There should be something around syncToServer + HtmlUnit causes the issue (I guess RPC gets called back sooner than timeout(0) but didn't debug) but I don't think this in practice will effect anything else.
Enabling batch mode by default for everything potentially effects timeouts also sometimes it might be incompatible with the test runner that drives the java side of the tests (for example it doesn't work well with our internal test sharding mechanism). Hence I don't think it worths the trouble. I think just having batch mode enabled for junit tests is good enough to move forward. Daniel's patch is important for both hand written jsinterop code and Elemental2 and yes unfortunately Scheduler.scheduleFinally will be affected. PS: pls avoid the urge to discuss technical stuff in steering commitee and try to keep it in the contributor list ;) On Wed, Oct 11, 2017 at 9:59 AM, Thomas Broyer <t.bro...@gmail.com> wrote: > As discussed minutes ago in meeting: here's the patch to enable -batch > module for all our HtmlUnit tests: https://gwt-review. > googlesource.com/c/gwt/+/19740 > Once that one and the HtmlUnit workaround are in, we can rebase Daniel's > patch about trapping window.onerror by default. > > On Wednesday, October 11, 2017 at 10:52:41 AM UTC+2, Thomas Broyer wrote: >> >> >> >> On Wednesday, October 11, 2017 at 4:52:06 AM UTC+2, Goktug Gokdogan wrote: >>> >>> tbroyer: Are you using batch mode while testing? We are using -batch >>> module internally and maybe you guys do not externally? (though you linked >>> the code that only runs in batch mode, IIRC). >>> >> >> That's right, we do not use batch mode, **and** it fixes the tests! \o/ >> >> I wonder, should the default batching strategy be changed to module? or >> should we use "-batch module" for all tests? only HtmlUnit tests? (we only >> run those on CI anyway; also I tested in Firefox with -runStyle >> ExternalBrowser:firefox, and the tests pass without the need for -batch >> module) >> What would you recommend? >> >> Colin: do we try to slip this into 2.8.2 at the last minute? (and >> possibly revert if there's any red flag during smoke testing) >> >> Daniel/Goktug: iiuc, this change means that you no longer need to wrap >> everything into $entry() to get the UncaughtExceptionHandler called on >> errors, right? This makes things easier with JsInterop and @JsFunction >> callbacks (or exposing objects whose @JsMethods will be called from JS), >> but as a non-negligible side effect you won't get the >> Scheduler.scheduleEntry and Scheduler.scheduleFinally commands called. Am I >> getting things right? >> >> >>> >>> On Tue, Oct 10, 2017 at 1:04 PM, Thomas Broyer <t.bro...@gmail.com> >>> wrote: >>> >>>> I'll try with the manual runstyle, i.e. with a real browser, and see >>>> how it goes. Thanks for the feedback. >>>> >>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "GWT Contributors" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to google-web-toolkit-contributor >>>> s+unsubscr...@googlegroups.com. >>>> To view this discussion on the web visit https://groups.google.com/d/ms >>>> gid/google-web-toolkit-contributors/1fab1f52-557b-4857-a0a9- >>>> f017c1d4822d%40googlegroups.com. >>>> For more options, visit https://groups.google.com/d/optout. >>>> >>> >>> -- > You received this message because you are subscribed to the Google Groups > "GWT Contributors" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to google-web-toolkit-contributors+unsubscr...@googlegroups.com. > To view this discussion on the web visit https://groups.google.com/d/ > msgid/google-web-toolkit-contributors/f9fcbd96-ae33- > 430e-8fda-1cca8ad8b07a%40googlegroups.com > <https://groups.google.com/d/msgid/google-web-toolkit-contributors/f9fcbd96-ae33-430e-8fda-1cca8ad8b07a%40googlegroups.com?utm_medium=email&utm_source=footer> > . > > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "GWT Contributors" group. To unsubscribe from this group and stop receiving emails from it, send an email to google-web-toolkit-contributors+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAN%3DyUA2%3DjAqhYoWKTBYriWDp3ufvR3_N5t8LaXyfSd-gwrUOAQ%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
