Life in the Faster Lane?
I was out for a drive on the xda-developers highway and came across what appears to be a pretty good tweak. dearnasfamily swears that he’s found a way to speed up Internet connectivity. Though not absolutely conclusive, a number of folks on the thread tested the regedits and the majority report increased speeds.
You can make the registry changes manually or you can use the cab on page 4, post #35. Take it for a test drive and report back.
If you enjoyed this post, please consider to leave a comment or subscribe to the feed and get future articles delivered to your feed reader.
Comments
drjim I believe it’s only for the TP/Fuze. But you should check the thread.
FYI, I installed and used my Opera and it does seem faster.
drjim An honor to have the man behind blownfuze.org respond.
Yes, that is what some have argued. But, having a Ph.D. in Psychology, I know that placebos can have a very real effect on us physically. I don’t suppose that could work with technology though! But if you belive it’s faster, then it has the same psychological effect, namely, less frustration and impatience than before making the registry tweaks. That’s a benefit, I suppose. :->
BTW, I installed the cab and it didn’t do anything. So I just made the registry changes manually. Very easy to create a new string and change the value of another.
So the question remains: Is it real or is it memorex (that goes back a few years!)?
nikhil s i personally just like to use opera turbo on the 9.7beta. does the trick for me 
(except when i have to download stuff and turn off opera turbo…that’s just a pain in the butt…then i just use the newest opera mini )
The Jack of Clubs this is just allowing more tcpip connections to ie. itll work on any phone that has ie. you can do this too in opera mobile 9.5 and 9.7 too. it should be under performance in opera:config. just put 32 for everything, dont bother with that 16.
Clint I did over 50 seperate speed tests on my Fuze, uising PIE, Opera, and google’s android browser. This did absolutely nothing to speed up the connections. On average the post mod speeds were slower.
I believe people are doing one premod test, then seeing faster speeds after mod due to traffic changes.
P.S. I am using one of the latest Energy ROMS, maybe thats why mine is already fast?
Doug Simmons Torrent downloaded from http://www.Demonoid.com
Hot damn Clint that’s some serious dedication.
Keep in mind people there are a lot of other factors at play to confound your results. For starters AT&T’s network ain’t what it used to be. Once upon a time (when I had the 8525 and in the first few months of the original Tilt) I could get over 2000Kbit/s (record was 2150Kbs), or 250KByte/s, which most would qualify as broadband speeds. I could even get those speeds downloading Linux torrents on a laptop tethered (with wmwifi of course). Since the Fuze I get under 800.
Now AT&T either is chronically saturated somewhere along the routes between your phone and their main gateways or they have some kind of throttling, either sophisticated shaping or just speed limits either based on congestion at a given time or how much you’ve downloaded that month or what server ports you’re connecting to. Who knows. But with factors like that and speed tests with results that tend to fluctuate a couple hundred kbits per 1mb test — which is why it’s good, Clint, that you did this so many times. I just did three 1MB tests — full URL is http://www.dslreports.com/mspeed?t=1&s=1024&w= — and got 620, 170 and 251Kbit/s back to back.
The only hack I’d be interested in would be this: You ever heard about people hacking their cable modems to remove a bandwidth governor in order to get ridiculous speeds, throughput uncapped? Give me a regedit for that and I’ll give you a share of google.
But for those of you with some faith for this thing and think it does more than just changing some IE tweaks you could do in advancedconfig or opera:config for max connections, or if any of you just want to do some real bandwidth testing for the purposes of, say, comparing radios, here’s what you need to do:
Tether up a computer to your phone *through wifi* not usb or bluetooth, make sure nothing is running else is running on the phone, make sure the computer isn’t connected to any other hosts for whatever reason by going to start > cmd > netstat -p tcp (good way to see if you got any trojans btw), then do many tests on http://speakeasy.net/speedtest, all the different servers listed.
Better yet, because I’ve heard a number of times ISPs being accused of doing a little sneaky packet bursting on these 1MB speedtest sites and then ramping down the throughput on you after, if you have access to a shell account or some site that you can download a huge file that you know is definitely super fast (test it first on your computer through fast broadband) and not too many hops away from your phone (find that out with tracert), perhaps by downloading something like the Anddroid for the Fuze port I posted a few articles back, break out the calculator and plot out the data, get an average. Preferably do this when most people are asleep. Make sure your phone is in the exact same place during all tests, like your window or wherever you can get it the best signal.
Then flash another radio, or in this case the regedit cab, and repeat.
Other than doing the same but with, maybe, level3’s public dns servers in your phone’s advanced connection settings which are 4.2.2.1 and 4.2.2.2, that’s the best way I can come up with for you to go into a thread here or on xda to brag about your amazing results as a result of whatever it is you did to your phone, do that kind of testing to weigh in more scientifically. You might want to add those dns servers anyway to see if it helps your latency (though it won’t help throughput), and latency’s important for us and our phones.
The registry tweak on this thing only appears to modify max connections per server per server meaning if you load a site like this it will try to download more images in addition to the index file at the same time rather than grabbing eight at once and then starting another and another as it finishes one of those eight. Based on that I have no idea why it would yield better results for a one-connection speed test. And again, not much new here; you can make these changes (i think) with advancedconfig. Whether or not these changes effect how many ports your computer connects to when tethering, I don’t know.
I’d wager that there’s a downside to using this tweak on blogs for example that have a lot of text and a lot of little images to load. Your CPU takes a beating as your phone grabs all the images it can and you have to wait longer before you can start scrolling throw the given thread.
That’s it I’m done.
Doug Simmons uhh .. don’t know how that link got up top there, I apologize. For that and the length of the post.
drjim @Clint: Okay, I admit it, this was a false post!! Sorry…
@Doug: What are you referring to?
Doug Simmons first line on my post… no idea how it got there (probably an accidental ctrl V) but I do go into the max connections per server and how that might be relevant to downloading Linux torrents when tethered.. nevermind. feel free to delete that top link unless things are more laid back here than xda.
Does this also work for the Tilt 2?