Hi Yoav, We don't have existing metrics that would help here, and I honestly don't even know what we would track for this specific issue (detecting that a certain exception came from eval, for example, isn't something we currently track).
Bug reports are how we tend to track issues like this. And fwiw my intuition matches Sathya & Ross's (that this is very unlikely to cause problems in practice). On Mon, Jun 17, 2019 at 9:42 PM Yoav Weiss <y...@yoav.ws> wrote: > LGTM2 > > Do we have metrics in place in case we got it wrong and sites do rely on > eval's error types in weird and unexpected ways? Or we're relying on bug > reports to alert us of that? > > On Tue, Jun 18, 2019 at 6:34 AM Chris Harrelson <chris...@chromium.org> > wrote: > >> I see. Given that lack of interop, I do agree that probably sites are not >> relying on the current behavior, and if they are then they are likely not >> interoperable. >> >> LGTM1 to try. >> >> On Mon, Jun 17, 2019 at 3:02 PM Ross Kirsling <rkirsl...@gmail.com> >> wrote: >> >>> >>> >>> On Mon, Jun 17, 2019, 11:42 Chris Harrelson <chris...@google.com> wrote: >>> >>>> Hi again, >>>> >>>> Sorry for the slow response. >>>> >>>> On Thu, Jun 6, 2019 at 7:17 AM Sathya Gunasekaran <gsat...@chromium.org> >>>> wrote: >>>> >>>>> >>>>> I don't expect to see code that special cases against ReferenceError >>>>> in the wild especially for just early errors. I expect to see code >>>>> that handles both ReferenceErrors and SyntaxErrors. Since we're not >>>>> adding a new type of error, I wouldn't expect to see any breakage. >>>>> >>>> >>>> Do you have any data to support this? Are there any use counters for >>>> how often this situation might occur? e.g. what fraction of pages >>>> trigger ReferenceErrors? >>>> Also, have any other browsers shipped this change yet? >>>> >>> >>> The concern here is purely around *early* ReferenceError. Since >>> ReferenceError has been ambiguous between parse time and runtime, it would >>> have been extremely difficult/fragile to use this information—as mentioned, >>> one would need to string-match on error messages. (I think folks are hung >>> up on the phrase "observable concern", but I meant this to indicate that >>> there's something non-cosmetic to be *gained*, not that there's worry about >>> breakage.) >>> >>> The new behavior shipped in Safari TP 83 three weeks ago, and it's >>> important to note that the major engines were not previously in alignment >>> anyway. V8 was conforming to spec, but JSC actually had *late* >>> ReferenceError for all four cases and SM was split (`0++` and `0--` were >>> early SyntaxError while `0 = 0` and `0 += 0` were early ReferenceError). >>> >>> -- > You received this message because you are subscribed to the Google Groups > "blink-dev" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to blink-dev+unsubscr...@chromium.org. > To view this discussion on the web visit > https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACj%3DBEhQtWJoxxi3PwfutmSbZFJGyNkKpgD1OeMFDM6%3D2sA4bw%40mail.gmail.com > <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CACj%3DBEhQtWJoxxi3PwfutmSbZFJGyNkKpgD1OeMFDM6%3D2sA4bw%40mail.gmail.com?utm_medium=email&utm_source=footer> > . > -- -- v8-users mailing list v8-users@googlegroups.com http://groups.google.com/group/v8-users --- You received this message because you are subscribed to the Google Groups "v8-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to v8-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/v8-users/CAEvLGcKV93MirncPshaX9z29%3D5-0aBcecEOJ-Z%3D0zZE1%2BQb%3D7w%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
