I use firefox for regular and explorer for downloads and stuff.

I use IE for the pages where the programmers can't program for generic browsing.  There are fewer of them than there used to be, but occasionally, I have to open IE to get scripts to run correctly.  I use Firefox for downloading - I like the way it handles the status.  I don't mind having to save to disk and then open the program to actually install.  In fact, I prefer it that way.  Gives me a backup executable that I save on my external drive.

I have NoScript, ReminderFox and AdBlock Plus add ins.  In addition to the plugins to make stuff work, I have a few others loaded for looks and such.  ReminderFox is a fully functional, easy to use pop up calendar.

I got used to the old Netscape Communicator years ago and Firefox is very similar.  I like the way settings are put on the hard drive in mostly plain text.  Makes troubleshooting and reloading very easy.

I am not overly concerned about the memory usage, my issue is hard drive space.  When the new computer gets here tomorrow, I may load up Chrome and try it out. 

Sony.

I use IE and Firefox.  This is because sometimes IE acts like a little bitch.  I have numerous times tried to open up IE, sat there watching it...do nothing, then opened up Firefox and it will pull up my homepage while IE is still sitting there blank.

Doesn't happen all the time.  But often enough.

Isn't Google's motto "Don't be evil"?  In The Mind of The Market, by Michael Shermer, he praises Google as being an excellent example of a good corporation -- Enron being the example of an evil one.  He said the CEO from Enron took the wrong message from Dawkins' book The Selfish Gene.  After I read that I liked Google a lot more; it seems like a cool place to work too.

Perhaps some background is in order, because Chrome does seem to be the best browser for people who are NOT heavy-duty web users.

What Chrome appears to do is create an entire Windows' process for just about each and every web page that is open.  This is a big win because it seems to make the threading work better, and that can give better performance.

However, I typically have a dozen or more windows, browsers, applications, etc. open ALL AT ONCE.  In that kind of environment, Chrome's footprint grows to the point where I've seen more than 1,000MB of active memory at a time.  For example, 90KB is nothing, and 113KB for Firefox is also nothing.  But my current Firefox process is 783MB and the peak working set was 1,260MB.  The worse part of it is that at 20MB per second off the drive -- which isn't too bad for Windows on a laptop -- paging a 1,260MB working set takes 25 seconds.  If 1,260MB has to be paged out, that's another 25 seconds.

In my experience, Chrome would be about 20 to 30 percent =more=.

Something else to keep in my -- I literally wear computers out.  There are 13 keycaps on my keyboard that no longer have print on them.  Of those 13, four are showing signs that I may well wear THROUGH the keycap at some point.  The "S" key has gouges more than 1mm deep from my finger nail striking the keyboard and I will likely wear through it first.  My keyboard at my old job had a 1/8" (3 or 4mm in the Metric world) ridge that was worn into the space bar.  Other keys had very large amounts of wear -- measurable loss of material on them.

As regards being Kosher -- it isn't a plant or a fish or a bird that's not a bird of prey.  It lacks cloven hooves and doesn't chew its cud.  Based on that, it is not kosher.  However, it isn't an insect, reptile, amphibian, crustacean =or= any other forbidden animal.  So perhaps it is.  But I am =not= going to ask a Rabbi.

My bad -- that 113KB should have been 113MB in my reply.

There is a point where paging I/O ceases to be reasonable because the operating system can no longer maintain enough free space to readily page whatever processes are being paged in, without engaging in massive amounts of paging out.

In the case of Chrome, the design point (so far as I can tell ...) is to trade "more memory" for "more processing", which is always a perfectly valid design decision (writes the woman who's spent 25+ years of her life on operating system design and implementation ...), but design decisions can result in pathological cases where performance hits a wall.

To give you an idea, 20 years (or so) ago, I worked on a operating system where certain assumptions lead to a simplified algorithm with had performance that was in the order of objects cubed (n^3, O(n^3) for geeks) with a small constant in front -- it was very fast and very reasonable for small data sets.  Over time, these assumptions proved to be invalid and I redesigned the same code about 7 or 8 years ago so that it had performance that was closer to the logarithm of the number of objects, but was more computationally complex and required more memory.  Initial benchmarks showed a 10,000 times speed-up on the data sets which were in use at the time.

Chrome is fast only so long as the user is viewing a small number of different websites.  Which I don't do.

 Been messing around with waterfox lately, it's firefox for 64-bit machines.