I agree! I think you and I are actually in agreement, just using different terms. I'm using portrait and landscape to literally mean the viewing area is taller or wider. You are using it to mean the device is upright or sideways. I agree that their are applications that will need to know both, and I am all for adding a (generic, if possible) API that supports telling that to the app.
But if I am an application writer and wanted to use different layouts depending on the tallness or wideness of the viewing area (e.g., as iPad Mail does), then I would make my canvas 100% and look at it's dimensions (but I'd also like to rely on the LFC to automatically keep my app "upright"). I'd want the LFC on a tablet to make my app think that all that happened was the screen resized from 960x480 to 480x960 and not make me worry about which way is up. I can imagine plenty of other apps, like the references you sent, that will want to _not_ have that happen. That will want to handle their own rotation, so we should do that too. I guess this is what the iOS "do you want me to auto-rotate you, because the world is spinning" event is about? I want it all. I think you do too! On 2010-07-12, at 19:19, Raju Bitter wrote: > So my last question is: if I have a device with a square display and > tilt it, shouldn't the screen rotate as it would for a device with an > uneven aspect ratio? Autorotation of the display shouldn't be bound to > devices with non-square displays, I'd say. > > But if I have to take a beating for trying to improve OpenLaszlo, > maybe I shouldn't ask questions in the future. > > On Tue, Jul 13, 2010 at 1:11 AM, Raju Bitter > <[email protected]> wrote: >> No, Tucker, we have the same concept of what portrait and landscape >> mean. But Android will judge that an application is in landscape mode >> using the accelerometer data, not caring if you have a square display. >> Take that Motorola phone. If you hold it upright, tilt it left 90 >> degrees then the accelerometer will send the Android app an event >> saying: "We are going into landscape mode", no matter what the screen >> resolution of the display is. Maybe that's Google/Android's engineer's >> misinterpretation of landscape, so complain to Google if you want. >> >> >> On Tue, Jul 13, 2010 at 12:37 AM, P T Withington <[email protected]> wrote: >>> On 2010-07-12, at 18:09, Raju Bitter wrote: >>> >>>> You are welcome, Max. >>>> >>>> Thanks for your comment, Tucker. I understand what you mean, but I >>>> think it's more complex than it initially seems to be. >>>> >>>> What about browser APIs? If the browser has orientation information, >>>> what would be the proposed way by the OpenLaszlo team to retrieve that >>>> information in both Flash and DHTML runtime? >>> >>> I have no objection to exposing this information. >>> >>> LZX has two ways it does this: >>> >>> 1) We try to come up with a platform-neutral API and map the >>> platform-specific into that. >>> >>> 2) We never prevent people from calling directly into the >>> (undocumented/unsupported, on LZX's part) platform API when the need to >>> (unlike Apple). Go ahead. Make your innovative app that depends on an API >>> we have not yet codified in LZX. We want to see people innovate. But >>> please don't complain when your app breaks! >>> >>>> @Tucker: Regarding your question "If the app wants to present >>>> different layouts based on landscape or portrait orientation, why >>>> should it base that on some hardware device sensor? I should be able >>>> to resize my browser window and get the same layout." >>>> There are quite a few applications which don't autorotate the whole >>>> app but content within the app. Here's a blog post of a developer >>>> describing that: >>>> http://blog.sallarp.com/shouldautorotatetointerfaceorientation/ >>> >>> I don't get your point here. Except that I think it re-enforces my point. >>> I want simple apps to not have to even know they are running on a device >>> that can be turned upside down. My eyes glazed over as soon as I saw the >>> diagrams on that blog where the coordinate system goes upside down and >>> backwards and he has to deal with it in his app! >>> >>>> Square displays and portrait/landscape mode: If you an Android based >>>> phone, the OS is relevant, and Android generally supports portrait and >>>> landscape modes. Within Android apps, you can have different views >>>> with different behavior regarding autorotation of content for each >>>> view. If such an application will be launched on a device with a >>>> square display, the views will rendered based on the setting for >>>> orientation for the app, or pars of the app. >>> >>> I must just have a different concept of what portrait and landscape mean >>> than you do. >>
