I've read every forum post, and the Core Concepts article on the web many times over (which needs to be much better written, and I am working on that
and I still don't really get it. I would much appreciate answers to these questions, and please assume a persistent tce/opt/home:
1. When I "Install" an extension, my understanding is that the tcz gets downloaded to my persistent tce, then it is loaded completely into ram. So when I run the program there is zero disk access at that point because it is already completely loaded into ram. Is this correct?
2. "Installed" extensions are all completely loaded into ram on boot (barring special boot codes and such) - so if I "installed" firefox, chrome and opera, I would basically be having 3 entire browsers running simultaneously in ram even though I have not yet opened any of them. Is this correct?
3. "OnDemand" only downloads the tcz, tells TC of its existence and then stops there without loading itself to ram until it is invoked. Is this correct?
4. Following, once an "OnDemand" extension is actually run, it is then completely loaded into ram and for all practical purposes it becomes almost exactly like it would be had I "instaled" it. Is this correct?
5. Some stuff is weird or awkward to make "OnDemand" like 915resolution or audio drivers, however bigger, higher-level programs like vlc or chrome make perfect candidates for loading "OnDemand". Correct? In fact, why would I ever not want high-level programs to be "OnDemand"?
6. I still haven't gotten anything "OnDemand" to actually work. In some posts it mentions running the ondemand command but that doesn't seem to do anything for me. And I never see menu entries for things "OnDemand". Why can't "OnDemand" be completely invisible in that in terms of UI everything is identical, however behind the scenes TC knows to first load it to ram?
That's all for now, thanks for your time!
Baz