Qupzilla is dead slow on scrolling some web pages.

[pfalcon]

Example: https://github.com/pfalcon/micropython/blob/master/py/vm.c . That specific page almost impossible to scroll (it takes few seconds after clocking scrollbar to actual scroll action), and CPU usage jumps to 100%.

This happens on 1.6.6 installed from PPA on Ubuntu 14.04.

[pfalcon]

Extra info: this issue is not specific to Qupzilla, but affects other QtWebKit browthers too, e.g. Rekonq. But it doesn't affect other WebKit browsers, e.g. no issues with Chromium 36 (from Ubuntu package).

[JHooverman]

You already gave the answer to this issue: this issue is not specific to Qupzilla, but affects other QtWebKit browthers too, e.g. Rekonq

Maybe the performance will increase after switching to QtWebEngine. You could try to compile and run QupZilla's experimental QtWebEngine branch. Qt-5.4.0 is required for this.

[pfalcon]

Yeah, but please don't write off this matter - after all Rekonq is just a random niche browser which will die soon like Arora did before, and Midori seems to do now, whereas "QupZilla's main aim is to be a very fast and very stable QtWebKit browser available to everyone." ;-). And currently QupZilla is not very fast and actually slower than bloat like Firefox and Chrome, so people who'd love to drop bloat, actually can't, because QupZilla don't offer benefits (and thus risks to die like many previous attempts).

Writing "lightweight browser suitable for everyone" is very ambitious goal, and surely, WebKit hacking should be included to achieve it.

Thanks for hint about new branch, added to my queue of things to try.

[pejakm]

Writing "lightweight browser suitable for everyone" is very ambitious goal, and surely, WebKit hacking should be included to achieve it.

Feel free to try the hacking. :) Issue here is the slow rendering of tables. You can also just report it to the qtwebkit tracker. In any case, nothing that we can (directly) do about it in QupZilla's code.

[pfalcon]

I'm trying to do the hacking, but don't feel much encouraged by closure of tickets w/o any help of consideration: #1450.

Then closure of this ticket gives even less encouragement. Issues with qtwebkit make your browser suck, so it's your problem, not qtwebkit's (that's of course if you're writing the browser, not a weekend hack to die soon). And there're things you can do - first of all, acknowledge the fact that issues with upstream libraries affect users' experience with your browser, then research the issue (which may involve reporting bug upstream - somehow you're better suited at them as random newbie), then give easy to find and follow recommendations to users (like minimum version X.Y required for non-lousy experience), then build yourself against that version for all platform (heck, you must be doing something like that for windows already, no?), and give easy way for users to do the same.

[pfalcon]

(Just as a background, it's 4th or 5th alternative browser I'm trying to user over last 3 years, the end result being that I still stuck with Firefox and Chromium, which I hate very much. Thinking what I'm myself doing wrong, I figure I never tried to submit bugs and feedback to all those aroras and midoris. So, I decided to try it with qupzilla - until I give up on it like with all the previous stuff. At least I will have answer to myself why I hate Firefox and still use it. (The answer would be because even the best of alternatives are worse in multiple areas, and their developers aren't really set to resolve it)).

[The-Compiler]

I noticed the same in qutebrowser a while ago and opened an upstream bug: QTBUG-40559.

[pejakm]

@The-Compiler Thanks, we appreciate your effort.