6 ms·
https://github.com/gorhill/uBlock/wiki/Firefox-version:-benchmarking-memory-footprint-of-the-early-preview https://github.com/gorhill/uBlock/wiki/Firefox-versio
by msl09 12y ago
https://github.com/gorhill/uBlock/wiki/Firefox-version:-benchmarking-memory-footprint-of-the-early-preview https://github.com/gorhill/uBlock/wiki/Firefox-version:-benc...
- zobzu 12y agoWah Chromium uses way more ram than Firefox. 425M (firefox+ublock) vs 1.04G (chromium+ublock) in all cases memory is reduced :)
- gorhill 12y agoI did warn on that page: > You can't compare directly the figures between the browsers Still, yes, Chromium uses more memory. A good part of this is because per-process tab. Once Firefox get the same per-process tab architecture, it will be easier to compare both browsers together.
- wooger 12y agoYep, Chrome / Chromium are now the #1 reason to bump my ram to 16GB. While using a laptop for coding / VMs etc. I still find my browser eating more ram most of the time - it simply doesn't handle paging out gracefully, and is horrific for my uses with < 4GB ram on any OS.
- zobzu 12y agowell, firefox seems set to use a process for UI and a process for all tabs (such as in nightly right now), the consumption is still far lower.
- exprL 12y agoI'd appreciate details of CPU consumption as well. Minute amounts of memory are of no concern for many people, but everyone prefers snappier browsing, every time.
- gorhill 12y agoHere: https://github.com/gorhill/uBlock#cpu https://github.com/gorhill/uBlock#cpu The details of the benchmark (if you want to reproduce) are in the spreadsheet linked in the caption.
- exprL 12y agoI was aware of the contents of the front page, having skimmed it again today (and read more carefully before, although with little interest for Chromium extensions). I'm only interested in performance characteristics in Firefox, but am unlikely to test this myself before it arrives in AMO.
- karon 12y agoHow is that even possible? Firefox with blocker uses less RAM than without? Does it block elements from issuing requests?
- wtallis 12y agoOne would expect any ad blocker that actually blocks ads to reduce memory consumption significantly. Ad blockers that merely hide ads exist and have succeeded in earning reputations as ad blockers despite only doing half the job.
- scholia 12y agoTrue, but when I uninstalled Adblock Edge, I gained almost a gigabyte of memory in Firefox. (I'm now using Ghostery to block ads. It isn't as effective but I'd rather reduce the memory footprint.)
- nnethercote 12y agoAlas, it's not that simple. See these bugs about AdBlock Plus's effects on Firefox's memory usage: https://bugzilla.mozilla.org/show_bug.cgi?id=988266 https://bugzilla.mozilla.org/show_bug.cgi?id=988266 https://bugzilla.mozilla.org/show_bug.cgi?id=1001426 https://bugzilla.mozilla.org/show_bug.cgi?id=1001426
- wtallis 12y agoThat second bug is lacking in technical details, but the first one is clearly about the memory consumption of the CSS rules used to hide elements. If you only use rules that match URIs instead and not element ids, ABP is extremely efficient. It's worth remembering that ABP only implemented the element-hiding feature so that things like inline text ads could be blocked, but now the popular rule lists are using a large number of broad rules to hide all kinds of elements, many of which could probably have been effectively blocked with URI-matching rules. Additionally, it's very unfair to the ABP devs to be criticizing the performance or memory usage of ABP when the problem is really the memory usage of EasyList. Using a more restrained and targeted rule list makes the problems go away.