cordova-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ian Clelland <iclell...@chromium.org>
Subject Re: [Testing] No content to speak of
Date Mon, 25 Aug 2014 23:31:14 GMT
The changes to Entry.js introduce this failure; for some reason the
reference to entry.filesystemName was replaced with entry.filesystem.name,
and so the entry is not being returned with a reference to the correct
filesystem.

I think that this was a change made to accommodate Windows Phone, which
broke Android (and probably iOS). There's a fundamental difference between
how we have to pass the filesystem object across the bridge vs. how we can
pass it around in JavaScript, and this change makes one style work, at the
expense of the other.

I'll see this evening if I can work up a version that will work for both,
and get some WP8 testers to look at it in the morning.

Ian


On Mon, Aug 25, 2014 at 7:06 PM, Jesse <purplecabbage@gmail.com> wrote:

> Which tests are now failing, what do they test, and how have you verified
> that they should pass? Just because they used to pass, does not mean they
> are valid, or should have passed.
>
> This change to DirectoryEntry.js makes perfect sense, and may have
> uncovered a previously passing test that should have failed.
>
> -    if (!/\/$/.test(nativeURL)) {
> +    if (nativeURL && !/\/$/.test(nativeURL)) {
>          nativeURL += "/";
>      }
>
> Similarily, the changes to www/resolveLocalFileSystemURI.js and
> www/Entry.js do not appear to introduce a new failure, but could be
> uncovering previously falsely passing test(s)
>
>
>
> @purplecabbage
> risingj.com
>
>
> On Mon, Aug 25, 2014 at 3:56 PM, Martin Gonzalez <
> martin.c.glez.glez@gmail.com> wrote:
>
> > The tests shouldn't be failing. I agree with Ian, the tests started to
> fail
> > after:
> > 0ffb969 (Fixes multiple mobilespec tests
> > errors). I've verified with the previous commit and everything works as
> > expected.
> >  On Aug 25, 2014 5:46 PM, "Jesse" <purplecabbage@gmail.com> wrote:
> >
> > > I was just noting a possible reason for the tests to now fail.
> > > Has anyone verified whether they should fail or not?
> > >
> > > @purplecabbage
> > > risingj.com
> > >
> > >
> > > On Mon, Aug 25, 2014 at 3:27 PM, Michal Mocny <mmocny@chromium.org>
> > wrote:
> > >
> > > > The 4 failing file tests were reported on JIRA (
> > > > https://issues.apache.org/jira/browse/CB-7093) and I thought
> > > Martin/Jesse
> > > > were looking at them.
> > > >
> > > >
> > > > On Mon, Aug 25, 2014 at 5:39 PM, Marcel Kinard <cmarcelk@gmail.com>
> > > wrote:
> > > >
> > > > > Yes, all the other automated tests (except the 4 File tests as
> > > previously
> > > > > noted) are running successfully, so that does seem to point
> somewhere
> > > > else
> > > > > besides the bridge.
> > > > >
> > > > > The issue appears to be related to CB-6764. If I remove from
> > > > mobile-spec's
> > > > > index.html the script reference to "doesnotexist/notcordova.js"
> then
> > > the
> > > > > device object and the yellow box gets populated properly.
> > > > >
> > > > > On Aug 25, 2014, at 4:23 PM, Ian Clelland <iclelland@chromium.org>
> > > > wrote:
> > > > >
> > > > > > Usually an empty yellow box on startup means that the bridge
> isn't
> > > > > working,
> > > > > > but if you're seeing ~300 other successful tests, then that's
> > > unlikely
> > > > to
> > > > > > be the problem.
> > > > >
> > > > >
> > > >
> > >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message