Hi Philip, please find my comments enclosed in green
Rgds Andre
-----Ursprüngliche Nachricht----- Von: Philip Withnall [mailto:philip.withnall@collabora.co.uk] Gesendet: Freitag, 8. Januar 2016 11:04 An: Barkowski Andre (CM-CI1/PRM1) Andre.Barkowski@de.bosch.com; devel@lists.apertis.org Betreff: Re: AW: [Devel] Preferences and Persistence design 0.3.0
Hi Andre,
Sorry for my delayed response. Instead of diving straight into updating the document, I feel this topic (§5.1.5, user interfaces) needs a little more discussion. I think we are broadly on the same page, although your e-mail has introduced a couple more use cases which the design currently doesn’t cover. I will try to explain below.
In various places below I have said that the design covers a particular point. On re-reading the design when writing this e-mail, I realise there are places where it could be restructured to be clearer. I will try to do that.
On Mon, 2015-11-30 at 16:18 +0000, Barkowski Andre (CM-CI1/PRM1) wrote:
Enclosed my feedback.
- secrets & passwords ==> move to security paper
Its ok for me to move this scope to the security paper, with that we can also add more context to it like discussed in another thread. (see enclosed email). But please do not forget a reflection of rollback and its impact the "user secrets and password" approach and where to add this reflection (this paper - e.g in chapter 5 approach - or the security paper - which I feel would may fit better).
OK, I will expand the section on this in the Security design. I will leave the current section in the Preferences and Persistence design as a pointer to the Security design.
Andre: ok.
I have the feeling we are not on the same page so far. To make it easier to understand, let's call it "an iOS like" solution. There is a settings App, bundled with the system like an App-Launcher Application - providing the preferences to the User. An App carries stuff to be configured in some way, e.g. a manifest, this gets interpreted by some "Settings App" and shown to the User in a consistent way.
This is actually quite similar to what we have so far in the design: a GSettings schema (‘manifest file’) which specifies all the preferences for a given application. The design assumes some central preferences application, like iOS, which has sub-pages for each application.
Andre: ok.
The design currently gives a couple of options for how those sub-pages are implemented, but the differences are technical rather than user- facing. In all cases, the intention is that the preferences application lists applications, the user selects one, and all the preferences for that application are displayed.
Andre: the difference is responsibility & deployment. The settings App belongs to the system chrome, gets defined by the OEM and we expect to write (by us) one new implementation per product. But we keep the Apps out of that scope. They are not ruling how it looks like (by providing the UI implementation aka look & feel), they are only defining the data. In a model/view/control pattern, their focus is limited to the model.
However, no general rule w/o exception, we would like to keep the flexibility for special topics/data, which may need some more complex interaction to gets defined. But this doesn't change the overall approach, it only extends it. This would mean for very special entities (a very small quantity and with that an exception and not the rule of thumb), Apps can provide a UI which gets used to modify a value (e.g. like a wizard). For this part we again think about how to customize it to the maximum degree, but we accept limits.
Where the design currently differs from this is that it advises //against// automatically generating the preferences UI from the GSettings schema file. This is advice borne out of the experiences of various of us, who have seen and worked with auto-generated preferences UIs. They are almost universally not a pleasure to use, unless they list a very small number of very simple preferences (on the order of a couple of checkboxes, for example). You can think of this as a model–view–controller split: in all such patterns, it is rarely possible to satisfactorily generate the view from the model.
Andre: yes, I see, but I don't share this recommendation against it. Its not black or white, the majority is very simple and on top we provide flexibility for the exceptions and last but not least we have the need for OEM customization.
Instead, it advises creating preferences UIs using code. This allows for ultimate flexibility in how the UI works, which can make for a much more pleasant user experience.
Andre: again, I see but don't share it because it does not fit to our business model. Which user experience is pleasant decides the OEM, even if we find it very disappointing. With that we are talking about different kind of "flexibility". Since the majority of preferences is quite easy, the "flexibility for Apps" is only important for some special entities, but since we adopt the look&feel of each and every product accordant to the OEM interest, we do need "flexibility for OEM". And therefore we restrict flexibility for Apps. However, for parts with very good reasons, we will provide it, but not as a rule of thumb.
And implementing it using code is not the issue what I am talking about, the responsibility is the topic. The system chrome application providing the UI for the App settings (which has been created based on code) will be done by the product team, not the App Developer. With that we keep the maximum level of flexibility for the settings app (belonging to the system chrome) in our own hands (i.e. the product team, out of the App Developer hands) And like said a new implementation will be done for each new product. Its the same for the App-Launcher, the status bar / Home screen, etc.. But its not based on code to provide flexibility for the configuration, its based on code to provide flexibility for the OEM, even for the standard / easy stuff. The complex, very App specific stuff can not realized outside the scope of the App responsibility, so we keep this in their hands with limitations in customization (even though we push also the limits as much as possible to enable customization for that part to a maximum degree)
However, the approach which I explain is not going for one way without enabling the other one. So we have to enable both, minimize the part which is in the hands of the App Dev and maximize the part which is out of their scope.
The amount of code needed does not need to be large at all: there are various GSettings convenience APIs which mean binding a particular setting to a particular widget is typically 1 line of code.
Andre: great.
I think the main attraction of creating preference UIs using code, rather than auto-generating them is that it provides a simple progression path: if version 1 of an application has a couple of boolean preferences, it makes sense to auto-generate the preferences UI for them. However, if version 2 adds some more complex preferences which cannot have their UI auto-generated, the application developer has to reimplement the entire preferences UI in code, including the old preferences.
Andre: see above, its not black or white as also the flexibility for App Developers is not the first/only priority in our business environment.
So instead I propose a new solution (which is not in the current version of the design): write preference UIs in code, but Apertis should provide a widget which can auto-generate the UI for a specific set of preferences from the GSettings schema for the application. Let’s call the widget ApertisSettingsList.
This means that version 1 of the application above would implement its preferences UI in a small amount of code to instantiate the ApertisSettingsList widget for its two boolean preferences. Version 2 would keep this widget, and add additional widgets to the UI for the new, complex, preferences.
Wizards could be implemented as a similar progression.
Andre: like said, I do not have an issue with realizing the settings based on code, the topic is the responsibility of code as also the differnent kind of flexibility.
I would like to go for a solution, where the majority of flexibility / code is in the hands of the product team, which adopts it accordant to OEM interest, and the minority is in the hand of the App developer.
In your sample above, its still too much responsibility in the scope of the App-developer. The "ApertisSettingsList" is not able to cover enough customization from my point of view, we would need more like a a setings applications, complety in the responsibility of the product team, providing an implementation of all the easy / standard stuff accordant to the OEM interest, and hook to Apps for special entities, which then gets limited customization capabilities based on widget customization.
- A user do not has to step into an app and to search for it to
configure it, it’s at a central place.
This is already handled in the design. Andre: ok
- We also do not have to coordinate all the different kinds of
possibilities to incorporate it within Apps (at different places and in different way)
The design already covers this by specifying a standard way for an application to be launched so that it shows its preferences immediately (§5.1.5¶2; ActivateAction).
- An OEM can decide to remove configurable parts in his product, so
that only a subset gets published to the user for a specific device.
This is already handled by the vendor lockdown system supported by GSettings. Applications can check whether a preference is locked down by calling the g_settings_is_writable() method. I will mention that in the document.
Andre: "can" check is not sufficient. We need to deliver the product w/o contacting the App-developer to adopt this part. Again, it’s a matter of responsibilities. So the code is in responsibility of the product team to the maximum degree. For the remaining stuff we try to push the limits as much as possible, so "can" would be ok for that.
- An OEM can also change the look and feel of the Settings-App and it
keeps consistent, no need to change the App
This can be handled by applying CSS classes to the preferences windows in applications, which allows the OEM to override the CSS for that class and have the changes be applied to all preferences windows without needing to modify applications. This is something covered by the UI Customisation design. I will mention that in this document too.
Andre: like said above, this is sufficient for the special cases but not enough flexibility for the general one.
- One can optionally re-organize / combine configurations of same
kind over all Apps into one view (e.g. similar like "privacy" settings)
This should be possible for the simple preferences supported by ApertisSettingsList by creating a new view which instructs that widget to display the desired preferences from //several// GSettings schemas.
Note that it would not be possible for preferences whose UIs cannot be auto-generated.
Andre: see above
- We can easily add Search capabilities (over all Apps), to help
Users to find the topic of interest independent of their organization
This can be implemented by searching the summaries and descriptions of installed GSettings schemas.
Andre: yes, GSettings is not an issue.
Often there are "configurations" which modifies the App behavior but used by the system-integrator at point in time of product creation. By intention, they are not transported to the Used for configuration. It limits the capability of the User, but also reduces complexity. Its maybe used for variant handling. However, it is highly OEM specific, since another OEM my decide differently which configuration capabilities to provide to the User.
This is already handled by the combination of vendor lockdown and vendor overrides, as provided by GSettings.
Andre: I am not talking about the GSettings capabilities, I am talking about the code responsibility for the presentation, and who will change what in case an OEM adoption is running into issues. The Settings App / UI shall be in the responsibility of the product team to the maximum degree and interact with the OEM for adoption, but not with each and every App developer to check if its Code has issues with the new presentation solution.
But to be honest, settings configured by the central Settings App are not covering all App internals. There are settings changed dynamically during usage, e.g. Sat-View/Hybrid-View/Standard-View in a Map, routing parameter like fast/short/eco etc or even stored Favorites. This will be done App-internally, and typically spread over the App-UI to be at the place where it is needed (instead of collected and presented all together at a single place). These App internals are managed by the App itself. Also for this the prepared (preferences & setting) services can be used by the App-developer, and we provide prepared basic widgets, to keep the UI consistent and to customize it easily (and we should have some words about it in the document) but when I say we have to have a central App for the App settings, then I am talking about the "static" ones.
Indeed, the design recommends that preferences are integrated into the application UI where appropriate. This is definitely a use case we want to support.
For that we use some basic data structures like "bool", "int", "string" etc and provide widgets as visual representation of it. Lets compare it with the iOS way.
Since I haven’t mentioned it so far, both the iOS approach and the Android approach are relevant here (and I will include them in the document as background research):
iOS uses settings bundles, which are lists of preferences from which a UI is auto-generated. They support very simple data types, and can be loaded by the application or by a central preferences application.
Android uses preferences fragments, which are implemented in code as part of the application, and are launched by the preferences UI when needed, using a special launch argument to get the application to display that fragment (or present it as a separate Settings Activity).
Both approaches result in roughly the same user experience.
So as it stands, our design is aligned more with the Android approach than the iOS approach.
Andre: yip, but it looks like the iOS approach is fitting better to our business needs.
In addition, there could be a sequence needed to ease the input of values, e.g. a wizard (e.g. to configure an email account). This should be provided by the App-Developer. An isolated executable (Agent), using some PopUp-Sequence and finally set the key-value pairs shown in the settings App. With that, the Wizard gets shipped with the App, but launched by the Settings Apps.
The wizard, as you describe it, is effectively what we have in mind for all preferences UIs — you could think of them as similar to one-page wizards.
Andre: yes, but only for very special cases, not as the general approach.
Once we do have both mechanisms - the wizard together with the standard key/value types - , we have a great tradeoff between the 80% of easy to use Standard stuff and the 20% complex parts not fitting well into the standard patterns. And being well prepared to handle it in a central App, we can easily cover a look & feel where it behaves like bundled to each App (for example a dedicated "Settings Hard Key", which shall show only the Settings of the App (in Foreground at the point in time of key press), by initiating a flip Animation and showing the App-Settings at the Back-Side of the current App Screen. That said, from a solution for central handling we can derive a decentralized setup, but not as easy vice versa. So lets go for the one covering the wider scope.
I will add the settings hard key to the document as a requirement and figure out how to integrate it into the design.
However, even if we support the central solution, an App developer may decide not to use it. In this case, App settings - which still exists but not exposed centrally - are somehow internally handled, in an app-specific way. Its not consistent, etc, but its possible. And we shall support this too. Its like different levels of integration, similar like with the different level of prepared UI elements (from basic widgets via container to complete views/templates) - we recommend the deepest integration but if the App developer decides not to use it he can go for it but then he has to take the burden for build in functionality. So for "settings", we may expose a "settings key pressed" signal to the App for the case it does not participate on the central handling. Or the OEM customization in look&feel but also in scope. Or the "search settings" functionality. An App Developer can even decide not to use our low level services to handle key-value pairs and uses its own. In this case, he has to take care about under voltage etc,
OK.
With that in mind, I do have the feeling that the explanations on page 26, like
- disadvantages
- ways to embed UI from an application (A) into the system
preferences app (B) are not perfectly fitting.
Hopefully this e-mail has managed to clarify my reasoning behind page 26 a little more. At the moment I am leaning towards recommending ‘alternative model 2’ as the default recommendation.
Andre: yes, thx for that. However, it doesn't convince me and I hope you see why in our OEM business flexibility for App Developers is not the only driving goal.
Could you please reflect the things said above
- different kind of App settings (which ones are in-app handled,
which ones are central)
- going for the widest scope solution (central handling,
incorporation of wizards)
- flexibility in tailoring for a specific product (customization, de-
centralized setup)
I will ensure those are covered in the next iteration of the document, but would appreciate your thoughts on my explanations above first, to make sure we’re on the same page.
And also
- reflect the iOS User Interface approach for presenting the App
settings
Will do (and the Android one).
Philip
-----Ursprüngliche Nachricht----- Von: Devel [mailto:devel-bounces@lists.apertis.org] Im Auftrag von Philip Withnall Gesendet: Freitag, 27. November 2015 13:05 An: devel@lists.apertis.org mailto:devel@lists.apertis.org Betreff: [Devel] Preferences and Persistence design 0.3.0
Hi all,
Please find attached version 0.3.0 of the Preferences and Persistence design, plus a document showing the changes between version 0.2.3 and 0.3.0.
This version includes more background research, some alternative designs for the system preferences application (§5.1.5) and clarifications of recommendations for storing secrets and passwords.
Is this one OK to upload to the wiki?
Philip