Optimizing nglayout.initialpaint.delay is a frequent subject of discussion among many Mozilla enthusiasts. Some of them claim that changing its value to 0 makes "a huge difference". I couldn't verify this personally, but I don't rule out the possibility that this pref does make a significant difference for others, as it's greatly dependant on connection speed and other factors. In addition, David Hyatt has recently described a new algorithm that he implemented in Safari to further optimize the initial paint delay. More details here: http://weblogs.mozillazine.org/hyatt/archives/2004_05.html#005496 Prog.
> that changing its value to 0 makes "a huge difference" This is intriguing as it would mean that the whole gymnastics/coding of the paint suppression is not useful for those users.
Paragraph 1 and paragraph 2 of comment 0 are two totally different bugs. Since the summary of this bug indicates it's about the first paragraph, *** This bug has been marked as a duplicate of 82777 ***
I guess you're using (inappropriately) the term "paint delay" in the second paragraph. What Dave Hyatt described is *not* paint delay and shouldn't be called that. If you want to file a bug entitled "Optimize when and how to start displaying content for a loading page" with the second paragraph of comment 0, feel free. Reusing the current bug will lead to lots of users whining about the paintdelay pref in what could otherwise be a useful bug.
Re: "Optimize when and how to start displaying content for a loading page" That would be bug 51202. The bug that prompted me to file bug 82777 was bug 72138. Following a few links from there led me to that bug (bug 51202), which seems to be much closer to what the reporter intends. Currently, there is only one pref (content.notify.interval -- see bug 72138) to decide when the available content is flushed for display (a.k.a. incremental reflow). Maybe having another pref (say content.notify.initial) could allow the initial display to be treated differently, while waiting for the idealistic, but more complex, bug 51202.