Firefox is too slow...

**djr33** · 11-11-2012, 09:20 PM

I've always preferred FF. But speeding it up a bit is just another reason to like it and perhaps eliminate one major complaint others have had.
(On the other hand, I do admit that FF seems to be going in weird directions at the moment, such as having a new full version number-- apparently now I'm at v15 or v16-- every couple weeks. And the more it starts to become like the other browsers and do weird things, the less I like it. But having the option to override these things is good and overall I'm not planning to switch from it.)

I did install Chrome and Opera the other day to do some testing though. So maybe that's my first step away from Firefox. Not for the moment though

**keyboard1333** · 11-11-2012, 10:14 PM

Does the workaround for firefox increase bandwidth significantly?
I'm on a fairly limited bandwidth per month...

**molendijk** · 11-11-2012, 10:53 PM

Everyone has his/her own specific reasons for preferring one browser over another, of course.
My specific reasons for preferring a given browser over another are:
- there should not be flicker (white flash) when we go from one page to the other, or when we load flash elements;
- clearing the cache (f5) shoud be easy;
- the pages of any website should load; they should not have NSL (never stops loading) problems (Opera and Chrome have a problem here, see also this).
So I choose FF, and don't trust Opera and Google Chrome.
IE9 is OK, but it's a target for hackers having bad intentions.
Arie.

**djr33** · 11-12-2012, 12:06 AM

Does the workaround for firefox increase bandwidth significantly?
I'm on a fairly limited bandwidth per month...

That's an interesting and important question. I really don't know.

I was originally suspicious of it and also the fact that it might be frowned upon by ISPs if you're overloading the system. But after looking into it, I think it's ok. I'd like someone to confirm it.

The workaround increases the number of simultaneous requests. That's certain. But it's also ambiguous. There are two possibilities:

1) The browser sends a lot of requests and hopes that one goes through quickly; it's excessively bothering the server (and ISP?) and might increase bandwidth (just for the requests? does the data come through eventually for each and then get ignored?). And it seems like a bad idea just in general-- a hack, not a solution.

2) The browser sends all requests at the same time rather than waiting to send them later; so if there are 5 images, 2 scripts, 1 stylesheet, 1 movie and 1 flash object on a page, then you end up with a total of 10 requests (including the HTML page itself). Normally, this takes the time of 10 requests (each one in sequence), but now it takes the time of one request-- whichever of those 10 is the longest, because they're all going simultaneously.

If (2) is the case, it makes sense and I like it. I think that's right. I hope it's not (1). The first link (in this thread) makes it sound like it might be (1). But in fact, from doing a bit more research I think it does end up as (2), where it's just doing parallel processing rather than serial processing, a good idea, not a bad idea.

If anyone else has some thoughts on it, I'd love to hear them though!

**traq** · 11-12-2012, 12:44 AM

I haven't tested (or even tried it out), but from reading that article, I would suspect it's #2 that's going on: not "extra" requests, just making all of the requests asynchronously.

**keyboard1333** · 11-12-2012, 02:22 AM

I just did what the tutorial suggested and there is definetely an increase in speed!!!
I'll keep an eye on the bandwidth meter though...

**djr33** · 11-12-2012, 02:41 AM

Ok, let me know if you see something different. If traq and I are right, then this will just speed things up (basically multitasking) rather than increasing the amount of transfer.

Technically, there is an increase in "bandwidth" in some sense-- the actual speed of transfer will increase.

So if you have 5 images that are each 20kb, and we assume that it takes one second to transfer 20kb, then:
--Total amount ("bandwidth" in some sense): 100kb.
--Speed/time with normal FF: 5sec
--Speed/time with faster FF: 1sec
--Transfer rate with normal FF: 20kb/s
--Transfer rate with faster FF: 100kb/s

As on most webhosts and for most ISPs, I assume "bandwidth" refers to the amount of actual transfer (eg, over one month), so that won't change anything-- it's always 100kb.

But if you have any limits on the other type, the actual speed (which is really what "broadband" refers to and there's another kind of "high bandwidth" related to that), then that might be an issue. My guess, though, is that it would simply slow things down so that the transfer is, let's say, 50kb/s, and that would take you 2sec to load everything.

But... I don't have any evidence to back this up. It just makes sense.

**keyboard1333** · 11-12-2012, 02:46 AM

But if there are no downsides to doing this, then why wouldn't firefox make it the default setting?
One of the main downsides of firefox is the speed, why not use the solution?

**traq** · 11-12-2012, 02:53 AM

I suppose something weird might happen, once in a while, if requests -particularly scripts- come back in random (i.e., fastest) order.

**djr33** · 11-12-2012, 04:10 AM

What timing!! I was just convinced by FF to update to 16.0.2. And now my "Firefox 3 theme" addon is disabled. And I'm very upset with FF. It's so bad looking now. Why can't they allow the user to choose how it looks, or better yet just not make these odd changes? I've tried a couple other options (various addons) and they're no better. Maybe I'll get used to it. There are about 70,000 people complaining now, that is the group of everyone who was using "Firefox 3 theme"... hmm...