cordova-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Carlos Santana <csantan...@gmail.com>
Subject Re: [DISCUSS] Core Plugins and Android API 23
Date Thu, 14 Jan 2016 22:02:43 GMT
I agree with final compromise, I also had a discussion with Nikhil during
holiday break offline and understood the situation, some 3rd party plugins
are not open source and customers/devs don't move as fast as we wish, but
having a CLI 6 with latest plugins being backwards compatible with lower
Android API is good base for developers to start upgrading.


On Thu, Jan 14, 2016 at 1:23 PM Steven Gill <stevengill97@gmail.com> wrote:

> Awesome! I'll kick up a discuss thread
>
> On Wed, Jan 13, 2016 at 5:42 PM, Richard Knoll <riknoll@microsoft.com>
> wrote:
>
> > Hey all,
> >
> > This change has been merged into the file, camera, geolocation, and
> > contacts plugins (media did not actually require any change). I was able
> to
> > build and run mobilespec on cordova-android 4.0.0 and cordova-android's
> > master. We should probably push for a plugins release soon, since there
> are
> > also some major bug fixes to the contacts and camera plugins that were
> > recently pushed.
> >
> > Thanks,
> > Richard
> >
> > -----Original Message-----
> > From: Nikhil Khandelwal [mailto:nikhilkh@microsoft.com]
> > Sent: Friday, January 8, 2016 12:12 PM
> > To: dev@cordova.apache.org
> > Subject: RE: [DISCUSS] Core Plugins and Android API 23
> >
> > We had a good discussion about this offline with Joe, Simon, Steve,
> Jesse,
> > Parashu and Jason. We came to the following agreement:
> > This is a temporary fix to the plugins which we need to take to ensure
> > Cordova 5.x tools behavior of using the edge plugins does not break for
> our
> > users. We will pursue the change for all the affected plugins. This will
> go
> > out in the next PLUGINS release which will happen before the Cordova 6.0
> > release. Since this a temporary fix, we aim to remove it in 6 months
> > (giving reasonable time for our user base to move to Cordova 6.x+ which
> > will have a better plugin version pinning mechanism).
> >
> > Thanks,
> > Nikhil
> >
> > -----Original Message-----
> > From: Simon MacDonald [mailto:simon.macdonald@gmail.com]
> > Sent: Thursday, January 7, 2016 4:00 PM
> > To: dev@cordova.apache.org
> > Subject: Re: [DISCUSS] Core Plugins and Android API 23
> >
> > I'm going to play devils advocate a bit here:
> >
> >
> > > There are valid reasons why many app developers might not be ready to
> > > move to API level 23:
> > >   1) They have an app which is stabilizing or in maintenance mode and
> > > they don't want to risk destabilization by moving to a new major
> > > Cordova version, Cordova Android platform, and Android API level.
> > >
> >
> > If the developer is not ready to move to a new version of Cordova Android
> > or the Android API would it not also follow that they shouldn't be
> changing
> > their plugin versions?
> >
> >
> > >   2) They are using a 3rd-party plugin which has not yet been updated
> > > to request Android permissions as required by API level 23. There are
> > > probably a lot of plugins affected, since access to any of the
> > > following things on Android M requires runtime permission requests:
> > > calendar, contacts, phone, camera, microphone, location, beacons,
> > sensors, SMS, storage.
> > >
> >
> > Right, so if the 3rd party plugin they are using doesn't support Android
> M
> > they should either a) not upgrade or b) send a PR to the plugin
> maintainer
> > so everyone can benefit.
> >
> >
> > >   3) They might not have the capacity or ability to test their app on
> > > devices running Android M. Because API 23 enables the new permissions
> > > model only on Android M, it requires testing on that platform.
> > >
> > >
> > Everyone has access to the Android emulator.
> >
> >
> > > In any of the above cases, developers might still like to benefit from
> > > some of the major bug fixes in those 5 popular core plugins mentioned
> > > below. Or even if they weren't specifically looking for bug fixes, it
> > > would be a much better experience if adding or updating one of those
> > > plugins would just work, rather than failing on Android. The
> > > explanation for the failure will not be obvious to many users, if they
> > > overlooked the warning when installing the plugin or if they were
> > > using another tool to add the plugin where the warning wasn't surfaced.
> > >
> >
> >  Seems like you are describing a tooling problem here. If the tool
> doesn't
> > surface the warning or allows the user to add a plugin that is
> incompatible
> > with the version of Cordova Android that is being used really sounds
> like a
> > bug in the tooling to me.
> >
> >
> > > Of course developers should be encouraged to upgrade to the latest
> > > most secure highest-quality version of Cordova. But the encouragement
> > > does not need to be so forceful. This proposed change gives developers
> > > more time to upgrade, and allows for more choice about when to upgrade
> > > individual parts
> > > (plugins) rather than limiting them to all-or-nothing.
> >
> >
> >  I would argue that the developer has an infinite amount of time to
> > upgrade. Nothing is forcing you to upgrade to the latest Cordova Android
> or
> > Android API. They can continue to use the same version of Cordova Android
> > and plugins that are currently working in their app. If the developer is
> > using semver properly and we do our job right they shouldn't pick up
> > breaking changes. For instance, if the API of a plugin changes we bump
> the
> > major version of the plugin so users who setup their config.xml to use:
> >
> > <plugin name="cordova-plugin-camera" spec="^1.1.0" />
> >
> > will effectively prevent the user from picking up the breaking change in
> > camera version 2.0.0.
> >
> > Simon Mac Donald
> >
> >
> https://na01.safelinks.protection.outlook.com/?url=http%3a%2f%2fhi.im%2fsimonmacdonald&data=01%7c01%7cnikhilkh%40microsoft.com%7c84fb6405edf74f47453208d317beb5e5%7c72f988bf86f141af91ab2d7cd011db47%7c1&sdata=B8DuMuPF8aUU6xvIlpYTi3LJPs1HbENcHXS8JrHqEu0%3d
> >  B KKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKCB
> >   [  X  ܚX K  K[XZ[
> >    ] ][  X  ܚX P  ܙ ݘK \ X  K ܙ B  ܈ Y  ] [ۘ[    [X[     K[XZ[
> >    ] Z [    ܙ ݘK \ X  K ܙ B
> >
>

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