[dev] Dynamic view prefs removed?

[Michael J Rubinsky]

Quoting Michael M Slusarz <slusarz at horde.org>:

> Quoting Michael J Rubinsky <mrubinsk at horde.org>:
>
>> Why were the dynamic view prefs removed in develop?
>>
>> A considerable amount of time was spent making those work - these  
>> prefs were added to allow using a single application in traditional  
>> mode, while still using dynamic mode for everything else that  
>> allows it. This was a requested feature from users that wished to  
>> continue using a traditional interface when it contained  
>> functionality that was unavailable on the dynamic version.
>
> Not in IMP's case.  It was created to allow a user to indicate their  
> view preference in the case where hordeauth was being used, so the  
> user would NEVER be able to select their view mode.

That may be why it was originally added pre-4.0, but it is not what it  
was being used for in 4.0.

> It was a hack.  It was NEVER meant to allow a user to login via the  
> dynamic mode and then decide that they want to view IMP in  
> traditional mode instead.

Of course you are right, that wouldn't make sense if we are only  
talking about IMP. In 4.0 though, we moved this choice outside of  
IMP's code and into Horde proper. With the addition of the other  
dynamic apps (and somewhat incomplete implementation of some features)  
it made perfect sense to allow some of the applications to be shown in  
dynamic view, and other ajax enabled apps to be shown in traditional  
view - at the user's discretion.

> That may be possible with the current setup, but that is incorrect  
> behavior and an unfortunate side-effect.

Well, I actually spent a fair amount of time, with help from others,  
getting this side-effect to work this way so I would say it's not  
unfortunate. I would say it provides flexibility for the users.

> Going forward, if you login with dynamic mode, an application (if it  
> has a dynamic mode) should show you its dynamic mode.  There should  
> not be a preference overriding this.  Which makes sense - if I log  
> via traditional mode on my mobile device, I should never see an  
> application give me dynamic mode.
>
>> I know going forward we are aiming to only support a single desktop  
>> view, but until the traditional views are actually removed I think  
>> we should still support this feature.
>
> I disagree.  And I did not realize there were plans to remove the  
> traditional view.  In IMP's case, that should not happen.  Although  
> I don't personally agree, many people prefer the traditional mode.

..and this is a perfect reason to keep this ability. Some people may  
like the traditional version of IMP for whatever reason, but yet still  
want to use the dynamic view of Kronolith. As long as we keep both  
versions, we should offer the ability to use them in a flexible way.  
With the current code, this is impossible.

>> This will also prevent us releasing an application with a new  
>> dynamic view until it is *fully* functional. As an example, Hermes  
>> has a work-in-progress dynamic view that is functional for 90% of  
>> the use-cases, but still lacks some functionality. It is far easier  
>> to deselect the option for using the dynamic view by default then  
>> it is to have to logout then login again under traditional to use  
>> those features, then log out and in again to continue using e.g.,  
>> IMP's dynamic view.
>
> The new system is designed to allow applications that don't have  
> multiple views to work seamlessly.  It is the application's decision  
> which view to display.  That is why 4.0 applications work fine with  
> 4.1 applications.

Yes...and all I'm asking is that as part of this decision these  
applications can continue to consult a user pref as part of this  
decision, as they had been doing in 4.0. I feel this is requirement as  
long as we provide both interfaces...or at the very least until any  
dynamic interface has 100% of the feature set of the traditional  
interface. Otherwise, we are forcing the user to logout/login anytime  
they wish to use on of these features.

-- 
mike

The Horde Project (www.horde.org)
mrubinsk at horde.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 6096 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.horde.org/archives/dev/attachments/20120204/e9d90108/attachment.bin>