On Mon, 8 Mar 2021 07:46:54 GMT, Yasumasa Suenaga <ysuen...@openjdk.org> wrote:
>> src/jdk.accessibility/windows/native/common/AccessBridgeDebug.cpp line 80: >> >>> 78: uli.LowPart = ft.dwLowDateTime; >>> 79: uli.HighPart = ft.dwHighDateTime; >>> 80: return (uli.QuadPart / 10000ULL) - 11644473600000ULL; // Rebase >>> Epoch from 1601 to 1970 >> >> This is good and true to the original change; >> >> I am not even sure the epoch rebase is needed. All 8196681 did was to print >> out the timestamps verbatim. I do not know enough about how that debug trace >> is used, and if the timestamps really have to be 1970 based. > > It seems to be fine not to use UNIX epoch at first glance as long as we can > know the timing of events, but I'm not sure (Thus I rewrote to comply with > the original code). So I want to hear from the others. Did you check via some a11y tool that the new code actually works? ------------- PR: https://git.openjdk.java.net/jdk/pull/2859
