Btw, looking back, you are talking about making a talking compass that uses information from the sensor hardware... this is the exact kind of thing you need to run on real hardware to see how it actually works.
On Sep 3, 3:40 pm, hackbod <[EMAIL PROTECTED]> wrote: > It sounds like you will generally be in good shape, but again I will > say -- you should always always run your app on at least one piece of > real hardware before releasing it. This app will be running in a > completely different environment (interpreted dalvik code), and the > general user experience on real hardware just can not be emulated. > Things like the screen density, interaction with the touch screen, > etc. Personally I don't think running a previous J2ME version on some > other hardware counts as running the current Android version on the > corresponding hardware. > > On Sep 3, 3:07 pm, blindfold <[EMAIL PROTECTED]> wrote: > > > Good points, hackbod. My app is mostly a port from an existing J2ME > > app that runs on many physical Nokia phones, so I guess I have a fair > > idea of the performance that I may expect because my CPU intensive > > parts are actually identical on J2ME and Android. If I may believe the > > rumors that the first Android phone will have a 500+ MHz processor, I > > would be surprised if Android did worse than what I get with J2ME, > > although there are still some Android-specific performance concerns. > > One is stemming from the inability to play byte arrays in Android > > (J2ME can do that), for which the current workaround is to write these > > arrays to file (slow flash on physical phones?) and play these file > > chunks with MediaPlayer. Another Android performance concern is that > > one can currently not use the byte arrays coming from PreviewCallback > > because of image format incompatibilities, while using PictureCallback > > is probably less efficient for small images. Both concerns I feel are > > more related to Android versus J2ME than to the physical phones, and > > may be addressed through issues 398 and 739. Of course you are right > > that I can only tell how badly these two factors limit performance > > after trying a physical Android phone - or waiting for feedback from > > Android phone users. > > > Thanks > > > On Sep 3, 9:20 pm, hackbod <[EMAIL PROTECTED]> wrote: > > > > Testing across multiple devices may be needed, though our goal is > > > certainly to have much more consistency across devices than J2ME does. > > > > Testing on at least -one- device, however, should be a basic > > > expectation. If you are running on the emulator, you have no idea how > > > your app will behave on actual hardware. > > > > I don't think we are expecting developers to have their apps up on the > > > marketplace the first day devices are shipped. At least, if I was a > > > developer, I certainly wouldn't want to do that, because I would have > > > no idea how my app would actually behave on real hardware. Since > > > Android is about a third party developer ecosystem as much as it is > > > about a phone platform, we are in the somewhat tricky situation where > > > initial hardware is available at the same time for both users and > > > developers, which does imply that developers won't have their > > > applications available to users the first day devices are on sale. > > > > On Sep 3, 11:13 am, blindfold <[EMAIL PROTECTED]> wrote: > > > > > > Nobody should publish their app without first running it on actual > > > > > hardware. > > > > > I don't know what your assumptions on resource constraints are, but > > > > while developing for J2ME, I've had to deal with a wide variety of > > > > firmware issues making that no single phone is representative either > > > > any more than an emulator, and the cost of testing on the myriad of > > > > Symbian based phones is for most developers prohibitive. I do not > > > > see how Android will be much different in this respect. > > > > > The emulator had better be fairly representative, although of course > > > > I'd love to play with the first physical Android phone if I can find > > > > one. :-) > > > > > Regards > > > > > On Sep 3, 6:27 pm, hackbod <[EMAIL PROTECTED]> wrote: > > > > > > On Aug 31, 3:58 am, blindfold <[EMAIL PROTECTED]> wrote: > > > > > > > That's right. My own app includes a talking compass, but I cannot > > > > > > really test it and I may first have to wait for user reports with > > > > > > the T-Mobile G1. > > > > > > Nobody should publish their app without first running it on actual > > > > > hardware. Once phones become available, anyone who is developing > > > > > should have a phone to test and run their app on. You can't expect > > > > > the emulator to provide 100% fidelity with real hardware, and it > > > > > certainly won't give enough fidelity to be able to judge the real > > > > > experience on hardware. If you only ever run in the emulator, you > > > > > will ultimately end up with a poor application. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Android Developers" group. To post to this group, send email to android-developers@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] Announcing the new Android 0.9 SDK beta! http://android-developers.blogspot.com/2008/08/announcing-beta-release-of-android-sdk.html For more options, visit this group at http://groups.google.com/group/android-developers?hl=en -~----------~----~----~----~------~----~------~--~---
