On Thursday, July 18, 2013 at 6:05 PM, Mounir Lamouri wrote:
> On 04/07/13 12:50, Marcos Caceres wrote:
> > ## Outstanding issues
> > If version is greater than the highest version number supported by the user
> > agent, should the user agent treat the manifest as the highest manifest
> > version it supports? Or reject the manifest outright as invalid?
>
>
>
> I think the application should be marked as incompatible. Imagine I have
> an old phone with a runtime that only understand manifest_version { 0,
> 1, 2 }. My phone doesn't supports newer versions but there are newer
> versions out there like { 3, 4, 5 }. If an application is advertising
> itself as manifest_version = 3, it is a hint that my phone will not be
> able to run it. It *might* be able to run it but nothing is less sure.
I'm really worried about this, as it's not very "webby": the manifest should
hopefully not specify things that would cause the application itself to not be
runnable (i.e., manifest stuff should just be harmless metadata with limited
side effects … though CSP tightening already breaks this principle a bit).
Having apps accessible through the Web browser, but then not installable
because of a manifest version would be weird for hosted apps (which shouldn't
need to be installed to be usable - as not all user agents may support
installation - or users may want to "try before they buy/install").
It would also mean that certain devices would be tied to certain manifest
versions (it's a nasty way to lock people out, without providing graceful
degradation). This might lead to all sorts of version/device mappings, which
would not be great IMO.
I think we should adopt a "do minimal harm" principle when adding new stuff to
the manifest: meaning that stuff that is not understood is just ignored, ala
HTML parsing style.
> > If a manifest version number is invalid (i.e., NaN, <= 0 ), is the manifest
> > invalid?
>
>
> I think we should use manifest_version = 0 as the current default
> version. Hopefully no one will ever write manifest_version = 0 because
> we would introduce that feature at the same time as we would introduce
> manifest_version = 1.
I don't think the version number we start from actually matters from a
technical standpoint. However, it might confuse people to see version 0 now and
again. Version 1 is more "natural".
> Also, I would like to think about the path to standardisation. Google
> Chrome Apps have the same concept (manifest_version) using numbers too.
Yes. The proposal is more or less identical - though I don't know how they do
their parsing.
> Should we have a special token like "w3c" or "standard" that will be
> required to use the standard manifest version? I think that would be
> reasonable even though that means that if some applications were using
> "w3c" as a value, it might break them :(
>
I would like to treat the formats as completely separate (we have different
ways of installing these manifests as well as different MIME types, which
serves as the differentiator). Any overlap between the two should be seen as
coincidental - otherwise, things are going to get really messy.
_______________________________________________
dev-webapps mailing list
dev-webapps@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-webapps
