Since we moved to the country 2 months ago, I am now withput high speed compuiter connection, and suffer through on a dial-up modem. (which I can begrudgingly live with btw)
My avge connection speed is 46.6 Kbps.
While surfing tonight I found this website:
They claim I can download up to 400% faster.
Any truth to this guys? I don’t mind buying it if it works, but don’t want to waste my cash on smoke and mirrors.
Logically I just can’t see how they can squeeze so much more data into my itty bitty phone line, but I thought I would pose the question to be sure.
For dialup users it’ll do exactly nothing. Your connection speed is a physical limitation, and there’s just no way around that.
For broadband users it can have some effect.
The way these download “accellerators” work is that they establish multiple connections to the same file, each connection downloading a different part of the file. In cases where the site the file is located on serves only a certain amount of bandwidth per connection, it can speed up the overall download speed by virtue of multiple connections being faster when combined than one single connection. It isn’t a whole lot faster – two concurrent connections doesn’t mean twice the speed as the server does have finite bandwidth; each connection itself would be slower than a single connection, but overall there’s a general increase.
It also has no effect on broadband users if each individual connection is able to serve data as fast as you’re able to receive; at that point multiple connections could actually be slower due to the overhead involved in maintaining multiple connections.
Any connection speed has a physical limitation; this software isn’t claiming to increase it, but rather to put it to fuller use; more peaks/plateaux and less troughs in the graph. This can be done for dialup connections.
However, I don’t expect it works nearly as well as they’re claiming except in the very specific conditions they have no doubt engineered in which to test it.
Proxy compression won’t work all that well on high-speed modem, or rather, it will, but what you gain in reduction of bytes transferred, you lose in real-time compression implemented by the modem. It is appointed unto data once to be compressed, but after this, it just ain’t worth it.
If the server sending the data experiences periodic hiccups in transmission (the troughs) then multiple connections will keep the transmission stable. If you’re downloading from a stable connection in the first place however, multiple connections wouldn’t accomplish anything. Either way – as you say below – the gains would be extremely slight for dialup users.
I can’t even think of how they could claim such speed increases even in the most carefully engineered test scenarios. If we’re speaking strictly in terms of file downloads (as opposed to general data transfer as occurs during the normal course of web surfing) then how is this possible? You could speak of compression, but this requires A) That the server employ the same compression algorhythm that the receiving end will use to decompress, and B) That the compression algorhythm is up to 400% more efficient then the compression that already exists in the modem. I don’t know what compression method they use in modems these days (I remember v.42bis used LZW or somesuch; not sure if that changed for v.90) but I’m fairly certain nothing can achieve a 400% better lossless ratio.
Dunno; back in the days when I was on dialup, I remember watching the bandwidth graph dip and dive all over the place; I’m not sure exactly where the bottleneck was with this, but one way to have solved it would be for a proxy server (with a fast connection of its own) immediately upstream from me to be performing extensive prefetching with multiple connections, so that my thin pipe could be used to maximum benefit all the time. There are probably other ways around it too, but as you say, it isn’t ever going to make dialup comfortably fast, especially with today’s increasing weight of content.
Another possibility, for typical web use: Most of the time when you’re web-surfing, you’re not downloading, just reading. If the program anticipates what links you might click on, it could start downloading those in advance, while the modem would otherwise be idle. The downside, of course, is that this puts a huge strain on the server, since everything gets downloaded, whether it’s read or not. There was some speculation recently in ATMB that this might be a contributing factor to the chronic speed problems here.
I think the speed problems here are related to the forum software’s search function. vBulletin is notorious for bogging right down to nothing when searches of large databases are performed. Another forum I belong to switched from vB to Invision last year for precisely that reason. The symptoms were identical: Speed was fine normally even under significant user load, but there would be almost no response from the server if anyone was performing a search.
As for web accellerators – there have been dozens of those programs and they all either cache relatively static data (images and the like) and/or prefetch every link on the page. The first took up a great deal of space over time (depending on what cache expiry was set to) and took a while before any significant gains were realized; the second was just a bandwidth hog – a bad idea for anyone whose monthly bandwidth is capped by their ISP. (The last is irrelevant to dialup users, but that has its own problems; prefetching made link clicking less responsive as it would take a few moments to stop prefetching to acknowledge the clicked link)
Depends on how much bandwidth Unca Cece’s webhost provides, what sort of bandwidth pipe those using web accellerators have, and where they’re located. In general, I’d be inclined to say it would take quite a few simultaneous bandwidth hogs for any significant slowdown to occur.
I still maintain however that the biggest contributor to the problem is the vB search function.
From what I’ve heard, searches are in fact the major drain on server resources here, but that does not rule out the possibility of web accelerators being a contributing factor. If they are a non-neglible problem, though, it would be due less to the bandwidth usage than to the processor usage, since all pages on this board are constructed on the fly according to scripts.
I recollect that some “download” speed enhancers also compress HTML code in a proprietary way, which is then expanded by a client program. This sort of makes sense, as HTML is pretty wordy and could be compressed significantly, except that even at full wordiness the actual HTML doesn’t take up much space compared to graphics, sounds and all the other things you’re downloading.
In other words something that could lead to huge speed increases in a very unrealistic test, but wouldn’t have a noticable difference in the real world.
The FireFox extension FasterFox will do many of the things mentioned, for free. It’s unlikely to have any noticeable effect, but why not give it a go and see? Ya never know.
Of course, the test bed under which such an increase occurred was probably A) A local server B) compressing C) plain text D) containing nothing but the letter “A” repeated 1000 times D) sent over 10 feet of RJ45 to the receiving client.
What? It’s perfectly accurate. Just because the ad is true and entirely possible doesn’t mean has to be practical or likely.
First, why buy it?
It’s got a FREE ad supported version, just download it and see if you like it!
400% faster? No you’ll never see that much of a speed difference.
Faster than the Windows downloader? Yes!
The bigger the file, the bigger a speed improvement you’ll get.
Here’s links to two other FREE download managers, Star Downloader Download Express
Be very careful about those download managers! Many free “download managers” and “web accellerators” contain adware and/or spyware, and we’re all quite familiar I’m sure with what that’s all about.
Anyone remember Gator? There are still plenty like them and worse out there. Read thoroughly through their documentation and EULA before going any further with installations. Even then you aren’t necessarily safe.
