Reply 40 of 89, by Jorpho
- Rank
- l33t++
wrote:I use 4 different browsers all opened at once:
I really don't see why exactly you would need that, but have you considered Lunascape?
http://www.lunascape.tv/
wrote:I use 4 different browsers all opened at once:
I really don't see why exactly you would need that, but have you considered Lunascape?
http://www.lunascape.tv/
wrote:I just have 4 gigs in my PC. Two years ago you could buy this amount for €15 ($20) but the prices seem to have doubled now. Meh...
A RAMdisk of 512 MB works as a charm, I assure you.
wrote:Isn't using a ramdisk for a browser by definition the same as disabling disk caching entirely? Admittedly, I've never been too find of Firefox's insistence on keeping around urlclassifier3.sqlite, which is presently consuming 56 MB of space and seems to constantly be growing – apparently it is used for nothing more than storing FF's database of phishing sites.
It's nothing of the sort. Instead of relying on the physical HDD in your computer, let them be mechanical HDDs or SSDs, you instruct your browser to use a specific volume in your computer which has been previously configured as a RAM disk. A RAM disk reserves a small portion of your system memory and makes it available as conventional disk space, with its corresponding filesystem, where you can write anything you want - but remember that any reboot or shut down will wipe the contents of the drive.
It's ideal for using on applications that write in short bursts to disk, like web browsers and its main advantage is that it is blazingly fast. Here I attach some benchmarks from a couple of years ago.
wrote:Isn't using a ramdisk for a browser by definition the same as disabling disk caching entirely?
Nope nope, browser still caches, only instead of doing it on slow, crawling devices like an HDD ...
... or an SSD ...
... on a ramdisk it caches at RAMming speed 😈 :
Admittedly, I've never been too find of Firefox's insistence on keeping around urlclassifier3.sqlite, which is presently consuming 56 MB of space and seems to constantly be growing – apparently it is used for nothing more than storing FF's database of phishing sites.
Worse still, Firefox is continuously making a ton of small writes on urlclassifier3.sqlite :
So if the Firefox Profile is in a HDD, we lose speed scratching.
If it is in a SSD, we are flogging the SSD with all those writings.
But if we put Firefox Profile in a ramdisk, we can have our cake and eat it too: Max achievable speed and no flogging for the SSD 😉 .
Let the air flow!
wrote:It's nothing of the sort. Instead of relying on the physical HDD in your computer, let them be mechanical HDDs or SSDs, you instruct your browser to use a specific volume in your computer which has been previously configured as a RAM disk. A RAM disk reserves a small portion of your system memory and makes it available as conventional disk space, with its corresponding filesystem, where you can write anything you want - but remember that any reboot or shut down will wipe the contents of the drive.
Yes, I know all that – but instead of configuring the browser to use disk space that actually exists as a portion of RAM, why not just configure the browser to stop writing to disk and/or to use more RAM directly?
Interesting but no matter how bad that it is IE is still worse and Chrome is getting bloated in some ways on its own.
On a far away planet reading your posts in the year 10,191.
wrote:... why not just configure the browser to stop writing to disk and/or to use more RAM directly?
That can be done in Linux, but as far as I know not in Windows, unless via ramdisk.
Let the air flow!
/dev/shm is effectively a ramdisk anyway.
VogonsDrivers.com | Link | News Thread
wrote:wrote:I use 4 different browsers all opened at once:
I really don't see why exactly you would need that, but have you considered Lunascape?
http://www.lunascape.tv/
Interesting, I may have to check that out.
Having extra RAM allows you to open more apps at once anyway so it's no big deal on my machine.
In fact I run a VMware Server app or two at the same time without any performance issues.
No matter where you go, there you are...
Anyone using Opera?
theres dedicated RAM 'disk' that connected to SATA interface, so it wont using your RAM for it,
and as system (BIOS/OS) concerned it just another HDD.
http://techreport.com/review/16255/acard-ans- … al-ata-ram-disk
-fffuuu
Anybody else use Waterfox?
Yep, using Waterfox. Fast, reliable and doesn't seem to get patched too frequently either.
Today I just switched to Pale Moon after finding that Opera does not run well on my machine because of how it is supposedly not GPU accelerated. Pale Moon has been running fantastically for me and does not have the bloated feel of normal Firefox.
The main drawback of using RAM disks is that information is lost after powering down.
Disable browsing cache, at least in my experience, makes continued browsing sluggish and unresponsive. Modern browsers rely in a lot of caching, not only of webpages as such. Having a large and fast caché helps a lot.
opera lets you choose disk and/or ram cache
Does anyone know how to test real-world, real-time, the performance difference between all these new browsers? And a way to put up our results, if possible to compare? Personally I can't see much difference between IE10 and the latest Firefox. So I do not know if I'm missing something, looking at various issues others face.
In other words, how do you know if your browser is "fast"? And what is a "fast browser" by definition? Can we quantify it and compare our results?
Problem is, putting up such results, will be influenced thousands of numerous configurations of pc systems, and hard to pinpoint what may be the dragging or slowing factor. But still, any practical testing will be useful than nothing.
A lot of program "speed" relies on perception...particularly, I'd say, when talking about Web browsers.
Sure...Firefox can run x or y Javascript (or, funnily enough, in my case some IE specific tests 😜) test faster than IE...or maybe Chrome does, but, sometimes, in this case, I'd say it isn't that important. Why?
First and foremost it's about responsiveness; for some people, speed at which the program can be launched and how fast or slow pages display...something and are useable is the real factor (the processing speed maybe quite similar; but if a browser displays a page while loading the images and the other doesn't, perceived speed is different, even if, in the end, to get the real, complete page it took the same time). A lot of pages rely on plugins such as Flash, that counts too.
To further complicate matters, connection speed *is* an important issue (and hence the need for cache, that seems to be *IMO* wrongly smacked a lot in this thread).
I'm not sure if people use really slow hard disks (or don't defrag them) or load really poorly optimized or "image heavy" websites, but the idea behind cache is that not all internet connections are all that fast and, in the end, the internet is, at times, a flimsy beast...cache is there to alleviate some of those problems.
There's a lot of reliance on Javascript (or some form of ECMA script) these days, and...like any program it can be written poorly or properly optimized, also, most textual web components can be "minified"...again, just like any program, you can (and should, if you really want speedy websites) have your developement or debugging source and your release version...depending on how much you "minify" it with at least unnecesary whitespace removed (this can be done at the very least with Javascripts, style sheets and the html itself).
To further complicate perception matters on speed, there's PHP, another useful tool that relies on server load and speed (that varies greatly through the day).
There's also the user interface: different people like different interfaces: to me the Chrome one is awful and the IE one is useless, yet some people love them, this factors in general program perception; then there's the so called "bloat": different people have different needs and what some may consider bloat others consider an indispensable feature, and, at least in the user's mind, if they think a program is "bloated" they perceive it as slow (even if it's not...or if speed doesn't really depend on that feature, specially if turned off).
In the end, as with so many things in life, I think the real benchmark for a webbrowser is not speed...but user preference and needs.
Hell, want a really fast browser? Use w3m 😜
cache actually matters, i found plenty sites that tell the browser to NOT caches the javascripts,
but those very same javascript files actualy repeatedly requested in different pages, in this case browsers that actualy obeying and not caching it would perform poorly than those who cache'em anyway.
Forgot to add, this mostly apparent for those with slow internet connections.
-fffuuu
wrote:The main drawback of using RAM disks is that information is lost after powering down.
Any decent ramdisk can save an image to HDD before shutdown, and load it again after booting. I use batchs to load just the PaleMoon Profile to ramdisk after boot, and save it to HDD before shutdown (much faster).
Also we can turn the drawback into a security/housekeeping advantage, by using the ramdisk as an automatic garbage disintegrator on shutdown (temp files etc.).
Let the air flow!