Firefox 1.1 Boasts New Features 479
Distro Jockey writes "The Fedora Core Blog gives a review of the features we can expect from Firefox 1.1. Many uses have been running the latest trunk builds and seeing dramatic improvements in page rendering, managing many tabs quickly, and the much-anticipated fix for the /. layout bug. From the article: 'One major new feature in Firefox 1.1 is the "Sanitize" feature. This enables secure browsing with much more ease. Select the "Sanitize" option in the preferences and Firefox will scrub your profile of sensitive information (which you select in the preferences).'"
What I'm curious about (Score:5, Interesting)
(2) Does it finally start to reverse the recent trend for firefox to become a huge RAM hog, or does it continue this trend?
Does the status line work properly now? (Score:5, Interesting)
Re:What I'm curious about (Score:5, Interesting)
Another "hope they fix this" post. (Score:5, Interesting)
Copy? (Score:4, Interesting)
Sounds something like the "Private Browsing" feature in Safari [apple.com].
erm (Score:2, Interesting)
How About That Memory Leak, Fixed? (Score:5, Interesting)
Firefox also boasts remote code execution. (Score:2, Interesting)
Re:google maps (Score:5, Interesting)
Imitation is the sincerest form of flattery (Score:2, Interesting)
Sounds like the "delete all private data" feature that Opera has had for several years.
Going down! Since this is Slashdot, I'll be modded flamebait for making an honest observation.
/. Rendering (Score:3, Interesting)
A few setbacks, UI wise (Score:5, Interesting)
The new preference dialog sucks. I suspect it's design is an attempt to match what OS X users expect, since Firefox devs have this (IMHO) crazy notion that the product should look as identical as possible across OSes.
The whole thing looks much more cluttered, and it has the same bugs that the UI did in pre 1.0 where the text was rendered inside of windows all the time (Like in the toolbar customization pallete, or in the current prefs). Which makes me worry that actually it's an XUL problem. If text placement is a thing that's hard to get right in XUL, it makes me worry about it as a platform.
However, performance did increase noticably for me, and the sanitize feature could be handy. I don't offhand find it much more useful that the "Clear All" button under privacy now. But it is nicely customizable, and not loosing my login cookies is kinda nice . . .
Re:What I'm curious about (Score:3, Interesting)
Mine is presently using 229MB. Of course, my X server is presently using 303MB, which, together is more than the amount of physical RAM that I have. Does Firefox map the video RAM into its address space?
Re:Another "hope they fix this" post. (Score:5, Interesting)
Thing about FireFox I don't like... (Score:4, Interesting)
1.) If I hit the middle mouse button and use auto-scrolling for something like this slashdot page, Firefox will use 30 to 40 percent CPU. And I wouldn't classify my system as slow(Athlon64 3200+ w/512Mb of RAM). Hopefully the can do something about this.
NOTE: Prior to making this post, I observed that IE holds at around 7 percent for the same action.
2.) Unexpected browser closing in v1.01 and above that wasn't present in the pre-v1.0 releases, such as when I'm holding down several keys or typing something in the browser and then switch to another page with the mouse, causing the browser to close (or crash, though I don't get an error message).
Make Firefox Look Like Maxthon? (Score:2, Interesting)
URL history sort in address bar (Score:4, Interesting)
Just like commercial software (Score:1, Interesting)
Re:back/forward (Score:5, Interesting)
YES! Finally!
Instantaneous back/forward (with mouse-rocker) navigation is one of the major killer features that has kept me using Opera as my main browser for years now. And if the tab switching and general snappiness of Firefox v1.1 has also improved to Opera's level, as some attest, then I can ditch Opera for good...
Yep. That's it. I can live without the rest of the kitchen sink.
MOD PARENT UP (and another suggestion) (Score:3, Interesting)
Re:time for a new icon? (Score:3, Interesting)
Re:What I'm curious about (Score:5, Interesting)
google search (Score:2, Interesting)
Re:What I'm curious about (Score:4, Interesting)
You have that complete bass ackwards. Memory usage shows how much physical memory (something which is usually somewhat limited) is currently allocated to the process, while VM Size shows the virtual memory (something which is practically limitless).
If a process starts up, allocates 100MB, and then never touches it, the VM Size will be significantly larger than the real memory usage, and in the real world this makes a big difference - having some seldom-used space in a paging file set aside for a task is a lot less relevant than having a block of physical memory set aside. If, on the other hand, a process allocates 100MB and then perpetually scans through it looking for Waldo, it won't be paged out and it'll consume real physical memory.
Of course memory usage can include shared memory blocks, but overall it is the best indicator of the real, practical memory usage of an application. No one cares how many new statements exist in the code - they care how much finite physical memory is practically used by the app.
Uh-oh (Score:5, Interesting)
Maybe I'll see if I can send some info to the developer... It's such a useful extension. But I've never seen FF fly like this.
x64 version too? (Score:3, Interesting)
Re:What I'm curious about (Score:3, Interesting)
You are just making that up. Look at the commit charge on the performance tab, which corresponds to the memory graph. Commit charge is the sum of all processes VM Size, not memory usage. VM Size is the amount of memory the process has allocated in memory and swap combined. Mem Usage is like RSS (resident set size) under top/ps, which is the amount of physical RAM being used, which is meaningless because of the fact that the amount of RAM being used is NOT the amount of memory the process has allocated, because it depends greatly on how the OS manages memory behind the scenes.
Even the Task Manager help file is pretty clear here:
Memory Usage: In Task Manager, the current working set of a process, in kilobytes. The current working set is the number of pages currently resident in memory.
Virtual Memory: In Task Manager, the amount of virtual memory, or address space, committed to a process.
Again, the amount of memory allocated (committed), is the number anybody cares about. Whether or not the OS has decided to swap some of the memory to the paging file or not is completely useless when all you care about is the amount of memory the program has gobbled up.
Normally, one might think that the RSS/Mem Usage should never be higher than the VM Size, because a process should never have more RAM being used than it has total memory being allocated, and I believe under Linux this is true, but under Windows, Task Manager will frequently show the Mem Usage as being higher simply because it hasn't subtracted freed memory yet, because the OS hasn't used those pages for anything else yet.
Re:What I'm curious about (Score:2, Interesting)
Astoundingly you start off with this, and then basically repeat exactly what I said.
However your conclusion, that VM memory is more important, is absurd - if an app has large swaths of memory that it basically doesn't touch, and thus can be paged out, that reduces the real memory load of the application - basically it's free memory management for apps (which a lot of apps rely upon). However if Windows deems that the memory shouldn't be swapped out, keeping the memory usage high, because a process keeps touching the pages, then that matters. That really bogarts real live physical memory.
You see, perhaps this might be the reason that those crazy Microsoft folks put that crazy mem usage on the task manager, and left vm memory as an extra option. Or maybe they just didn't consult with you.
Re:Some good and some bad (Score:3, Interesting)
While this may at first glance seem like a good idea, the browser absolutely must respect no-cache headers.
A trivial example is if you go to an online bookshop say, and you order a book. If you hit the 'back' button, you may get very confused if you start to see out of date information 'huh? It says that I haven't ordered this book yet, but I did'.
Then extend this further to more critical areas that web applications are getting used for these days. Bringing up medical information, for one. Say you order a drug through a web interface, and then hit back to go to the summary screen of all the meds the patient is currently taking, to print it out for your ward rounds. In this case showing the incorrect cached data is very very bad indeed.
how about moz (Score:1, Interesting)
some don't like FF's layout/key codes
A REAL download manage (Score:2, Interesting)
Re:Uh-oh (Score:3, Interesting)
If the latter ends up being true, perhaps it would be worth creating a separate profile that contains the Web Developer plugin. When you say you "uninstalled" it, do you mean you actually totally uninstalled it, or did you just disable it?
Re:Uh-oh (Score:3, Interesting)
I totally uninstalled it; I didn't want to make my test messy. I would imagine that disabling would give the same results though.
I've posted [chrispederick.com] on the Web Developer extension forum.
Re:What I'm curious about (Score:3, Interesting)
Another example of why people should be looking at VM Size instead of Mem Usage is that many people run a program, open something then close it, and repeat a few times, and expect that each matching close operation should bring the Mem Usage back to the value it was before the open operation, otherwise they get all fussy and claim the program has a memory leak. The VM Size will typically settle after a few iterations, while the Mem Usage will not. If the VM Size grows after each iteration and can grow continuously, you might just have a memory leak, but if the VM Size remains relatively constant and the Mem Usage increases after each iteration, it means nothing! Just minimize the application and all of a sudden the Mem Usage is back to normal.