On the subject, is there any possibility to test a plugin at another platform?
I'm developing my plugin on Windows machine and I have a Linux server, which I can test the plugin on but when it comes to OSX I have no machine to test it on. Does anyone have experience/good cookbooks how to set up a OSX virtual box or something similar to test a plugin in? Best regards Axel On Wed, 6 Feb 2019 at 13:05, Matthias Kuhn <matth...@opengis.ch> wrote: > Hi > > On 2/6/19 12:47 PM, Nathan Woodrow wrote: > > QGIS is for the users in the end, if this opens a new world of users, > > and note this is government funded, I don't think we should avoid > > allowing these kinds of things in. This isn't a core feature of QGIS > > and more of an exception than a rule. I don't see most plugins going > > this route but the few that do because of X reasons they have no > > control over should still be part of the community I think. > > > > I really don't see the harm here as it's not something that most > > people will know or care about in the end. Getting plugins to work > > cross-platform is a lot of work and I don't really like pushing people > > away because they can't do it for whatever reason. If this was going > > to be a core feature I would be on your side for it but the plugins > > are for the users. > > > > I would also note we have some features of QGIS that only work on some > > platforms and not others and we are fine with that because it makes it > > better for those users, so this is the same here IMO. Pushing them > > out to a self-hosted plugin repo just adds extra overhead and limits > > user base and visibility of well-funded work that we should embrace. > > > I am tempted to think in the same direction. Most plugins will be > cross-platform anyway because Python and because Qt and because QGIS. > > If they are not and it's trivial to solve (e.g. someone is using \ as > path separator instead of os.path.join...) than things can go through an > issue report. If it's more complex it's a question of effort vs. > benefit. If it's a single-purpose use-case specific plugin which is > intended to be run inside a specific organization / scenario, then I > don't see the benefit of investing a lot of effort into something that > will never be used. > > IMO we should be open to anyone and in parallel encourage people to > support as many platforms as possible. E.g. adding metadata with > supported platforms might help to be more transparent about supported > platform, filter plugins in the plugin manager and offer specific help > to make plugins cross-platform down the road. > > Bests > > Matthias > > _______________________________________________ > QGIS-Developer mailing list > QGIS-Developer@lists.osgeo.org > List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
_______________________________________________ QGIS-Developer mailing list QGIS-Developer@lists.osgeo.org List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer