On Dec 13, 2010, at 6:03 PM, James Cameron wrote: > On Mon, Dec 13, 2010 at 01:12:04PM -0500, Bakhtiar Mikhak wrote: >> On Dec 10, 2010, at 6:09 PM, James Cameron wrote: >> >>> On Fri, Dec 10, 2010 at 05:50:54PM -0500, Bakhtiar Mikhak wrote: >>>> If I am reading the current version of HIG correctly, it is not a >>>> requirement for developers to implement a portrait layout for their >>>> activity: >>>> >>>> Screen Rotation >>>> >>>> While in Hand-held mode, the laptops support screen rotation; by >>>> pressing a small button on the bezel of the display, the interface >>>> will rotate 90 degrees to provide a portrait layout of the >>>> currently active activity. Just as any activity can implement >>>> Hand-held mode, those which can benefit from a vertical aspect >>>> ratio may also implement this feature, and we encourage developers >>>> to take advantage of this functionality. >>> >>> Agreed, it is not a requirement for the activity to implement portrait >>> layout. When it is not implemented, part of the activity would be >>> invisible after rotation. The learner will rapidly find the activity >>> does not work well when rotated, and will avoid rotating. >> >> I was not suggesting to support a UI rotation without accounting for >> the change in the dimensions! > > I wasn't suggesting that either. I mentioned it in order to be > comprehensive. I've seen several activities or applications react in > this way to rotation.
Thank you. I appreciate your point. My point was more basic: if one has the option to disable rotations for one's activity, one can ensure that the users of one's activity don't encounter a compromised rotated layout. :-) >> Setting aside the wisdom of iPhone and Android allowing developers >> control over UI rotation, there are however practical considerations >> on XOs that lead us to want to disable UI rotation. For example, as >> one can easily verify in Record and other activities which use >> Gstreamer, xvimagesink does not rotate its contents. > > Ah, so you're saying the reason you do not wish to support rotation in > your activity is that xvimagesink does not support rotation. You are in > the same situation as the Record activity then. I was not aware of > that. Yes, whether or not we want to support rotation, we can't on XOs because of the documented bug with xvimagesink and UI rotation. >> Your comments did also make us wonder if the case that you seem to be >> concerned with can be seen anywhere on the laptop right now. And, we >> discovered the bug that you can see in [ http://imgur.com/z8s4i ] on >> the latest Sugar build on an XO-1.5. Note that the speaker icon gets >> positioned incorrectly after a UI rotation as well. > > That's not the latest Sugar build, and that's not an XO-1.5. It is an > XO-1, with OpenFirmware Q2E45, running OLPC OS os357. It's nearly the > latest OLPC OS build; os358 follows. These are development builds > leading toward 10.1.3 release. I should have provided links to pictures from XO 1.0 and 1.5 machines; the bug effected both machines. >> We have just filed tickets for these bugs. > > Where? I've just checked on http://bugs.sugarlabs.org/ and > http://dev.laptop.org/ and couldn't find them. The bugs were filed but apparently were still awaiting approval when I wrote my last message. They have both since been approved, closed, and noted as bugs that should be submitted to OLPC instead. >>>> I therefore think it is worth considering if the control over if/how >>>> one's activity takes advantage of screen rotation should be exposed >>>> through the Sugar API. >>> >>> Sounds good, please propose a design and patch. >> >> There are many people on this list who are more deeply familiar with >> the Sugar source code and would know best know how to go about doing >> this, but I am happy to work on it, and it would definitely be a good >> constructionist learning experience. In the meantime, however, as >> activity developers with limited time and resources, we were initially >> asking if this functionality is already supported and exposed through >> the API. >> >> The larger point being, sometimes Activity developers would not mind >> foregoing an opportunity to extend the platform they build on in >> exchange for access to tools that help them be more expressive and >> productive with that platform. Inviting our developers to extend the >> system when they ask whether a feature is supported in the API, >> however well intentioned and welcoming it is meant to be, can serve as >> a deterrent. I am sure that is not what we want. > > Yes, it would be a deterrent to many. The last thing Sugar needs is > changes that break things. Having been responsible for one recently, > I'm being very cautious now. I'm not sure that it has helped, my > patches still don't get in. I really doubt that I alone could achieve > the API you desire. > > By the way, a request for an API for an activity to prevent rotate > was logged in 2007 in ticket #2791. http://dev.laptop.org/ticket/2791 > > See also > > http://dev.laptop.org/ticket/6477 (record-activity suddenly breaks if > screen is rotated) > > http://dev.laptop.org/ticket/5797 (Record doesn't display image if the > screen is rotated) > > http://dev.laptop.org/ticket/4112 (Need option to rotate recorded > video.) > Thank you for these links, Cameron. I appreciate you taking the time to consider and offer feedback on our questions. > -- > James Cameron > http://quozl.linux.org.au/ > _______________________________________________ > Devel mailing list > [email protected] > http://lists.laptop.org/listinfo/devel _______________________________________________ Devel mailing list [email protected] http://lists.laptop.org/listinfo/devel
