Thanks for the link. The reporter clearly says he only has a problem on Seti. All other projects worked fine. So it seems to me that this is a Seti related problem that now got improperly fixed on the BOINC level where it can potentially do harm to other projects that don't have this particular problem.
Using third party APIs is always a bit problematic because if they change something we have to adjust our code. But this also means that you have to correctly use it. Using an undocumented parameter just increases the risk of breaking the API in the future (when google rightly decides to actually use this parameter) and nobody remembers why this parameter was added to BOINC. We could at least use one of the documented parameters if that solves the problem for Seti. Or we could investigate the real reason why this is not working on Seti but is working on other projects. There is already enough project specific code in BOINC. We should strive to decrease the amount and not put more project specific workarounds in. Regards Christian On 02.02.2017 23:29, Jord van der Elst wrote: > The report was done via > https://setiathome.berkeley.edu/forum_thread.php?id=80955 > I did ask the user if it truly was IE, seeing how he's using Windows > 10. It may have been Edge. No reply on that though. > But in that case it would be IE 11. Or Edge. > > > -- Jord van der Elst. > > On Thu, Feb 2, 2017 at 8:43 PM, Christian Beer > <[email protected] <mailto:[email protected]>> wrote: > > Quoting my own email for the other commit that actually adds the > parameter: > > Again, which problem is this trying to fix? Is there a report? > The used > > parameter is not listed as a supported one. See: > > https://developers.google.com/recaptcha/docs/display > <https://developers.google.com/recaptcha/docs/display> > > If the parameter is not supported this is not a fix. It's most likely > breaking again soon. I also don't see a difference when requesting the > script with "?ver=2" added (through wget with a faked user agent). > Also > what version of IE are we talking about? Is this workaround mentioned > somewhere else? Why doesn't the official implementation mention this? > > Regards > Christian > > > On 02.02.2017 20:29, David Anderson wrote: > > html/inc/recaptchalib.php > > > > <script src="https://www.google.com/recaptcha/api.js?ver=2 > <https://www.google.com/recaptcha/api.js?ver=2>" async > > defer></script> > > > > added ?ver=2 > > > > On 2/2/2017 2:53 AM, Christian Beer wrote: > >> So the extra linebreak after the script line fixes this? I > don't see any > >> other change that affects how the page is rendered. As it is > now the > >> commit message is misleading. > >> > >> Regards > >> Christian > >> > >> On 02.02.2017 11:36, David Anderson wrote: > >>> recaptcha didn't work on IE > >>> > >>> On 2/2/2017 12:00 AM, Christian Beer wrote: > >>>> On 02.02.2017 07:13, GitHub wrote: > >>>>> Branch: refs/heads/master > >>>>> Home: https://github.com/BOINC/boinc > >>>>> Commit: c2350c7f5747dc3109d7a33aa2f12a3161258fda > >>>>> > >>>>> > > https://github.com/BOINC/boinc/commit/c2350c7f5747dc3109d7a33aa2f12a3161258fda > > <https://github.com/BOINC/boinc/commit/c2350c7f5747dc3109d7a33aa2f12a3161258fda> > >>>>> Author: David Anderson <[email protected] > <mailto:[email protected]>> > >>>>> Date: 2017-02-01 (Wed, 01 Feb 2017) > >>>>> > >>>>> Changed paths: > >>>>> M html/inc/account.inc > >>>>> M html/inc/recaptchalib.php > >>>>> M html/inc/util.inc > >>>>> > >>>>> Log Message: > >>>>> ----------- > >>>>> web: try to get Recaptcha working on IE > >>>> I don't see how this commit fixes anything that is related to a > >>>> specific > >>>> problem with Internet Explorer. Is there a bug report? > >>>> > >>>> Regards > >>>> Christian > >>>> > >>>> _______________________________________________ > >>>> boinc_dev mailing list > >>>> [email protected] <mailto:[email protected]> > >>>> http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev > <http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev> > >>>> To unsubscribe, visit the above URL and > >>>> (near bottom of page) enter your email address. > >>> _______________________________________________ > >>> boinc_dev mailing list > >>> [email protected] <mailto:[email protected]> > >>> http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev > <http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev> > >>> To unsubscribe, visit the above URL and > >>> (near bottom of page) enter your email address. > >> > >> _______________________________________________ > >> boinc_dev mailing list > >> [email protected] <mailto:[email protected]> > >> http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev > <http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev> > >> To unsubscribe, visit the above URL and > >> (near bottom of page) enter your email address. > > > > _______________________________________________ > > boinc_dev mailing list > > [email protected] <mailto:[email protected]> > > http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev > <http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev> > > To unsubscribe, visit the above URL and > > (near bottom of page) enter your email address. > > > _______________________________________________ > boinc_dev mailing list > [email protected] <mailto:[email protected]> > http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev > <http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev> > To unsubscribe, visit the above URL and > (near bottom of page) enter your email address. > > _______________________________________________ boinc_dev mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
