Cool Ideas Fibre ISP - Feedback

Status
Not open for further replies.
John Tempus said:
It is rational specially in this example you did not test wifi/lan to the same test server, how about that ?

Not saying you dont have a legitimate issue but when comparing the two you should atleast compare the same test server.
Click to expand...

See above, I made an ethernet connection to MWEB, Cape Town.
 
View attachment 480988
Cool Ideas 100/100 Octotel over Wifi, to Brussels, Orange.

I don't get anywhere the speeds Zoidberg gets (I should get over 60% of local speed), whether I go LAN or Wifi, the difference LAN to Wifi is not the deciding factor. It has to be something else. Maybe it's the CI supplied router, or Octotel, or who knows what. The nature of the internet, maybe. :)
 
Last edited:
Mirai said:
Cool Ideas 100/100 Octotel over Wifi to Cape Town, MWEB.
View attachment 480974

Cool Ideas 100/100 Octotel over Ethernet to Randburg, Cool Ideas
View attachment 480976

I just made these posts to show that blaming 5GHz AC wifi in same room as the router for speed differences is not rational.
Click to expand...

The latency on your ethernet test looks completely out of whack.

This may have been during today's time when we had issues in Joburg, as the 143ms latency would imply.

Would appreciate a re-test. And sorry to say, that cabled vs wifi test is *very* rational.
 
Mirai said:
View attachment 480988
Cool Ideas 100/100 Octotel over Wifi, to Brussels, Orange.

I don't get anywhere the speeds Zoidberg gets (I should get over 60% of local speed), whether I go LAN or Wifi, the difference LAN to Wifi is not the deciding factor. It has to be something else. Maybe it's the CI supplied router, or Octotel, or who knows what. The nature of the internet, maybe. :)
Click to expand...

Just remember that my line is twice as fast and that means the effect of contention on the lines will be different. Hence the main reason I decided to go for the 200/200 option.
 
Mirai said:
See above, I made an ethernet connection to MWEB, Cape Town.
Click to expand...

Ok I think you missed my point.

If you want to compare wifi vs ethernet then you need to test to the same test site. Your example was wifi tested to capetown and ethernet tested to randburg which would not be a valid test.

So assuming you still have issues, care to run wifi and ethernet test seperately to any of the same test sites to get an indication of which one is more stable than the other. If you have a really rock solid wifi connection you would essentially get same results as with ethernet but the point everyone is trying to make is that if you solely test over wifi then you are more likely to get inconsistent results vs testing over ethernet under stable conditions.

ps. I suspect there is still ongoing issues between capetown and johburg because my test results on 200/200 to london is 190/195 and to randburg 40/40 , luckily I couldnt care less about servers in randburg right now. :)
 
Last edited:
John Tempus said:
Ok I think you missed my point.

If you want to compare wifi vs ethernet then you need to test to the same test site. Your example was wifi tested to capetown and ethernet tested to randburg which would not be a valid test.
Click to expand...

Thanks for the advice John but please don't say someone misses the point when you missed the point. The point was that the variability between wifi and LAN is less than between local and remote location. That was really the point.

So assuming you still have issues, care to run wifi and ethernet test seperately to any of the same test sites to get an indication of which one is more stable than the other. If you have a really rock solid wifi connection you would essentially get same results as with ethernet but the point everyone is trying to make is that if you solely test over wifi then you are more likely to get inconsistent results vs testing over ethernet under stable conditions.
Click to expand...

The point is that CONSISTENLY I get SLOWER overseas server connections whether on WIFI or LAN vs local locations.

This is not just today and yesterday but for a while now what I've been seeing from downloading files and so on. Point was really to dispel those who say oh it's because of Wifi, and it's not because of that because wifi performance issues on my setup are less than the margin of difference between local and remote connections.

Even on LAN local vs foreign server is faster on LAN.

Now I was prepared to live with that until Zoidberg came along with his Octotel Coolideas account which seems to have a very similar performance for local and remote servers.

You kind of missed the point here and you suggested I missed the point and it's kind of rude to just assume someone misses the point, if you know what I mean. In the future I'd appreciate it if you would keep your points about the issue at hand and not whether you think someone missed your point or not. :)
 
Last edited:
DrJohnZoidberg said:
Just remember that my line is twice as fast and that means the effect of contention on the lines will be different. Hence the main reason I decided to go for the 200/200 option.
Click to expand...

So I can even get 90+Mb/sec local, but I get up to say 30Mbit/sec overseas out of 100Mbit/sec, and you get 90-130Mb out of 200mb/sec, so that implies you get a better contention ratio right?

Maybe I get 1/3 foreign server speed and you get 1/2-2/3 of it? And so guys on 300mbit lines get like 4/5 of the foreign speed then? Is that what you mean or do you mean I should expect 50-65% of my top speed the way you get?

Also as I originally said in response to you, local is maybe nice, but foreign sites are not so hot on 100mbit. It's really 100mbit for local connections. Maybe these things should be advertised as such, "get typically 100% local speed, 33% overseas speed", which was really my original response.

I don't really mind, I can get much faster speeds in Europe for a fraction of the price, just when I'm here, but I expect it's 3rd world internet.
 
TheRoDent said:
The latency on your ethernet test looks completely out of whack.

This may have been during today's time when we had issues in Joburg, as the 143ms latency would imply.

Would appreciate a re-test. And sorry to say, that cabled vs wifi test is *very* rational.
Click to expand...

It is not rational if the difference in margins between wifi and lan are less than the margins between national and international servers.

Also I provided ethernet connections and as you see they are very similar to the wifi ones.

In a world where tablets and wifi are the main ways of connecting, ISPs actually have to get off their butts IMO and not blame wifi traffic. That's so 1990s. These days most laptops and tablets don't come with ethernet ports. That's just how the world is and if ISPS want to service IT geeks who just sit with desktop towers connected with ethernet cable they will have to drop 90% of their clients, in fact their light clients and just have to keep the heavy downloaders. Never mind they also supply the routers usually too.

Maybe this needs a Carte Blanche expose or some Cape Talk expose. It seems to be an industry wide thing and totally out of whack with the modern world, IMO.
 
Last edited:
Mirai said:
Thanks for the advice John but please don't say someone misses the point when you missed the point. The point was that the variability between wifi and LAN is less than between local and remote location. That was really the point.



The point is that CONSISTENLY I get SLOWER overseas server connections whether on WIFI or LAN vs local locations.

This is not just today and yesterday but for a while now what I've been seeing from downloading files and so on. Point was really to dispel those who say oh it's because of Wifi, and it's not because of that because wifi performance issues on my setup are less than the margin of difference between local and remote connections.

Even on LAN local vs foreign server is faster on LAN.

Now I was prepared to live with that until Zoidberg came along with his Octotel Coolideas account which seems to have a very similar performance for local and remote servers.

You kind of missed the point here and you suggested I missed the point and it's kind of rude to just assume someone misses the point, if you know what I mean. In the future I'd appreciate it if you would keep your points about the issue at hand and not whether you think someone missed your point or not. :)
Click to expand...

Sorry to reiterate but you did miss my point (yes the point I made specificly and not the point you tried to make) . The example you posted were wifi to capetown and ethernet to randburg. If you still dont understand why I am saying that is not a fair comparison and not an apples with apples comparison I cant help you further. As I stated you might and probably do have a real issue but it would help as I tried to point out if you actually show result comparisons between your wifi and ethernet to the exact same test server.

That aside, today coolideas had some wonky performance between capetown and jhb which explained that high latency to randburg.

The point about wifi vs ethernet is not that you cant get stable wifi connectivity with full speed. The point is simply that if you do have wifi performance issues then it "might" be a wifi issue that you can isolate if you dont have the same problems over ethernet. Its merely a troubleshooting approach to figure out the problem.

Can you just make it simple for all of us to follow and post your wifi and ethernet test to the same capetown server, same jhb server , same london server and same USA server then perhaps we can collectively try and figure this out.
 
Last edited:
In other news, PBCool is there still ongoing issues.

My torrents, no matter what I try is absolutely DEAD. This is the first time ever on coolideas that I see this behaviour and I hope its due to ongoing network fixes and not something to expect going into the future. Have thousands of seeds as before and it starts at a few KB and just slowly drop down to 0KB permanently as if its being shaped down to nothing.

The part that worries me the moment I connect to my encrypted VPN, BAM.. not exactly full speed like before today but about 20% of max speed and atleast not absolutely 0.

Would love a response/explaination to what is going on here or would we just have to use VPN's from now ? shrug.
 
Last edited:
John Tempus said:
In other news, PBCool is there still ongoing issues.

My torrents, no matter what I try is absolutely DEAD. This is the first time ever on coolideas that I see this behaviour and I hope its due to ongoing network fixes and not something to expect going into the future. Have thousands of seeds as before and it starts at a few KB and just slowly drop down to 0KB permanently as if its being shaped down to nothing.

The part that worries me the moment I connect to my encrypted VPN, BAM.. not exactly full speed like before today but about 20% of max speed and atleast not absolutely 0.

Would love a response/explaination to what is going on here or would we just have to use VPN's from now ? shrug.
Click to expand...
John are you in CT or JHB?
 
John Tempus said:
Sorry to reiterate but you did miss my point (yes the point I made specificly and not the point you tried to make) . The example you posted were wifi to capetown and ethernet to randburg. If you still dont understand why I am saying that is not a fair comparison and not an apples with apples comparison I cant help you further. As I stated you might and probably do have a real issue but it would help as I tried to point out if you actually show result comparisons between your wifi and ethernet to the exact same test server.

That aside, today coolideas had some wonky performance between capetown and jhb which explained that high latency to randburg.

The point about wifi vs ethernet is not that you cant get stable wifi connectivity with full speed. The point is simply that if you do have wifi performance issues then it "might" be a wifi issue that you can isolate if you dont have the same problems over ethernet. Its merely a troubleshooting approach to figure out the problem.

Can you just make it simple for all of us to follow and post your wifi and ethernet test to the same capetown server, same jhb server , same london server and same USA server then perhaps we can collectively try and figure this out.
Click to expand...

I can see you are trying to assist and I am grateful for this, but I still don't think I have missed anything. My wifi is pretty consistent and the major performance issues are really ISP/fiber provider related and for most troubleshooting I think Wifi is good enough. (If my Wifi can do 80Mbit/sec, it means a 12Mbit/sec foreign connection is not a wifi issue). To make things further consistent I have provided here some links to LAN only connections:

http://beta.speedtest.net/result/6833164612.png
http://beta.speedtest.net/result/6833169096.png
http://beta.speedtest.net/result/6833178565.png
http://beta.speedtest.net/result/6833173961.png
http://beta.speedtest.net/result/6833176302.png
http://beta.speedtest.net/result/6833183994.png

I'm happy not to troubleshoot wifi issues, as this goes beyond the scope of discussion, but rather this sort of performance between ethernet Windows PC direct to router which is direct to ONT. Do you think these speeds above are acceptable for a 100/100 connection?

I think Zoidberg hinted that the 100/100 plan is a potato plan and one should upgrade to an elite 200/200 to enjoy adequate contention ratios to enjoy true 100Mbit international performance.

Thanks for your help.
 
Mirai said:
I can see you are trying to assist and I am grateful for this, but I still don't think I have missed anything. My wifi is pretty consistent and the major performance issues are really ISP/fiber provider related and for most troubleshooting I think Wifi is good enough. (If my Wifi can do 80Mbit/sec, it means a 12Mbit/sec foreign connection is not a wifi issue). To make things further consistent I have provided here some links to LAN only connections:

http://beta.speedtest.net/result/6833164612.png
http://beta.speedtest.net/result/6833169096.png
http://beta.speedtest.net/result/6833178565.png
http://beta.speedtest.net/result/6833173961.png
http://beta.speedtest.net/result/6833176302.png
http://beta.speedtest.net/result/6833183994.png

I'm happy not to troubleshoot wifi issues, as this goes beyond the scope of discussion, but rather this sort of performance between ethernet Windows PC direct to router which is direct to ONT. Do you think these speeds above are acceptable for a 100/100 connection?

I think Zoidberg hinted that the 100/100 plan is a potato plan and one should upgrade to an elite 200/200 to enjoy adequate contention ratios to enjoy true 100Mbit international performance.

Thanks for your help.
Click to expand...

Glancing over your speedtest results I agree those results are not acceptable. I can only speak of my 200/200 with coolideas and vumatel and I believe you stated you are on octotel so I would not be the best person to compare with. But for sure you should not get those results at worse I get about 40% of my 200/200 to some obscure international server and on average I get 70%+ to vast majority of international hosts.

ps. I could not check for sure but zoidberg might not be on octotel himself ? I started off with a 50/50 then 100/100 and then 200/200 on vumatell and coolideas and I did not see any big difference to my average speed for the package all the way up to 200/200 so contention should not be the problem just cause your on a cheaper 100/100 package. Something definitely does not seem right and I am pondering if its octotel related perhaps.

pss. Today is a horrible time on coolideas to do any reliable speedtest unfortunately. I am getting potato results myself for the first time in the 3 months ive been on coolideas so id take most of those speedtest today with a bag of salt until I can confirm my results are back to normal if that matters at all.
 
Last edited:
Mirai said:
It is not rational if the difference in margins between wifi and lan are less than the margins between national and international servers.

Also I provided ethernet connections and as you see they are very similar to the wifi ones.

In a world where tablets and wifi are the main ways of connecting, ISPs actually have to get off their butts IMO and not blame wifi traffic. That's so 1990s. These days most laptops and tablets don't come with ethernet ports. That's just how the world is and if ISPS want to service IT geeks who just sit with desktop towers connected with ethernet cable they will have to drop 90% of their clients, in fact their light clients and just have to keep the heavy downloaders. Never mind they also supply the routers usually too.

Maybe this needs a Carte Blanche expose or some Cape Talk expose. It seems to be an industry wide thing and totally out of whack with the modern world, IMO.
Click to expand...
In a perfect world where wireless worked perfectly this would make sense, but it's not that simple.

Each household is different so there is no static wireless solution that will work every time, as the ISP, for us to manage such a thing would be unfeasible. The protocol is dynamic and so would be the solution.

You can imagine in this country when many people would pay for 4Mbps and only yet 2Mbps. Now when you don't get full line speed over wireless this is the ISPs issue?

90% of speed issues are wireless related, if wireless was so feasible and perfect, it would be wireless to the home, not fibre to the home.

Use wireless for mobile devices sure but cable up fixed ones like TVs etc.
No one complains when they have to cable up their DSTV system for a stable stream.
 
Mirai said:
USA NYC Speedtest and LA Speedtest servers.
http://beta.speedtest.net/result/6833232146.png
http://beta.speedtest.net/result/6833236975.png
Click to expand...

Before today I would say you are in your own boat but I am seeing some rather strange results today even over http results (aside from my torrents being absolutely dead as reported few posts up)

Here is test to USA host that I frequently test just as example and getting some of the strangest upload speed. Now I wish I knew whats going on today with coolideas but for now I can only hope this is temporary and gets resolved, zero access to torrents is the most concerning for me right now.

6833246676.png
 
PBCool said:
In a perfect world where wireless worked perfectly this would make sense, but it's not that simple.

Each household is different so there is no static wireless solution that will work every time, as the ISP, for us to manage such a thing would be unfeasible. The protocol is dynamic and so would be the solution.

You can imagine in this country when many people would pay for 4Mbps and only yet 2Mbps. Now when you don't get full line speed over wireless this is the ISPs issue?

90% of speed issues are wireless related, if wireless was so feasible and perfect, it would be wireless to the home, not fibre to the home.

Use wireless for mobile devices sure but cable up fixed ones like TVs etc.
No one complains when they have to cable up their DSTV system for a stable stream.
Click to expand...

We will have to agree to disagree.

Anyhow setting that aside, I have just provided pure LAN readings. So as per the comment from user Zoidberg and the reason why he or she went with 200/200 do you think that explains my international speeds? Is my 100/100 plan a potato plan then and is only the 200/200 the true 100mbit internet plan or do you think there are other issues here? Thanks.
 
John Tempus said:
Before today I would say you are in your own boat but I am seeing some rather strange results today even over http results (aside from my torrents being absolutely dead as reported few posts up)

Here is test to USA host that I frequently test just as example and getting some of the strangest upload speed. Now I wish I knew whats going on today with coolideas but for now I can only hope this is temporary and gets resolved, zero access to torrents is the most concerning for me right now.

View attachment 481066
Click to expand...

Wow those are very good speeds. You must be on 200/200. The upload sucks though.

So maybe it's really a local Octotel issue... as you and Zoidberg get very good speed performance. Or Maybe the 100/100 is just a junk oversubscribed potato plan. But after midnight?
 
Mirai said:
Wow those are very good speeds. You must be on 200/200. The upload sucks though.

So maybe it's really a local Octotel issue... as you and Zoidberg get very good speed performance. Or Maybe the 100/100 is just a junk oversubscribed potato plan. But after midnight?
Click to expand...

These are horrible speeds which is why im bringing up a point that today something is wrong even for me. I never had a day on this host as example below 190/190 so just making it known that im also seeing some wonky results.
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X