| Message ID | 20210825092023.81396-4-thuth@redhat.com |
|---|---|
| State | New |
| Headers | show |
| Series | softmmu/vl: Deprecate old and crufty display ui options | expand |
As an alternative, you may want to turn it into "-display sdl" rather than poke at dpy. This isn't much more code, but it keeps the shortcut isolated within a single "case". This follows a lot of recently cleaned up command line parsing code such as -no-hpet, -enable-kvm, -smp etc. In the end (spoiler alert for my upcoming KVM Forum presentation—slides are already on sched.com :)) what really produces complexity is the lack of isolation/modularity. As long as UI code doesn't care about command line parsing, and command line parsing doesn't care about global variables from all over the place, the cost of shortcuts is so small that it may tilt in favor of keeping them. Paolo Il mer 25 ago 2021, 11:20 Thomas Huth <thuth@redhat.com> ha scritto: > It's not that much complicated to type "-display sdl" or "-display curses", > so we should not clutter our main option name space with such simple > wrapper options and rather present the users with a concise interface > instead. Thus let's deprecate the "-sdl" and "-curses" wrapper options now. > > Acked-by: Peter Krempa <pkrempa@redhat.com> > Signed-off-by: Thomas Huth <thuth@redhat.com> > --- > docs/about/deprecated.rst | 10 ++++++++++ > softmmu/vl.c | 3 +++ > 2 files changed, 13 insertions(+) > > diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst > index 868eca0dd4..d5bec67a78 100644 > --- a/docs/about/deprecated.rst > +++ b/docs/about/deprecated.rst > @@ -148,6 +148,16 @@ Use ``-display sdl,grab-mod=lshift-lctrl-lalt`` > instead. > > Use ``-display sdl,grab-mod=rctrl`` instead. > > +``-sdl`` (since 6.2) > +'''''''''''''''''''' > + > +Use ``-display sdl`` instead. > + > +``-curses`` (since 6.2) > +''''''''''''''''''''''' > + > +Use ``-display curses`` instead. > + > > QEMU Machine Protocol (QMP) commands > ------------------------------------ > diff --git a/softmmu/vl.c b/softmmu/vl.c > index 613948ab46..bb59dbf0de 100644 > --- a/softmmu/vl.c > +++ b/softmmu/vl.c > @@ -2897,6 +2897,8 @@ void qemu_init(int argc, char **argv, char **envp) > dpy.type = DISPLAY_TYPE_NONE; > break; > case QEMU_OPTION_curses: > + warn_report("-curses is deprecated, " > + "use -display curses instead."); > #ifdef CONFIG_CURSES > dpy.type = DISPLAY_TYPE_CURSES; > #else > @@ -3270,6 +3272,7 @@ void qemu_init(int argc, char **argv, char **envp) > "-display ...,window-close=off instead."); > break; > case QEMU_OPTION_sdl: > + warn_report("-sdl is deprecated, use -display sdl > instead."); > #ifdef CONFIG_SDL > dpy.type = DISPLAY_TYPE_SDL; > break; > -- > 2.27.0 > >
On 31/08/2021 15.53, Paolo Bonzini wrote: > As an alternative, you may want to turn it into "-display sdl" rather than > poke at dpy. This isn't much more code, but it keeps the shortcut isolated > within a single "case". This follows a lot of recently cleaned up command > line parsing code such as -no-hpet, -enable-kvm, -smp etc. > > In the end (spoiler alert for my upcoming KVM Forum presentation—slides are > already on sched.com <http://sched.com> :)) what really produces complexity > is the lack of isolation/modularity. As long as UI code doesn't care about > command line parsing, and command line parsing doesn't care about global > variables from all over the place, the cost of shortcuts is so small that it > may tilt in favor of keeping them. Honestly, I'd rather like to see them removed in the end. Our user interface is so terribly inconsistent here that I think that these options are rather confusing for the users than helpful. For example, why do we have -sdl and -curses, but no -gtk ? And as a normal user, I'd also always wonder what's the difference between "-display sdl" and "-sdl", since the difference in the amount of characters that you have to type here is not that much that it justifies the shortcut option. So IMHO let's rather clean this up completely than dragging the shortcut options along forever. Thomas
On Thu, Sep 02, 2021 at 12:51:02PM +0200, Thomas Huth wrote: > On 31/08/2021 15.53, Paolo Bonzini wrote: > > As an alternative, you may want to turn it into "-display sdl" rather > > than poke at dpy. This isn't much more code, but it keeps the shortcut > > isolated within a single "case". This follows a lot of recently cleaned > > up command line parsing code such as -no-hpet, -enable-kvm, -smp etc. > > > > In the end (spoiler alert for my upcoming KVM Forum presentation—slides > > are already on sched.com <http://sched.com> :)) what really produces > > complexity is the lack of isolation/modularity. As long as UI code > > doesn't care about command line parsing, and command line parsing > > doesn't care about global variables from all over the place, the cost of > > shortcuts is so small that it may tilt in favor of keeping them. > > Honestly, I'd rather like to see them removed in the end. Our user interface > is so terribly inconsistent here that I think that these options are rather > confusing for the users than helpful. For example, why do we have -sdl and > -curses, but no -gtk ? And as a normal user, I'd also always wonder what's > the difference between "-display sdl" and "-sdl", since the difference in > the amount of characters that you have to type here is not that much that it > justifies the shortcut option. So IMHO let's rather clean this up completely > than dragging the shortcut options along forever. There's also the elephant in the room "-vnc" which has never been mapped into -display, but which is also one of the most widely used options for display backends :-( Regards, Daniel
On 02/09/2021 12.58, Daniel P. Berrangé wrote: > On Thu, Sep 02, 2021 at 12:51:02PM +0200, Thomas Huth wrote: >> On 31/08/2021 15.53, Paolo Bonzini wrote: >>> As an alternative, you may want to turn it into "-display sdl" rather >>> than poke at dpy. This isn't much more code, but it keeps the shortcut >>> isolated within a single "case". This follows a lot of recently cleaned >>> up command line parsing code such as -no-hpet, -enable-kvm, -smp etc. >>> >>> In the end (spoiler alert for my upcoming KVM Forum presentation—slides >>> are already on sched.com <http://sched.com> :)) what really produces >>> complexity is the lack of isolation/modularity. As long as UI code >>> doesn't care about command line parsing, and command line parsing >>> doesn't care about global variables from all over the place, the cost of >>> shortcuts is so small that it may tilt in favor of keeping them. >> >> Honestly, I'd rather like to see them removed in the end. Our user interface >> is so terribly inconsistent here that I think that these options are rather >> confusing for the users than helpful. For example, why do we have -sdl and >> -curses, but no -gtk ? And as a normal user, I'd also always wonder what's >> the difference between "-display sdl" and "-sdl", since the difference in >> the amount of characters that you have to type here is not that much that it >> justifies the shortcut option. So IMHO let's rather clean this up completely >> than dragging the shortcut options along forever. > > There's also the elephant in the room "-vnc" which has never been mapped > into -display, but which is also one of the most widely used options for > display backends :-( Yeah, for -vnc, it likely makes sense to keep the shortcut since it's in wide use and also takes additional parameters ... but -sdl and -curses? They are IMHO rather niche, and don't take additional parameters, so it should be ok to mark them as deprecated. We can still reconsider in case anybody complains about the deprecation later. Thomas
Daniel P. Berrangé <berrange@redhat.com> writes: > On Thu, Sep 02, 2021 at 12:51:02PM +0200, Thomas Huth wrote: >> On 31/08/2021 15.53, Paolo Bonzini wrote: >> > As an alternative, you may want to turn it into "-display sdl" rather >> > than poke at dpy. This isn't much more code, but it keeps the shortcut >> > isolated within a single "case". This follows a lot of recently cleaned >> > up command line parsing code such as -no-hpet, -enable-kvm, -smp etc. >> > >> > In the end (spoiler alert for my upcoming KVM Forum presentation—slides >> > are already on sched.com <http://sched.com> :)) what really produces >> > complexity is the lack of isolation/modularity. As long as UI code >> > doesn't care about command line parsing, and command line parsing >> > doesn't care about global variables from all over the place, the cost of >> > shortcuts is so small that it may tilt in favor of keeping them. >> >> Honestly, I'd rather like to see them removed in the end. Our user interface >> is so terribly inconsistent here that I think that these options are rather >> confusing for the users than helpful. For example, why do we have -sdl and >> -curses, but no -gtk ? And as a normal user, I'd also always wonder what's >> the difference between "-display sdl" and "-sdl", since the difference in >> the amount of characters that you have to type here is not that much that it >> justifies the shortcut option. So IMHO let's rather clean this up completely >> than dragging the shortcut options along forever. > > There's also the elephant in the room "-vnc" which has never been mapped > into -display, but which is also one of the most widely used options for > display backends :-( There's -display vnc=... Option -help shows it, -display help doesn't, but that's just a bug, I guess. More serious: -display '{"type": "vnc", ...} isn't implemented.
Hi, > There's also the elephant in the room "-vnc" which has never been mapped > into -display, but which is also one of the most widely used options for > display backends :-( Other way around, -display vnc should be dropped. -display is about local displays, and there can be only one instance. -vnc / -spice enable remote access, and this can be done in addition to a local display. not possible: -display gtk + -display sdl possible: -display gtk + -vnc -display gtk + -vnc + -spice -display none + -vnc + -spice take care, Gerd
diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index 868eca0dd4..d5bec67a78 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -148,6 +148,16 @@ Use ``-display sdl,grab-mod=lshift-lctrl-lalt`` instead. Use ``-display sdl,grab-mod=rctrl`` instead. +``-sdl`` (since 6.2) +'''''''''''''''''''' + +Use ``-display sdl`` instead. + +``-curses`` (since 6.2) +''''''''''''''''''''''' + +Use ``-display curses`` instead. + QEMU Machine Protocol (QMP) commands ------------------------------------ diff --git a/softmmu/vl.c b/softmmu/vl.c index 613948ab46..bb59dbf0de 100644 --- a/softmmu/vl.c +++ b/softmmu/vl.c @@ -2897,6 +2897,8 @@ void qemu_init(int argc, char **argv, char **envp) dpy.type = DISPLAY_TYPE_NONE; break; case QEMU_OPTION_curses: + warn_report("-curses is deprecated, " + "use -display curses instead."); #ifdef CONFIG_CURSES dpy.type = DISPLAY_TYPE_CURSES; #else @@ -3270,6 +3272,7 @@ void qemu_init(int argc, char **argv, char **envp) "-display ...,window-close=off instead."); break; case QEMU_OPTION_sdl: + warn_report("-sdl is deprecated, use -display sdl instead."); #ifdef CONFIG_SDL dpy.type = DISPLAY_TYPE_SDL; break;