In order to make sure that partially loaded pages get in to SH and unresolved pages don't get in to SH, browser has to implement nsIStreamListener and hook in to OnDataAvailable(). This can be done better after the new webshell is in place and browser component has undergone similar changes

Move to M16 for now ...

hey radha this i done already right? (ie, this bug can be closed?)

This feature was needed so that webshell/browser can differentiate between a valid url that was partially loaded/stopped half thro' from a invalid URL. Once it knows the difference, it can decide better if a url is to go in to the SH or not. To answer your question in simple terms, if invalid urls don't go in to SH and partially loaded/stopped urls go in to SH then this feature is not required. I haven't looked in to whether this is happening. If you know what happens in this scenario, please update me. I'll close this accordingly. Thanks,

working with the 2000042611 builds: It would seem that invalid url's are NOT being added to SH and partially loaded/stopped url's are being added. I tried loading several bogus url's ranging from fully qualified (http://, www, .com, etc.) to unqualified and after they failed tried to reaccess them via the SH (Back/Forward/Go/Reload/etc.) but was unable. For every page that did exist I WAS able to access it via SH as long as I stopped it AFTER it BEGAN to load. That is to say if I tried a page that takes a moment to resolve, the Stop button becomes active before the page begins to load. If you stop at this stage, nothing is added to SH.

I think this is not necessary anymore because docshell which now controls what goes in and out of SH takes care of invalid /partially loaded urls.

VERIFIED