Hi Naoto, I think I understand what you say, though only one class extending X11InputMethod currently exists, which is sun/awt/X11/XInputMethod.
Probably the code could be refactored such that X11InputMethodBase gets renamed back to X11InputMethod and the coding that went to the platform specific implementations of X11InputMethod with our change should move further down to platform specific implementations of sun/awt/X11/XInputMethod. Is that what you mean? Best regards Christoph > -----Original Message----- > From: naoto.s...@oracle.com [mailto:naoto.s...@oracle.com] > Sent: Freitag, 1. Juni 2018 18:12 > To: Phil Race <philip.r...@oracle.com>; Langer, Christoph > <christoph.lan...@sap.com>; i18n-dev@openjdk.java.net > Cc: Ichiroh Takiguchi <taki...@linux.vnet.ibm.com> > Subject: Re: <i18n dev> JDK-8201429 - Support AIX Input Method Editor > (IME) for AWT Input Method Framework (IMF) > > The issue in the current naming scheme (sharing the same class name for > impls) is that it assumes one X11 im impl per *platform*. Same > IMs/WindowManagers can be shared across platforms, or one platform can > have multiple IM/WM (e.g., the example I wrote below MAWT/XAWT > co-exist), and the current design cannot deal with these situations. > > Naoto > > On 5/30/18 2:33 PM, Phil Race wrote: > > It is in a platform dependent directory .. and I gather this allows the > > build to pick up the right one > > without any work here. > > > > -phil > > > > On 5/30/2018 2:08 PM, Naoto Sato wrote: > >> Hi, > >> > >> I see that the change introduced the base abstraction class, > >> X11InputMethodBase. Back in the days I worked on it, X11InputMethod is > >> the base abstract class, and platform impls. have names representing > >> them, i.e, MInputMethod for Motif, and XInputMethod for XAWT. Now > >> X11InputMethod is a platform dependent implementation. It'd be a bit > >> clearer to use the different impl class names. > >> > >> Naoto > >> > >> On 5/30/18 1:27 PM, Langer, Christoph wrote: > >>> Hi, > >>> > >>> I was just asked to include i18-n dev in the review of the change for > >>> this Item. > >>> Bug: https://bugs.openjdk.java.net/browse/JDK-8201429 > >>> > >>> Last webrev: http://cr.openjdk.java.net/~clanger/webrevs/8201429.3/ > >>> > >>> Unfortunately this request came a bit too late and the change has > >>> already been pushed: > >>> http://hg.openjdk.java.net/jdk/client/rev/668463f93ec0 > >>> > >>> So this is now merely a notification. However, maybe you can still > >>> have an eye on this or even do some specific testing. In case you see > >>> some issue with it, we shall open a follow-up item. > >>> > >>> Thanks > >>> Christoph > >>> > >
