Slashdot Mirror
Mopping Up Mozilla Memory Leaks
mouseman writes "Geodesic Systems, a maker of memory management/debugging software, has a live demo of their Linux product running on the Mozilla nightly builds. It's pretty damn slick -- it detects memory leaks and can show where in the code the leaked memory was allocated and actually recover (GC) the leaked memory. The Mozilla reports actually look pretty good, which jibes with my own impressions of how much it has improved -- see for yourself."
Wish they'd come out with a port for Win32. I've had more programs (usually file sharing services) eat up both memory and processor cycles to the point that everything came to a standstill.
As an alternative to this one and purify, I've been using valgrind lately (http://developer.kde.org/~sewardj/). It's defintely worth a look for those of you who don't have access to these commercial tools.
Of course it doesn't do much in the way of garbage collection (well, anything) but it's still a great tool.
Cheers Koz
Normal Netscape tanks for Linux, it freezes too often, the process will remain after exitting the program, requiring a kill -9 to start a new instance..
'cause I think I'm of the age that most of my memory is leaking all over the floor....
Now where did I put that mop and bucket?
The race isn't always to the swift... but that's the way to bet!
I must confess that since Purify is not available for Linux, we are always looking for interesting pieces of software to do this job.
:
Now the only question I have are
1) can it be used as a debugger too ?
2) how does it compare to other systems like electricfence ?
-- Martial MICHEL
Ever since .9.7, Mozilla has been running so damn fast and stable that it replaced my old Browser (Nutscrape). I'm not surprised at all. IT renders pages faster than IE does. I'm quite happy with it.
JoeLinux
Will this directly affect and improve Mozilla on non-Linux platforms? Would it have a use flagging the leaks in the Linux code and then making the code corrections on the OS X platform?
Does this product actually integrate with the debugger. For example, can it break on a line of code before it uses a dangling pointer?
I've been using NuMega products within MSVC for a number of years. I used Purify before that. I haven't used the latest version from NuMega, but I've always had a lot of success with BoundsChecker (memory profiler, error detection, etc), True Time (performance profiler) and True Coverage (code usage profiler).
I wouldn't even know where to start looking for open source equivalents of these products. Can anybody give me any pointers? Preferably for Windows, but also Linux. How does this product compare with the likes of those from NuMega and Rational?
only a closed minded fool who does not use it or thinks that MS software is better than anything would even say that...especialy considering where Mozy has gone in 4 years and what has taken IE to get in 7....I would put both on par and infact, Mozy has many usfulfeatures that IE does not.
I am the Alpha and the Omega-3
CCMalloc also detects memory leaks and shows you where they took place. Even uses the GPL! It might not be as easy to use (you link your program to it during compile instead of running the it on a completed binary), and doesnt have a web interface, but its functional, and has been out for years.
./a.out' just like the 'gsinject -d ./a.out' of this product.
e cts/ccm alloc/
I bet someone could write a ld_preload kind of thing that calls ccmalloc, and you could run it on the completed binary as well, so you could run 'ccmalloc
Here is ccmalloc's page.
http://www.inf.ethz.ch/personal/biere/proj
The Segway has a built in web browser too? Cool!
for Opera... only you *nix geeks didn't hear the music. The 'saur rules, but still isn't king of the web... MS is and for good reason.
Next product they need is a bandwindth leak detector for their web server....the /. effect seems to have run it's course already.
(+1 Funny) only if I laugh out loud.
If you're writing C++ programs, you should be using a garbage collector, such as Geodesics, or the Boehm collector available freely on the net. I've heard various objections to collection, mostly boiling down to speed issues. At least with the Boehm collector, you can choose to manually manage some memory yourself. In that case, it should be treated as an optimization problem. Write your program with the collector, then profile it. Anywhere you can pinpoint the collector as a major slowdown, handle the memory yourself.
If tits were wings it'd be flying around.
Live Demo = Slashdoted before Slashdot even posts the story.
Actually, ccmalloc has a mode like purify to automatically link itself in during the compile process. Not the same thing as preload, but it's just as easy. If you're using, say gcc, change your compiler from "gcc" to "ccmalloc gcc" and ccmalloc will do the rest.
:)
ccmalloc is a great utility to have if you can't get your hands on purify
F0 07 C7 C8
It's really improved a lot. 0.97 was painfully slow on OS X, but 0.99 is actually faster than MSIE. There are still a few UI glitches - for example, command-H doesn't work - but this is the first time I've been able to switch to Mozilla and not have to switch back within an hour. I'm still going strong after three days. Things are definitely looking up for Mozilla.
that I seem to be missing?
After reading the Salon article about how improved and fast it was now, I decided to give it a whirl on my NT desktop at work.
I uninstalled it after about 15 minutes. It was just slow. Sluggish in loading pages, slow in creating new windows, everything. Not only was it slower than IE 5.5 but what surprised me was that it was slower than *Netscape* 4.7, which is what I primarily use. It's a shame because it had some nice options.
What am I missing? Is it not meant to run on NT? Is it debugging code? Would it run better on XP? I'd love to give it a chance but apparently, I got some other Mozilla browser instead of the one everyone here is raving about.
Once they've recovered from their thorough slashdotting I'll have to go check 'em out. Efence alone has saved me multiple weeks of debugging time, allowing me to track down memory leaks in our horrid application instantly. This translates into thousands of dollars of development time that can be put to more productive use. I'd estimate that efence has saved my group over 12 thousand dollars in debugging time this year. Now if only I could convince the boss to channel even a small portion of that back into the open source community...
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
This is a great example of the Open Source community giving back to private companies. Geodesic tests an OSS web-browser, and Slashdot stress-tests their web server.
You can hear me screaming down the street.
It is a little better under IE, but I prefer to use Moz.
"It is a greater offense to steal men's labor, than their clothes"
noooooo its a FEATURE :D
----- Whats wrong with this picture? http://www.revoh.org:1234/whatswrong
Tabbed browsing is far better than anything IE has.
and, guess what, your pressiouse Mem foot print, well I looked at my memory used before I IERadicated my IE, after ward, I gained a ton of memory with out IE on my system.
IEs memory foot print is hiden from you, that is why it takes windows so long to load, because IE is in memory at boot.
I am the Alpha and the Omega-3
Somebody loose these?
:D
Now if you where running OS X, it would just shutdown that application *shudder* nasty way of managing a leak. I sure hope it saves the data first
What about BoundsChecker instrumented builds?
----- Whats wrong with this picture? http://www.revoh.org:1234/whatswrong
Well you're right and wrong. Yes Mozilla has caught up and in some ways surpassed IE. IE does not portend to be the end-all-be-all application or swiss knife killer app that Moz wants to be. But to suggest that moz made it where it is today in 4 years is not entirely accurate because the knowledge to get it where it is today was built while creating Mosaic up to 4.72. Not to mention work that other people provided that may have been going on prior to 4 years ago. It's an apples to oranges comparison because both companies didn't start development on their product simultaneously and therefore can't be compared in "coding speed" to each other. Additionally neither browser is fully backwards compatible so it could be said that each implementation is in essence a rewrite and if that were the case and criteria for judging then MS would be ahead.
oh yeh... that will definitely help when you've got a memory leak the size of ontario from a program that stopped executing an hour ago.
or, you've got a pointer that went off its bounds and started writing all over its instructions, and the program crahes 1000 lines after the viiolation....
sometimes, printf, jsut isnt enough.
... hi bingo
as far as I can tell, Mozy displays all pages that NS 1.x could. how is that not backwards compatable?
Or else you're doing something wrong.
I found it extreamely helpful in tracing memory leaks.
It's GPL, it is mature, it is multi-platform.
Very feature reach and can be easly integrated in your existing projects with includes and automake macros.
Over the last 6 months, every time I download a nightly or milestone, I've been recording the results of the following test.
Download the installer and run it. Moz will launch automatically after install. Kill Mozilla and restart it. Bring up the task manager to see how much memory is used.
Here's the results I've had on my W2k box:
2001 10/01 build 03 22,540k
2001 10/15 build 06 21,876k
2001 10/18 build 03 21,692k
2001 10/23 build 09 21,984k
2001 11/16 build 03 21,952k
2001 11/29 build 03 21,656k
2001 12/06 build 03 19,868k
2001 12/10 build 03 19,812k
2002 01/07 build 03 18,124k
2002 01/14 build 03 19,064k
2002 01/17 build 04 19,244k
2002 02/26 build 03 18,608k
2002 03/06 build 04 18,220k
2002 03/11 build 04 17,704k (build 0.9.9)
That's a decrease of almost 5 megs in memory usage over six months - all the while they've been adding features.
Significantly, they've added venkman, tabbed browsing, the dom inspector and favicon support.
Also significant - I don't know why I bother to get the talkback builds anymore - I haven't submitted a crash in 3 months. And I run it all day, every day.
Textbooks and Open Educational Resources
As far as I can remember, NVIDIA have a neat testing system that tests the nightly builds of their drivers to see if performance has gone up, down or the same. If I remember rightly the developers who submit patches that slow things down get a ribbing the next day :)
Computer automated testing is our friend.
' Ore stabit fortis a fine placet ore stat '
- found on a park bench
Their demo is live no longer.
Long life the slashdot effect.
Any spoon would be too big.
I never thought I'd use tabbed browsing, but the more I use it the more I love it. It's wonderful to load up links in background tabs while I finish reading the current page.
Georgia summer camps
fellow Slashdot user and hacker extraordinaire.
Thank you Bruce, for all you've done for Free Software.
I tried mentioning this in a thread, but it was too quickly drown out. There is a free product that does ld_preload style memory leak detection. It even does memory allocation profiling, so you can find that hidden 'for' loop that is responsible for that extra 15MB of allocated memory. The gui is done with gtk+, so it's easy to use, and will run on most any linux distro these days.
It's available at http://people.redhat.com/otaylor/memprof/
Freshmeat has an entry for it as well.
no, REAL progammers use kprintf()
HA HA...What a Retard!!!!!
Mozilla seems to be pretty efficient CPU-wise, finally, but it's still a huge memory hog (I've seen 0.9.x versions have problems on systems with 128MB of RAM!), and once it starts needing swap space, that CPU efficiency goes out the window.
adéu,
Mateu
"And we're happy here, but we live in fear, we've seen a lot of temples crumble..." - Concrete Blonde
Stupid job ads, weird spam, occasional insight at
no...YAAFI...Idiot!!!!!
Slashdot Effect..
Hehe....
Even a few builds ago I considered Mozilla unacceptable for every day use. There were just too many sites that Mozilla couldn't handle. But now that I've tried the latest release, I think it is finally going to be my primary browser. Two questions though....is there a way to make my scroll wheel move down the page faster, and I really wish shift clicking on a link opened it in a new window....
...and all I get is an empty page
It is nice to see a technology demo of this tool, but mozilla already has a tool for finding memory leaks (among other things) called spaceTrace:
which you can find out about here
Hyperbole is the worst thing ever.
The problem is that on Linux systems, it's hard to get 'free' fonts that look good. Most fonts on Linux are bitmapped, and thus don't scale well. The following upgrade will help all applications work better, especially if you alias your 'helvetica' font to 'arial' and your 'times' font to 'times new roman'. Go to Microsoft's "web core fonts" site and download the .exe files. Extract with a program like cabextract (they're just cab files with a self-extracting wrapper). Put them in your X server's fonts directory. Restart the X server (and the fonts server, if you use Redhat). Set up Moz to use the new fonts. Marvel at the beautiful scaling of TrueType fonts. All is better ;).
That's it. I'm no longer part of Team Sanity.
and here i thought that Netscape skipped a number and went straight to 6.0...
</sarcasm>
btw, they are trying to avoid the slashdotting they are getting by changing the frames to all be about:blank. seems sloppy to me, especially since they are still having you download three javascript files at the top of the page. oh well...
here's a link to their 'news' page on the mozilla stuff (not the demo):
http://www.geodesic.com/news/2002/gc_demo
It was a joke you dumbass!
Years ago in my employment, that is.
If there is interest, I might recreate it for the Public Domain. It should only take a weekend.
The principle is simple: go through all writable memory (including the stack and the registers). The idea is -- and I didn't invent this -- what looks like a pointer [to the heap] is a pointer. Those heap blocks that were not pointed to, are reported as garbage.
A special allocator package records the calling frame so you can tell where the block was allocated. Other niceties are included. The tools is implemented as function calls that you typically invoke from the debugger.
While the principle is heuristic, the practical experience (with Solaris and Interactive UNIX) is that it instantaneously finds all garbage.
Marko
So how much memory does the windows Kernel, Explorer.exe and IE6 use all together?
You cant find out how much memory IE6 uses on WindowsXP, maybe on Windows95 you can.
If you use Linux, please help development of Autopac
Edit>>Preferences>>Advanced>>Mou se wheel should have everything you need, you can manually set your #of lines per wheel turn.
LOTR: Elijah Wood is a munchkin asshat. Yes, asshat. LOL.
Mozilla requires 300mhz machine at LEAST.
You shouldnt be using it on an old as hell machine. Its slow because your CPU is slow, period. they cant make your CPU faster. I mean the problems you complain about are CPU load problems not problems with ram.
If you use Linux, please help development of Autopac
I'm gonna need people's help to add to the list below.
Here is a list of (freeware) Memory Leak Detectors and Garbage Collectors -
ccmalloc
http://www.inf.ethz.ch/personal/biere/projects/cc
Valgrind
http://developer.kde.org/~sewardj/
Boehm Collector
http://www.hpl.hp.com/personal/Hans_Boehm/gc/
Parallel Collector on Message Passing Environment
http://www.yl.is.s.u-tokyo.ac.jp/gc/dgc/
If there are more out there, would you kindly add what you have to the list, please ?
Thank you !
Muchas Gracias, Señor Edward Snowden !
It's a story about a product that's actually unrelated to Mozilla, but they put up a 'demo' of how to use it WITH mozilla, and it makes the front page of slashdot with the Mozilla logo. If I have a pr0n site but I have a chick masturbating with tux can I have it posted under "Linux News" ?
"How hard is it to manage pointers and free your memory allocations anyways?"
.NET with all their marketing strength.
(To those who didn't get it, I was being cynical)
That is the response I usually get when advising to write using high-level languages that take care of memory management. Finding enormous leakage in huge programs almost proves this response wrong. Finding leakage in almost all programs definitely proves it wrong, and you sure will find leakage in almost every large program written in a low-level language like C or C++.
The overuse of low-level languages with increasingly powerful hardware is becoming appearant to more and more people. It seems the "dark side" will be out of the low-level language hell sooner, as Micros~1 is pushing C# and
If you don't want the Linux/*nix world to stay behind, stop using C and C++ where Python and Lisp can be used. Writing in C or C++ instead of a higher-level language can be deemed as a premature optimization, as any specific part of the program can be optimized and written in C when necessary. We all know how evil premature optimization is.
If you write in C or C++ because that's simply what you know, you should not be wary of learning new languages. You should know that C++ is extremely hellish to learn, while Java is a lot simpler, and Python takes at most a day or two!
From my own personal exprience, functional languages are a bit more difficult to learn and grasp, but it seems they perfectly fit some mindsets.
Stop wasting human power on writing destructors and freeing unused memory, when it can all be done automatically.
It's definitely slower than NS4.7, but that's mostly because its rendering engine supports a lot more. (And also because the architecture is pretty bloated, but there are some benefits to that.)
You might need a faster computer. I still use NS on my laptop, but Mozilla is nice and fast on my desktop.
It happens to me with Windows 98 with IE, but W2K with IE does not exhibit the same problem.
Under Windows 95, 98, and ME, the USER.EXE and GDI.EXE resource heaps are each limited to 64 kilobytes. (Windows NT, 2000, and XP have no such limit.) This bites me in the butt on K5 too.
Will I retire or break 10K?
The Mac OS X's malloc library has built-in support for providing good leak detection. The command line tool leaks (installed with the free developer tools). Although leaks can be run on any program, it works best when you set the enviromment variable MallocStackLogging, which causes the system to provide complete stack frame information about when each piece of leaking memory was allocated.
Also of interest are the malloc_history and heap command-line tools. They tell you lots of interesting information about exactly what is on the heap and how it got there. The malloc library also detects double frees by default (making it immune to the recent zlib security vulnerability) and can also detect writing on free blocks, etc. Very nice.
Finally there is a graphical application, MallocDebug which provides similar information on memory leaks and memory use but also provides an easy to use memory-browser interface. Unlike simple leak detectors, you can explore your program's memory space and discover memory that isn't strictly "forgotten" by your program (it still has valid references to it) but should perhaps have been deleted anyway (e.g., finding memory allocated for a splash screen graphic). This last application requires that the code use a special library, but it is easy enough to make dynamically-linked programs use this library (thanks to Darwin's equivalent of LD_PRELOAD, DYLD_INSERT_LIBRARIES).
Although you can debug any code with these tools (from command-line tool to X-windows, Carbon or Cocoa app), they really shine debugging Objective-C, because Objective-C provides enough runtime information for the tools to tell you lots about the objects that are sitting there on the heap.
First things first: the new wolf rocks. Now to get back on topic...
I've had no problems with RTCW after mozilla, during a few instances of mozilla or otherwise. Looks like its fixed! Or at least reduced to where my P933 w/ 128MB RAM can handle it with no noticable strain on anything.
LOTR: Elijah Wood is a munchkin asshat. Yes, asshat. LOL.
You fuckers are able to admit that Mozilla is a memory hogging, memory leaking POS
It's not so bad on my windows machine, but on my linux machine.....jesus christ
It's an apples to oranges comparison because both companies didn't start development on their product simultaneously and therefore can't be compared in "coding speed" to each other.
Yes they did. I always thought Netscape Navigator (up to 4.x) was based on NCSA Mosaic, which became Spyglass Mosaic, which became Microsoft Internet Explorer (see IE's about box). After the Netscape fork of Mosaic (i.e. Mozilla Classic) began to show its limitations, development began on its replacement -- Gecko -- around which Mozilla.org built its next generation Internet browser.
Will I retire or break 10K?
Hey I wonder if this program can also be used to debug the memory leaks in squid proxy server...
I'd recommend trying a different browser based on Mozilla's rendering engine. For Linux, there's Galeon. I use this as my primary browser (I can't stand Mozilla - it responds slowly even on my 1GHz processor with 400 MB of RAM). K-Meleon is a similar program for Windows ("K-Meleon may be considered the unbloated Mozilla for Windows").
A more basic question to ask, however, is why something like Mozilla has memory leaks in the first place. Avoiding memory leaks in C is hard because there simply are no hooks in C to automate resource management. But C++ has constructs that make writing leak-free code and code that doesn't use bad pointers pretty easy. Since switching from C to C++ a few years ago, memory leaks and bad pointers have simply not been an issue anymore in my code.
It doesn't support many NS4isms like layers. Pretty much all browsers are back compatible with Netscape 3 however.
HAHAHAHA. No sir, YOU are the retard. I will not even validate your fucking moronic statement with a thought out response.
Have you tried out the one at http://spellchecker.mozdev.org?
the good ground has been paved over by suicidal maniacs
Netscape was the company who originally made Mosiac if i remember right, before they were called netscape.
If you use Linux, please help development of Autopac
With Purify, and presumably this Great Circle product you either just re-link the app (UNIX) or simply re-run it using the tool (Windows).
Purify++ has some preprocessor nonsense going on that takes forever to use.
I know that when I have moderator status here on slash, I can watch the system resources disappear down a block hole if I hit a story with a lot of comments and I use Moz.... It is a little better under IE, but I prefer to use Moz.
I've been tracking this particular bug for a LONG time, and couldn't get anyone at the former Netscape to do anything about it. I haven't pestered the current Netscape or Mozilla teams because I haven't actually used the current browsers, and consequently expected I'd receive a poor welcome. But it sounds like the bug IS in still in the codebase, per the above-quoted comment.
Specifically, the resource leak occurs when there are a lot of links INSIDE table cells, AND those links contain a lot of text. This apparently has a trigger threshold since if you sufficiently cut down the amount of text displayed as links, the leak goes away. Worst case, a large table with lots of long links can suck resources to zero in 30 seconds flat.
It appears that checkboxes, radio buttons, and dropdown boxes occuring inside table cells can also trigger the problem, tho I don't have this pinned down to exact cases.
Symptoms vary with the NS version and the WinVersion under it, but in general you will notice oddities like vertically overlapping text and checkboxes that go invisible. An example that affects NS3.04 is Slashdot's own "User Preferences" page, with its zillion checkboxes.
This bug is apparently in ALL versions of Netscape, but only visibly manifests in certain versions. To my certain knowledge, 3.04-16bit (running on Win16 or Win32), 3.04-32bit, and 4.64 are the worst affected, but apparently it can be triggered in ANY version if you work at it hard enough. 3.02-16bit Gold, 4.00-32bit, and 4.04-32bit are somewhat less affected than average.
People using the old Delphi forums (which were horrible for triggering the leak) griped about the symptoms constantly, regardless of NS version. Of course most had no idea what was happening, only that using Delphi crashed their computer.
So long as you haven't actually zeroed out the system, resources will usually recover by 90+% after you leave the triggering page.
The bug is ALSO definitely in IE3.0. I don't personally know about other versions other than that IE5.01 seems somewhat resistant to it (but I only use IE as a last resort, so...) But per the quoted comment, the bug is clearly not absent in recentish versions (I assume no one here is insane enough to still be using IE3 :)
Correct me if I remember wrongly, but don't IE and NS both ultimately derive from the Mosaic codebase? The implication is that this is a core bug in the foundation code.
BTW there is also a resource leak that happens if you view a page that has SWF on it, but you do NOT have the SWF plugin installed. The leak is approx. 25% per SWF box per page, and unlike the above-mentioned leak, resources do NOT recover when you leave the affected page. This affects NS 3.04-32bit and 4.04-32bit equally, haven't checked other versions. I suspect this is another manifestation of the same leak, since SWF-inside-tables also seems to be involved.
Affected sample of the SWF/tables resource leak: http://www.blkviper.com/index.html (however, NO leak IF you have the SWF plugin installed).
I'm not a coder, but I can contribute testing if desired.. I hope this information finds its way to the appropriate people (or if someone knows who I should contact, I'll do it myself), and that this bug is finally fixed!
~REZ~ #43301. Who'd fake being me anyway?
I have never been able to successfully run Boehm's GC replacement for malloc/free in my programs on both Solaris/Sparc and Linux/Intel - it crashes very soon after startup usually after spawning a thread, but can also also in a single threaded application. In spite of this, I would still like to use it. What's the secret? There are easily two dozen tweaking parameters (interior pointers, exterior, etc). I just want to run it - not have a PhD in it. Are there a pre-built GC shared libraries for Solaris and Linux somewhere?
Don't know about NT, but I have Win98 with only 128MB of memory, and the ancient process, and Mozilla 0.9.8 and 0.9.9 runs *much quicker* than the latest IE. In fact I'm using Mozilla practically all the time - except for the non-standard web pages (that is, 'viewable only with IE'). It doesn't happen that often, though. In addition, IE crashes on Win98, every once in a while, but quite regularly. It didn't happen with Mozilla yet. Not with 0.9.8 and 0.9.9.
> I've been tracking this particular bug for a LONG time, and couldn't get anyone at the former
> Netscape to do anything about it. I haven't pestered the current Netscape or Mozilla teams
> because I haven't actually used the current browsers, and consequently expected I'd receive a
> poor welcome. But it sounds like the bug IS in still in the codebase, per the above-quoted
> comment.
Given that current builds of mozilla contain virtually no code from Netscape 4.x[1] it's probably a different bug. And yes, logging a netscape 3.x or 4.x bug in bugzilla would not make you popular.
[1] about 2%, I believe.
[snip]
It would not be unlike Microsoft to introduce an undocumented API which allows apps to behave differently in swap, but that's just speculation
[snip]
I like conspiracy theories as much as anyone, but this one I think I can debunk. IE shares a lot of code with the rest of the GUI part of Windows (iexplore and explorer are intimately connected == uses many of the same DLLs), and explorer.exe has many parts that are used all over. Since these will get constant (at least regular) usage, their pages just doesn't get swapped out that much.
Another possible explanation is that Mozilla is written in C++ (it is, if memory serves my correctly). Often, OO programming splinters code into many methods with little code in each. Since most compilers place methods in order of appearance in the sourcefiles, calling only a few selected ones of these and having a relatively deep inheritance hierarchy (which further distances code parts), the actual code executed will be scattered all over the memory area. If different classes in the hierarchy are located in different cpp files (they often are for overview and simplicity), they are also linked into mostly random parts of the memory area.
This all means that the 4Kb pages holding the methods called don't get many hits, so they are prone to being swapped out by the rather aggressive VMM in Windows.
It should be entirely possible to monitor the swap activity while doing a certain task using the standard monitoring system in WinNT/2000. Doing this on a standard installation after the initial loading of the application, and counting page faults all the way through a fixed sequence of web pages, once for IE and once for Mozilla, one could get a reasonable overview of whether page faulting is really to blame. Someone who cares enough should probably go do it ;-)
Black holes are where God divided by zero
Thanks a lot for the leads !!
I will add them to the list and hope that others will do too.
Thanks again !!
Muchas Gracias, Señor Edward Snowden !