InformAction Forums

So we'll get short term wins (theoretically) as they "improve" things (to the point where they were in Gecko 1.9, or Tom might even say earlier ).
Just hope they don't loose out in the big picture.
(Kind of sounds like, wholly crap, this is going to take a LOT, & I need a vacation.)

Then there is that thing that's bugged me for quite a while.
With (Windows) 64-bit OS's being quite pervasive for some time now ... where's the beef!
I've got to run a 32-bit SeaMonkey on my (just recently) 64-bit OS.
(At least I have picked up some leeway going to 64. Used to crash when SeaMonkey ate ~1.6 GB of RAM, now I'm good at least up to 1.9 GB.)
Other then IE-64, don't know that I have any 64-bit apps.
(Just changed my UNIX-like ones from ftp://ftp.delorie.com/pub/djgpp/current/v2gnu/ to http://gnuwin32.sourceforge.net/ cause djgpp won't work in an x64 OS.)

therube wrote:So we'll get short term wins (theoretically) as they "improve" things (to the point where they were in Gecko 1.9, or Tom might even say earlier ).

LOL! ... you know me well.

(At least I have picked up some leeway going to 64. Used to crash when SeaMonkey ate ~1.6 GB of RAM, now I'm good at least up to 1.9 GB.)

What the ... ? My Firefox on 32-bit Win XP uses about 50-60 MB RAM. With many tabs open, I can just barely get it into the triple-digit range -- *sometimes*.

I realize that SM is an entire suite, but still? -- seems like there's room for economy.

FWIW, the puter I'm writing this on has 1.5 GB RAM, and is very fast @ that amount. (Came OEM with 512 MB). Rarely goes above 400 MB *total* RAM usage. (dropped the pagefile -- didn't need it. Much faster.)
Guess I could never even think of trying SM.

I realize that SM is an entire suite, but still? -- seems like there's room for economy.

Actually, yesterday, it was with FF 8.0.1 where it hit 1.9 GB. (As memory was still being eaten - with no real input from me, at that point, I closed the browser.) So I figured if FF could run at 1.9 GB, SeaMonkey too should be able to hit that - at least.

There isn't going to be much difference between FF & the SeaMonkey "Suite" as far as memory usage is concerned.
(Or even file download size, for that matter. SeaMonkey clocks in at 18,083,259 & FF at 14,752,872. Been a long time since FF was "svelte".)

FWIW, I generally start out around the 700 MB mark & there is only one place to go from there. 12 GB RAM on one computer, 2 GB on the other. Of late GC & CC have been really painful, to the point where I reverted down one version on my slower computer. The 12 GB RAM isn't really much use (note the note above about 64-bit OS & ...).

> (dropped the pagefile -- didn't need it. Much faster.)

Heh.
I'll tell you about page files. My Windows 7 defaults to a pagefile.sys = 12 GB & a hibernat.sys = 9 GB!
(And now that I've sufficiently sized my Windows partition, I don't care. Otherwise, I'd resize them.)

http://www.pcqanda.com/dc/dcboard.php?a ... _id=537712

> Guess I could never even think of trying SM.

Why?
Though you'd likely like the 2.0 series better then current, currently at least.

therube wrote:

I realize that SM is an entire suite, but still? -- seems like there's room for economy.

Actually, yesterday, it was with FF 8.0.1 where it hit 1.9 GB. (As memory was still being eaten - with no real input from me, at that point, I closed the browser.) So I figured if FF could run at 1.9 GB, SeaMonkey too should be able to hit that - at least.

I don't remember seeing much of a jump in F8, but I'll recheck it. I can't fathom the browser using GB's of memory. Of course, I don't keep 50 tabs open, subscribe to RS feeds, headline feeds, horse feeds, don't have a local email client, don't use torrents or d/l entire movies, ... I just sit here in my cave.

> (dropped the pagefile -- didn't need it. Much faster.)

Heh.
I'll tell you about page files. My Windows 7 defaults to a pagefile.sys = 12 GB & a hibernat.sys = 9 GB!
(And now that I've sufficiently sized my Windows partition, I don't care. Otherwise, I'd resize them.)

http://www.pcqanda.com/dc/dcboard.php?a ... _id=537712

The linked topic seemed mostly to be about hiberfil, which was disabled on mine on the advice of OEM-authorized service. Didn't use hibernate much anyway; only standby.
Default Windows pagefile.sys is 1-1.5x physical RAM. I read that it takes 40,000 times as long for CPU to fetch data from pagefile as from RAM. Makes sense, when you think of what's involved. (lookup in table of pagefile entries, HD head search for it, HD read it into RAM, vs CPU lookup directly into RAM, and load....)

> Guess I could never even think of trying SM.

Why?

Because of the memory usage you quoted. As said, this laptop has 1.5 GB physical RAM. You're hitting 1.9 GB usage, and then there are the minor usages, like the operating system, firewall, AV, ... I'd need a huge pagefile.

Though you'd likely like the 2.0 series better then current, currently at least.

I do indeed prefer F3 to F8, and use 3 for all browsing. Only use F8 for doing support here, to make sure both current versions are covered in any issue or question.

I still have F2 on the machine, but that's a dirty little secret.
(seriously, it's been useful in diagnostics, where one user said that the issue applied to newer Fx, but not to 3.x and below)

Speaking of browser bloat, you might find this thread interesting. There were a couple of, uh, "strongly dissenting" opinions to mine, until a professional involved in ultra-high-security programming for Gov and corporate customers backed me up, and trashed the "new, improved" feature completely, on fundamental principles of security.

I tested Fx 8.0 for memory usage. About 80 MB "idling". 5 windows with a total of 7 tabs boosted that to about 110 MB. Playing a YT video went to almost 115 MB.
Total RAM commitment for the entire system never reached 500 MB.

The CPU usage spiked, in loading and playing the YT vid as quickly as possible, but this is still less than 1/10 of the RAM you're reporting for SM.
I can't fathom 1.9 GB RAM, but I know you do a lot more d/l, etc. Do other SM users report similar usage?

You might have a memory leak somewhere? -- in the computer, i mean.

In general Fx just keeps getting more bloated and bloated, I can only imagine that SM would have even more given the suite factor. Although not 100% accurate to the millisecond and mega/gigabytes, my general sense has been that Mozilla items (specifically Fx in my case) has been doubling in load time, reducing in half in performance, and doubling in resource usage with just about every major release. One of the reasons I have been itching to dump it and have frankly stayed with it for NS, no other reason whatsoever to be honest. They lost my loyalty when their claim that they can be extended by tons of great addons was followed by (the performance degradations are due to the extensibility via addons, not our fault). So on the one hand you boast that you can be extended and are awesome because of that and why you don't have certain common sense functions built in (until recently when they ripped off some from Chrome and other developers) and yet you blame any problem on that very boasted functionality? Excuse the french, but WTF?

So in terms of NoScript developement that means that Giorgio has more than enough time for porting and extending NSA to replace NoScript 2?

dhouwn wrote:So in terms of NoScript developement that means that Giorgio has more than enough time for porting and extending NSA to replace NoScript 2?

Effectively, which means my nightmare of having to deal with Fx for a longer time just crushes my spirit