain.<br>> <br>> Sorry for the confusion.<br>&g= t; <br>> Matthias<br><br>First, an apology for resurrecting a 20 year ol= d thread. I don't actually expect to make a difference here, but I am h= opeful to collect any historic insights tempered by later information and e= xperience. It seems to me that when _NET_CLIENT_LIST_STACKING was implement= ed and _NET_CLIENT_LIST_FOCUS was omitted from the standard, there was at l= east some expectation that the two lists would be the same in some, perhaps= most, use cases. Today this expectation is broken by tiling window manager= s that allow some floating windows, or floating window managers that allow = windows to be pinned to the background. Have there been any proposals consi= dered since then that might be useful in this scenario? Google suggests no = one ever mentioned _NET_CLIENT_LIST_FOCUS again after this thread.<br><br>C= larence "Sparr" Risher</div> --0000000000009d1ebc05803beee7-- --===============1364018586441107814== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ wm-spec-list mailing list [email protected] https://mail.gnome.org/mailman/listinfo/wm-spec-list --===============1364018586441107814==--