On Mon, Jul 14, 2014 at 5:02 PM, Colin Alworth <[email protected]> wrote:
> Sorry for the delay in getting back to you (and to my reviews, got at > least one up to date now), finally catching up after a few days off. > > I guess I was looking for "We want to use Object.create in Core" in your > initial email. If we also wanted any/all of the features I had listed (fast > byte[]/int[]/double[] for everyone? rpc-over-ws? cors?), dropping ie9 from > Core might have also made sense. > BTW, with Core, I mean literally com.google.gwt.core.Core (which includes the java.lang.emul as well) and we are not really dropping support; just changing the defaults when useragent is not available. Currently IE8 was good enough to get the benefit I was looking for (i.e Object.create(null)) but I think we can simply bump up the minimum version to IE10 if there are other good benefits. I think it is fair to require UserAgent dependency for anything older. > > I'm not actually encouraging cutting IE9 (or 8), esp from User, but if we > want to move some emulation code off to UserAgent or User, letting go of > IE9 may make sense. > > My email was written from the perspective of "huh, Goktug wants to drop > IE8 because it will make *something* easier - won't also dropping IE9 make > more something even more easier?". With the caveat that all you are > interested in is Object.create, targeting only IE8 makes sense. > > Yep, exactly :) > On Tuesday, July 1, 2014 12:02:50 AM UTC-5, Goktug Gokdogan wrote: > >> >> >> >> On Mon, Jun 30, 2014 at 8:46 PM, Colin Alworth <[email protected]> wrote: >> >>> Sounds great, but is there a reason that we're now starting at IE9+ and >>> not IE10+, thus giving us typed arrays, web workers, web sockets, etc? I >>> only ask because the kind of case where you are giving up User (and Widget, >>> RPC, Timer, and other fairly high-level apis) seems to suggest that you >>> might not be writing for a browser at all (htmlunit, nashorn, web worker, >>> node.js). >>> >>> >> A cross-compiled app is a good example that doesn't need User where you >> can, for example, use closure to develop the UI. >> >> I specifically pointed IE8 as it is the only supported browser missing >> Object.create functionality and such apps that just depends on java.emul >> are paying the price of IE8. On the other hand by just inheriting >> useragent.UserAgent (not necessarily the User) an app can target older >> browsers. >> >> >>> Dan definitely has a point that if we're supporting modern browsers for >>> a core chunk of functionality, we really shouldn't let 'modern' be >>> 'whatever junk still happens to be running rather tha updating'. And >>> besides, I can't always be That Guy pushing to keep all versions forever, >>> just because IE8 is still 11% of North America's browser usage (really: >>> http://theie8countdown.com/). >>> >>> If we're cutting a browser for being old/bad/whatever in Core, but >>> leaving support for it still in User, we should consider carefully why we >>> *aren't* cutting deeper. >>> >>> >> Can you be more specific? >> >> >>> >>> On Monday, June 30, 2014 2:59:12 PM UTC-5, Goktug Gokdogan wrote: >>>> >>>> We are planning to drop support for IE8 if the application doesn't >>>> inherit c.g.gwt.useragent.UserAgent and hence not have browser >>>> permutations. >>>> >>>> Nearly all of today's apps inherit User so they will not be affected by >>>> this change. In the future more apps will only inherit Core however they >>>> shouldn't need to pay price of IE8 support (currently they do because there >>>> are no permutations in Core). >>>> >>>> Let me know if you have any concerns. >>>> >>>> - Goktug >>>> >>> -- >>> 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 [email protected] >>> . >>> To view this discussion on the web visit https://groups.google.com/d/ >>> msgid/google-web-toolkit-contributors/45b68163-0d07- >>> 4a6c-9932-412232e2f71d%40googlegroups.com >>> <https://groups.google.com/d/msgid/google-web-toolkit-contributors/45b68163-0d07-4a6c-9932-412232e2f71d%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 [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/google-web-toolkit-contributors/8e78df45-4d71-4dcf-9fea-52d32fc58d65%40googlegroups.com > <https://groups.google.com/d/msgid/google-web-toolkit-contributors/8e78df45-4d71-4dcf-9fea-52d32fc58d65%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 [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAN%3DyUA1SA9EW%3Dz5-XGOZy2o-ykY%2BQiZvNrtkaJ8Jyg3h-p%3Dx6A%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
