Oh, i am so stupid. Ok, i correct myself once again and for the last time. After this, i go to bed, can sleep, and will continue with reading the original text again and think about that and how to put that together, before i write any mail again.
First of all, i noticed, i forgot the [i] after the .cache, as i listed the datatypes. Then i laughed internally about myself coz i´ve gotten it all wrong. After i bound the functions to applicationCache, i wanted to return the cache entries as an object to a variable and operate on the entries in the container After sending the last mail i thought i could move applicationCache.cache back to applicationCache. I speak here about my last message. applicationCache is of course the container of the entries[], i believe, for being sure i have to read the original again and that is what i will, afterwards. The functions of course have to be bound to the container of the entries, but that will be applicationCache, of course. This is what i assign to a variable and then i use the functions in the container for the entries (word from the spec should be used here). I am sorry, that i´ve put it out and moved it to the right and sent it to you. I am not shure, that filesystem and appcache would conflict. Thinking about it twice i couldn´t even use that together, both implementations have not much together. If i get the master file with cache or with network there is still nothing blocking my file writer or file reader. Replacing the master seems not possible to me, that´s all. I think i can go to bed now, have written enough about the topic, will read the documentation from scratch, think about it and mention it later again, before or after you go into last call and of course long before it becomes a standard. You should change this before. I am closing my argumentation now, until you write something. But i hope you´ll disturb the whatwg a few times again with, that they´ll enhance the application cache with the possibility to cache content managers and giving the chance of expanding or shrinking the cache by a few more api calls. Good night Edward