Michael Plitkins has just added APIs to nsIScrollableView to scroll the view by lines (or pages). The app can specify the definition of a line, and Steve Clark suggested that we use just-out-of-view line heights to decide how much to scroll, to ensure bringing a full line into view each time. It's a nice idea.

The App? Sounds like a job for Don's crew.

Re-assigned to matt@netscape.com and set target milestone to M7.

Just testing cc for new mail address.

Has this code made it into the tree? I believe it could be used as a fix for bugs 11751 and possibly for 16806

Re-evaluate later ...

Can you take this request for a feature. You can impletement it a million times faster than me.

I don't really like the idea that the browser would scroll by a different amount each time I press down.. I would like to be able to set how far I scroll, though. Netscape's default is right around 1 "standard-sized" lines, but I'm used to IE's default of about 3.

Doing something like this is complicated for lots of reasons. I think this code may have been enabled for a short time, and it had problems around tables. It's not clear what it should do around tables, floats, absolute positioning, etc.

Move to M16 for now ...

spam: qa contact to jrgm@netscape.com

nav triage team: Not a beta1 stopper, marking nsbeta1-

vishy

-> default owner

See also bug 200213, "Scroll step for up/down arrow keys is too small with smoothscroll on".

As far as I can tell, this is fixed. If I scroll on a page and then increase the font size and scroll again, I scroll by more. Looking at the code, this seems to have been fixed back in 1999: http://lxr.mozilla.org/seamonkey/source/layout/base/nsPresShell.cpp#3172 Is there any reason to keep this bug report open?

In case anyone values my opinion, I don't see why this bug should be kept open. As people have mentioned: not only is it complicated to implement, it's not necessarily a good user interface. Having one click scroll more or less than the very next is arguably inconsistent. If anyone really wants to do this "enhancement" anyway, go ahead and reopen the bug. :)