Hm, With
FireFox 9 it's improved (from the previous beta) on the overlap detection, but the horizontal scroll is way out of line, in fact I have to disable that to get it to do the margins and overlap at least a bit OK. When leaving the vertical scrolls on the default of 10 pages, it kind of gets confused. Setting it to the actual 3 is perfect ofcourse, but 4 or 5 it can handle. The margins are still troublesome though, all but the bottom margin is way off, like a top margin of 19 where 86 is needed, and a left and right margin of 2 where 9 is perfect.
After correcting the margins the Guess Overlap is perfect though.
Internet Explorer 8:
(I still refuse to upgrade to crap-o-rama 9 
)Same issues with selecting for Start Capture as Stoic had:
Ctrl-Click is apparently caught by IE to select text. Solution (by mouser) is to use the gear button, but that's kinda out of reach with a maximized window
Selecting the
IE8 internal browser area, and doing a scrolled capture, disabling horizontal scroll though (same issues as with FireFox), and leaving vertical scrolling at the default of 10, works just fine. It grabs 3 pages as expected (DC frontpage). Guess Margins did a fine job. Guess Overlap merged it all together just fine.
Doing a longer page capture, I had (could have been expected) to increase the vertical scrolls to 20 to fetch it all.
Guess Margins and Guess Overlap was almost right, mostly off-by-one, so some textlines where a bit compressed.
When setting the vertical scrolls to the expected number of pages (17 in my case) both guesses where exactly right! (And was that unexpected?)
Capturing
IE8's outside Windows frame, it requires the method to be changed to 'Simulate Keypresses (PgUp/PgDn)' to get the scrolling to work, but the Guess Margins is off by a few centimeters (instead of pixels) on the top and bottom margins, and the left and right margins by about 6 pixels, but after correcting that and setting the near correct number of 'pages' beforehand, the Guess Overlap is perfect. The method of selecting the inner page only is way better.
It might be worthwile to 'pre-detect' the actual number of pages that are in the scrollable object, and suggesting that automatically
Scroll Capturing a (wide) readme file from
NotePad++ was rather un-eventfull, it captured the page in the first try, and with correct margins/overlap both maximized and non-maximized. So not much to say there......
(Hm, a bit longer story then I thought
)
Recent Posts
but I'm jealous 
