Basic HTML5/offline app questions

[kepardue]

I'm increasingly interested in using HTML5 to convert some of my company's products into mobile-optimized versions with offline support.

It seems that all of the offline apps would need to have all of the site logic rewritten in JavaScript so that it can be cached offline, is this the case? I'm assuming that there's no practical way to also cache, for instance, the PHP that drives a page for offline use? Does this mean that we can expect to see a whole slew of client-side-only web apps in the near-to-medium-term future of the web?

With the offline database access, how difficult is it to get the offline version to sync with, say, MySQL once the user actually does have a connection again? There seems like there may be some potential issues in that regard to make sure that the client and the cloud are on the same page. Is there some kind of library that will check for online/offline status, and ensure that the two databases are in sync?

Can the manifest cache files be dynamic? One of the things we serve up is video for training, but I'd like to have some utility where a user can select the videos that he/she would like to view and provide them with a progress bar so that they can see when it's safe to leave the connected area.

[zcorpan]

You don't need to make the app completely client side to make it work offline. You could just make the pages available offline by putting them in the cache manifest, save the changes locally in localStorage or web sql database/indexeddb while the user is offline and offer to upload the changes to the server when the user goes online again.

For the dynamic cache thing you could build up a separate 'app' that you load in an iframe and that lists the videos of interest in its cache. You should be able to use the app cache API to detect when the videos are downloaded.