> As a developer you don't want to start at 0 reviews every time there is a OS update. Reviews are pretty important, and if it periodically looks like you don't

> As a developer you don't want to start at 0 reviews every time there is a OS update. Reviews are pretty important, and if it periodically looks like you don't have them it would be very bad. This is not true. Sure, if you were the ONLY developer with 0 reviews on a new OS version and everyone else had a bunch, it would look bad, but if every app had 0 reviews, you would be no different. And if you had a large user base on one version of the OS you'd likely have many reviews quickly. > You should not see any difference in the different versions of iOS. Though it might perform differently on different hardware. Certainly things are often mostly backwards compatible, but there are often changes at the API layer which break some things or add features that are available on the new version that weren't on the old one. For example, Apple added the backgrounding/multitasking mojo with iOS 4 but until an app was updated from iOS 3 and could make use of those APIs, the functionality wasn't enabled. Does it make sense for users of iOS 3 to see reviews of an app, not yet updated for iOS 4, of iOS 4 users bitching about how the app doesn't multitask right when multitasking wasn't enabled in iOS 3? > I feel as if Apple hasn't really done very much to help the developer note what platforms they are targeting in an easy way. If you don't want iPhone 3g users to use the app, they shouldn't be able to download it that that device in the first place etc. This is a decent idea, however most things DO work from one OS version to another, so barring people from using an app on their new device seems like it would often be an artificial limitation. Further, it is tough to beta test an OS if there are no apps to run on it. I think this concept of splitting the reviews would solve the problem of clutter in the current review system without barring people from doing the testing that ought to be done. Further, some kind of hybrid could be done, such as maintaining the star rating across OS versions (so people could have an accurate representation of the general reliability, etc. of a given app) without clogging the review systems of different OS versions with irrelevant data and unwanted complaining about incompatibility with the new beta.