kreport include files

Adam Pigg adam at piggz.co.uk
Fri Apr 7 11:31:12 UTC 2017


Yeah, you are right. Perhaps for 3.1 we should fix this as we are then
supposed to be offering abi stability.

Thoughts jaroslaw?

On Fri, 7 Apr 2017, 12:28 Dag, <danders at get2net.dk> wrote:

>
>
> Adam Pigg skrev den 2017-04-07 13:21:
> > The code is probably copied from kdb!
> >
> > I think this is saying that kreport will only load plugins which claim
> > to be the same version, so nothing would prevent an out of source
> > plugin, but its version number must match for it to be loaded, this
> > way multiple versions can coexist, and only load the appropriate
> > plugins.
> Yes, but what is an appropriate plugin?
> Imho releasing a new kreport version does not neccessarily invalidate
> plugins.
> I cannot see that a plugin developed for 3.0 could not be usable with
> 3.1 if the plugin interface is not changed, or am I missing something?
>
> >
> > On Fri, 7 Apr 2017, 12:12 Dag, <danders at get2net.dk> wrote:
> >
> >> Adam Pigg skrev den 2017-04-07 11:37:
> >>> I can try a clean build later, but it certainly should work!
> >> ok.
> >>
> >> While we are looking into kreport, I have a question about kreport
> >> plugins.
> >> I find this code in KReportPluginManager.cpp:
> >> const QString expectedVersion = QString::fromLatin1("%1.%2")
> >>
> >>
> > .arg(KREPORT_STABLE_VERSION_MAJOR).arg(KREPORT_STABLE_VERSION_MINOR);
> >> foreach(QPluginLoader *loader, offers) {
> >> //QJsonObject json = loader->metaData();
> >> //kreportDebug() << json;
> >> //! @todo check version
> >> QScopedPointer<KReportPluginEntry> entry(new
> >> KReportPluginEntry);
> >> entry->setMetaData(loader);
> >> const KReportPluginMetaData *metaData = entry->metaData();
> >> if (metaData->version() != expectedVersion) {
> >>
> >> I thought the metaData->version() should be some sort of plugin
> >> interface version, not the version of KReport itself?
> >> Afaics with current solution development of plugins outside of
> >> kreport
> >> is not possible.
> >> Have I misunderstood it completely, or maybe it is intentional?
> >>
> >> ---
> >> Cheers,
> >>
> >> Dag
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kexi-devel/attachments/20170407/7ab5a673/attachment.html>


More information about the Kexi-devel mailing list