Broadband Ratings

rpm

rpm

Admin
Staff member
Joined
Jul 22, 2003
Messages
66,806
Reaction score
5,057
Location
Johannesburg
Hi folks

Here is the link to the draft copy of the broadband ratings:
http://home.telkomsa.net/rudolph/broadband_ratings.pdf

The results in this document is accurate, but you will notice that I am not much of a journalist or graphic designer, so if you have any suggestions (grammatical, design etc.) please let me know. As the document improves I will simply replace the old one…

Your feedback is most welcome as usual.

Regards,

RPM
 
My feedback so far, impressive!!! - just looked at the layout & pictures so far - I will get to reading after a few urgent things are out of the way.

Looks v.goo<edit>d</edit> :)

Well done to RPM & Beyers.
 
Looks good. One comment though : although MyWi has not official cap, international bandwidth are definitely throttled to 25% of package after the first 500K (max) of any download. AFAIK, this is a definite on the MW128 and MW256 packages. This has been the experience of almost all the users, one exception being a reseller of the product and maybe also on the 512K version. Local is not affected at all and speeds are as advertised. IMHO, this makes MW unsuitable for international access, other than normal browsing.
 
Sweet. In general, gives a quick overview.
However, I cannot use the pdf to decide on the correct broadband provider to use if I want to play games (Or i'm too stupid to know what to look at in the presentation.).
<editdel>But, I don't suppose that that is not the purpose of the pdf.</editdel>
But, I suppose that that's not the purpose of the pdf.
 
Last edited:
Hi Stroke

Thanks for the feedback and good point! We have latency figures, but they were not comprehensive enough to valid comments about gaming. We are already testing new software to include this in the next report :D

Regards,

RPM
 
Have not read in detail yet but there is one HUGE error.

Telkom offer two ADSL line varients: 384 + 512
Telkom offer 3 ADSL account varients: 2 GB shaped, 3 GB shaped and 4 GB unshaped.

You can mix and match these at will!!!

I use home 384 + 2x 3g shaped accounts.

Please correct that else people will get a totally erroneus impression of home 384 package!
 
Last edited:
Hi RPM,

LSM latency results - I have been in touch with Dave Apelt @ TCPIQ. It is definitely a firewall issue - blocking of ICMP Echo Requests by our PCs. Unfortunately the IP addresses Dave emailed me don't appear in my firewall alerts & logs, otherwise I would just create an exception rule for those IP addresses. When he pings me from his own PC the ping does get blocked & logged.
 
Hi Harth

Thanks for the feedback. I agree that it makes a big difference and I have added a note to inform users about this. I think in future we might be able to test 384-3G combinations and give this service a result as well, but for now I think the note will have to suffice…

Regards,

RPM
 
rpm said:
Hi Harth

Thanks for the feedback. I agree that it makes a big difference and I have added a note to inform users about this. I think in future we might be able to test 384-3G combinations and give this service a result as well, but for now I think the note will have to suffice…

Regards,

RPM
Click to expand...

RPM, I'm pretty sure that all the home384 test results you received were for 384 + 3G. No one in their right mind would use the 2 GB package, it doesnt allow local access after being capped ;)
 
<suggestions revision="$ 1.0 $">
  • Please state that all prices are Inclusive of 14% VAT somewhere.
    .
  • In the rating blocks, change PRICE to Affordability.
    .
    The issue with price is that you don't know whether the rating blocks say: high price, or low price. You have to scroll up to compare prices to figure that out.
    .
  • Telkom charges differently according to Residential and Business accounts.

    Obviously HomeDSL384 is residential and not offered to Businesses.

    It would be good to have a price comparison between ADSL512 for Residential & Business customers.
    .
  • We also need unbundled comparisons that also show the 30 day notice contract pricing for DSL & iBurst (AFAIK R599).
    .
  • Similar to last point, not quite though. We also need startup costs.

    In the case of DSL:
    * ADSL Installation Fee on an existing analogue line R404.
    * No existing analogue line: Analogue Line Installation (AFAIK ~R2xx.00) as well as ADSL Installation Fee of R404.
    * If you take the 30 day notice contract, you need to buy your own Modem/Router ~R550.

    In the case of Sentech:
    * Activation Fee (AFAIK R500)
    * Delivery Charge (AFAIK R70).
    * If you take the 12 month contract R3500 (AFAIK) for the IPWireless modem.

    In the case of WBS iBurst:
    * Activation Fee (seems to change on a daily basis - says R456 on website, but others have been quoted different prices).
    * If you take the 30 day notice contract <edited old_text="R2200">R2100</edited> for UTC or R2800 for UTD.
</suggestions>
 
Last edited:
Well done!

I would add the business rates where applicable eg ADSL 512.

I suppose you will release the final version with a press release?

Is this going to be a regular event?

Regards
 
what is iBurst paying RPM???

RPM

(1) These iBurst reliability stats are way out of line. The system is EXTREMELY unreliable... many times I cannot even access the Internet.

(2) The comparision of SA ISPs should be normalized against international norms from the outset. Giving any SA ISP anything about 30% is a crime.

Regards,

J
 
Last edited:
is pricing all inclusive?

Hi RPM / All

Forgive my ignorance, but is pricing all inclusive? My understanding was that the 648 per month covers only line rental. What about a dialup account?

It would be handy if you break the all the costs down for each service. Consumers need to know all charges associated with getting connected.

Thanks for the good work.

J
 
Hi guys

Thanks for the feedback.

Kaspaas: We will try to do this at least every 3 months, but believe me when I tell you that the whole process takes a tremendous amount of time. I will take all the feedback from this one to create a better analysis for the next installment.

JamieB: We could only go on the results from the testers… While these services might be unreliable, the testing results showed that this was not the case. If we run the tests for a one month period it might give more accurate results, but remember that this whole exercise is based on volunteer work which makes it a bit tough…

If anybody has suggestions regarding better software or testing procedures we will be very interested to hear from you.

Regards,

RPM
 
Hi IC

Thanks for the great feedback…will make the changes tonight.

Regards,

RPM
 
Hi JamieB

The pricing used for the cost calculation (for the final percentage) involves a rather complicated formula which includes, among other things, basic monthly cost, cost per Gig, installation, cost per additional Gig etc. We also take international standards on capping, pricing structures and various other factors into account.

In the International comparisons section the local services are indeed normalized against international offerings (and as you mentioned none of them received more than 30%). The reason why we have two sections is mainly to ‘stretch’ the results for local services to create a clearer distinction between them. You should notice that the order of the services does not differ between local and international…only the percentages.

Regards,

RPM
 
Rpm

As you know, I participated in the tests.

As far as I can see, TcpIQ has no means of testing when a connection is down. My graph would have painted a very different picture had the program had this facility.

You ought to take in account the variability of the throughput in your assessment. If you take a look at the iBurst graphs, connection speeds are highly variable. One moment download throughput is at 200k then the next its 48k. The upload speeds are non-existant on most graphs - meaning that most users battle to post simple forms such as the the one I am about to submit now.

By definition a reliable connection is one that gives a consistent level of throughput equivalent or above the advertised throughput for both upload and download and has 100% uptime. iBurst had neither of these. It deserves a reliability rating of 1 bar at most.
 
RPM

Good work. Grammar looks okay to me, but I am an Engineer.

It would be good to include the statistics behind your summaries. For those who still need to be convinced.
 
I'm sure this is a first in South Africa!

RPM - point taken about the amount of work!

I guess that the ISP's involved will have something to say about this as well. Actually I'm looking forward to their comments/inputs.

Pity though that Telkom is still the no 1 supplier of 24/7 services to the consumer. It can't be called broadband in any way.
 
rpm said:
Hi IC

Thanks for the great feedback…will make the changes tonight.

Regards,

RPM
Click to expand...
Whoa, there's no rush, take a break, have a kit-kat or something - you do deserve it ;)
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X