[Interest] [Qml][type_id] Same component module have different type

[Jérôme Godbout]

I thing I found the culprit, it seem like the component was used into a Qml
(qmldir) singleton as QQmlComponent and QQmlEngine.trimComponentCache()
that was called to run different application part without relaunch.
But still weird that the component  got trim when still in used by the Qml
Singleton, since the singleton them self are never trim.
So I made a signal before/after the trimComponentCache() is called so I can
recreate the Qml Singleton component with Qt.createComponent().

On Tue, May 10, 2016 at 5:17 AM, Mitch Curtis <mitch.curtis at qt.io> wrote:

> Can you come up with a minimal example that we can run?
>
>
>
> *From:* Interest [mailto:interest-bounces+mitch.curtis=
> qt.io at qt-project.org] *On Behalf Of *Jérôme Godbout
> *Sent:* Monday, 9 May 2016 8:18 PM
> *To:* Interests Qt <interest at qt-project.org>
> *Subject:* [Interest] [Qml][type_id] Same component module have different
> type
>
>
>
> Hi,
>
> I have encounter a problem where Qml refuse to set a property because the
> type doesn't match, but they are the same type but got a different type
> name and id:
>
>
>
> *Unable to assign PartVisibleModel_QMLTYPE_1066 to
> PartVisibleModel_QMLTYPE_331*
>
>
>
> So I guess I did something wrong with it and end up with 2x the component
> definition.
>
> I checked I only have a single qmldir that declared that component
>
> PartVisibleModel 0.1 PartVisibleModel.qml
>
>
>
> No multiple version either, only the 0.1. The component only have 0.1
> version but the module have a 0.1 and 0.2 version. Is importing the module
> as 0.1 and as 0.2 can generate multiple type for component inside it that
> only have the lowest version defined?
>
>
>
> Is their a way to find out where this could have happen or why this is
> possible at all into a single QQmlEngine?
>
>
>
> Thanks,
>
> Jerome
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.qt-project.org/pipermail/interest/attachments/20160510/5a549e4a/attachment.html>