SG Broadband Community

Hi all,

I'm was doing a little research on RWIN values the past couple of days and I 'm starting to see a pattern in the size of values used with different Operating Systems and would like some input.

I'm finding that Windows XP users are getting better results with low values while 9x or ME users seem to be having better results with the higher values.

Please respond with the OS you're running and if the RWIN you're using is a low or a high value. You don't have to post the value.

Of particular interest are those of you who have upgraded from 9x or ME to XP and had to change the values.

TIA for responding.

Windows Xp 29200
Windows ME 372300

I hate to deviate from your developing theory but I'm on XP and I use 642400.

I have just completed an 18 megabyte download off of the AOL ftp server. I had an intial speed spike up to 1.73 MB/s and then it continued to fall. When my download finished, my speed was a sustained 577 KB/s

I guess my one-way cable is good for something - no cap.

Win2K - 256960 and 522680
98SE - 256960 and 522680

What can I tell you. They both work well. Hope I didn't mess up your test.

Win98se, RWIN 32767 1.5mb DSL PPPoE. I get 1300kbs

Windows XP Rwin 513920

Tried smaller and larger, this seems like the sweet spot.

ADSL 1.5/128, do to problems with isp I never know what my speed will be

Win XP 256960. Seems to work nice.

Windows 98SE - using 513920 as the rwin number. I have tried higher and lower rwin numbers but it seems that 513920 is the sweet spot for me. Plus, I'm having no packet loss, no slow speeds, no high latency that DSLReports users wants you to believe. When I did a tweak test at their site, it reported 90% retranmissions, but when I did a 20 time ping test to other sites, there's no packet loss with a high rwin number.

-TeenInternetAddict

rwin=18200

XP & ME: 64240

8000/800kbps Cable.

513920.

NYR56 you show off

Originally posted by HalfLifer
NYR56 you show off

Heh. Well RWINS dont mean as much without speeds so, why not?

I have set my Rwin higher than normal, that's all I am telling.

256960

Windows Xp pro

128480

1.7mbps from http://dfw.speakeasy.net

Windows 98, 768k/128k, 7502ft from Verizon CO.

Getting 736k/122k ( According to http://phl.speakeasy.net )

RWIN - 17520

TeenInternetAdict:

Plus, I'm having no packet loss, no slow speeds, no high latency that DSLReports users wants you to believe. When I did a tweak test at their site, it reported 90% retranmissions, but when I did a 20 time ping test to other sites, there's no packet loss with a high rwin number.

Yep, we DSLR users are out to get you. You'd better watch out.

My Specs are:

Windows Millennium Edition

640/90 ADSL w/PPPoE (Yes, I'm upgrading to 1.5/128 next month)

RWIN: 13068
MTU: 1492

My DSLR tweak test report which, by the way, reported NO PACKET LOSS and a 100% TRANSMISSION EFFICENCY RATE is available here: http://monitor.dslreports.com/tweak/blo ... a=raspppoe

If DSLR is so great go hang out in there forums would you

I do. What have you against DSLR ne ways?
I use DSLR and find it to be an excellent site. What should you care?

DSL doesn't suck. It's alot better than cable in that you do not have to share your connection with other users in your neighborhood, which can be as slow as dialup at times. DSL provides a rock solid connection which never slows down (or speeds up for that matter).

Originally posted by HenryHand1
I do. What have you against DSLR ne ways?
I use DSLR and find it to be an excellent site. What should you care?

DSL doesn't suck. It's alot better than cable in that you do not have to share your connection with other users in your neighborhood, which can be as slow as dialup at times. DSL provides a rock solid connection which never slows down (or speeds up for that matter).

Here

Your download speed : 2056684 bps, or 2056 kbps.
Browsers would show : about a 251 k/sec transfer rate.
Your upload speed : 1355625 bps, or 1355 kbps.
Your connection rocks .. above the 1mbit barrier!

my tests w2k with 642400

NY:

** Speed 3275(down)/124(up) kbps **
(At least 65 times faster than a 56k modem)
Finish.

Seattle:

** Speed 3150(down)/124(up) kbps **
(At least 63 times faster than a 56k modem)
Finish.

High rwin - Windows 98

Originally posted by joecool169

Here

Well, you just seem to have had a bad experience with DSL. I get the advertised speeds constantly no matter what time of day.

TeenInternetAdict,

Perhaps you thinking retransmissions are packetloss is the problem, and not DSLR being a conspiracy. Packetloss and retrans are different, and RWINS have not shown a pattern affecting the retrans.

Dont know why all you guys are against DSLR, but I can tell why DSLR doesnt like some of you...

I don't want you to get the wrong opinion DSLR is useful in my view but I get tired of seeing it here at Speed Guide because their theories are a little different and it always causes a 4 page thread of arguement. But I do like some of the stuff on DSLR

Originally posted by joecool169
I don't want you to get the wrong opinion DSLR is useful in my view but I get tired of seeing it here at Speed Guide because their theories are a little different and it always causes a 4 page thread of arguement. But I do like some of the stuff on DSLR

I know. It just frustrates me that people think that the DSLR test is purposely trying to make people believe that they have packet loss when they don't. I think DSLR tests for packet loss in a different way than just pinging manually, which could be why it reports loss when manual pings don't, although I'm not absolutely sure.

Originally posted by HenryHand1
I do. What have you against DSLR ne ways?
I use DSLR and find it to be an excellent site. What should you care?

Because it promotes a theory that's totally different from ours and we don't start a thread per week saying how bad Speedguide is. Plus, I doubt that we would try and spread our stuff in the DSLR forums.

Tweaking theories. There are many. Speed Guide goes down a road I find to be one of the best.

Opinions vary and that's okay.

Originally posted by Eugene4Pres


Because it promotes a theory that's totally different from ours and we don't start a thread per week saying how bad Speedguide is. Plus, I doubt that we would try and spread our stuff in the DSLR forums.

Yep, DSLR makes threads about you guys every week. What are you talking about? I spend a ton of time at DSLR, and visit this site fairly frequently, although dont participate as much. I havent seen threads on speedguide over at DSLR for a while, and whenever there is one, it isnt very often. Same here, there arent all that many threads about DSLR.

Just because it has different view doesnt mean its horrible. Do you discriminate against other religions? Their beliefs are different from yours, but Ill bet you accept that perfectly fine.

Considering the thousands of participants at DSLR, there will always be some that annoy you, or are immature. Dont think Speedguide doesnt have that either though, both sites are public forums.

Originally posted by NYR 56
Dont think Speedguide doesnt have that either though, both sites are public forums.

It’s like a family (any website you visit). A dysfunctional family. Some days I love it here and some days I want to smash my computer into the wall. Regardless of which site you visit… someone is going to hit your buttons just like I’m sure I hit other poster’s buttons here at SG.

It might be a dysfunctional family… but it’s “MY” dysfunctional family.

578160 win2k

5000/500 Shaw cable

Before this thread goes into Speedguide vs DSLreports, enough.

TeenInternetAddict;
I think you are mistaken. 90% retransmissions from DSLR?

If it says 90% from their test, that means 90% efficiency, not 90% packet loss. For your statement to be true, you'd had to see 10%.

You might see high latency because the mainframe is in New York.

I have a 29 ping to DSLreports and a 52 ping to Speedguide. Inherited distance latency is fun.

We have nothing against DSLR, it's and excellent site.


As far as RWIN, in order to really test the difference, you'd need to try a high latency transfer from a distant site... That's when high RWIN values really make a difference.