Jola
Honorary Master
Optimistic, are we ?
-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: Are you keeping your Prime subscription?
Shocking stats about SA's Internet quality
Optimistic, are we ?
Yeah you're right, 2020 was the date they touted in the past 4 years. The DOC will probably change that "aim date" to 2030 soon.
FlatspinZA
Expert Member
^This - My Telkom techie reckons that solving an intermittent fault with his gear takes 30 seconds, using software Telkom gave him. He has no clue how to launch a command prompt and ping the authentication server. It can't be my gear, because I have 3 modems that all do the same thing.
I wouldn't say the prices are ridiculous if you consider the area Telkom needs to cover, I don't have an issue with that. It's not like the UK where you have millions of people jammed into a tiny area - there are miles and miles of gaps where there are no subscribers. It's relative.
What gets on my tit is them upgrading exchanges to 20/40Mbps when they can't even give reliable service on 4Mbps connections. Get the basics right, for EVERYONE, and then move on.
By that logic we should be far above Brazil and perhaps even the USA. What's that I hear? Oh right, we are far below and far more expensive.
Bern
Expert Member
Right, so there are some really stupid comments that are being made here by a few people. Just to clear something up on the ADSL VDSL FTTH issue. Telkom is sorting things out, you need to start by getting the core upgraded (which I believe they have mostly completed now), then you need to start expanding this outward to the exchanges, but they are going one further and are implementing the MSAN option to move the exchange closer to you. This means a shorter copper loop to you and the later possibility for FTTH.
The issues around quality of the technicians sent out to do grunt work and the managements bad investment decisions is something completely different.
Back to the topic itself - can we please get some information on how this is measured? These kinds of figures are all highly subjective based on the test methodology, so without an explanation on how the tests are done this article has absolutely no meaning at all. A couple of points for starters:
Does this include wireless connections?
Does it take in to account the option the user has taken (i.e. if I have a 1Mbps line I am not going to get 4Mbps downloads)?
Who did the tests and measurements?
What is the sample size and the samples per geographic location?
What did the test actually comprise of?
While I am sure we are at the bottom of the list I don't like blind statistics, they need a reference to have meaning.
The issues around quality of the technicians sent out to do grunt work and the managements bad investment decisions is something completely different.
Back to the topic itself - can we please get some information on how this is measured? These kinds of figures are all highly subjective based on the test methodology, so without an explanation on how the tests are done this article has absolutely no meaning at all. A couple of points for starters:
Does this include wireless connections?
Does it take in to account the option the user has taken (i.e. if I have a 1Mbps line I am not going to get 4Mbps downloads)?
Who did the tests and measurements?
What is the sample size and the samples per geographic location?
What did the test actually comprise of?
While I am sure we are at the bottom of the list I don't like blind statistics, they need a reference to have meaning.
Rocket-Boy
Honorary Master
So by that same reasoning you shouldn't accept positive results for speed tests.
Come to think of it you shouldn't accept any results unless they have personally explained their testing details to you.
![[)roi(]](/forum/data/avatars/m/8/8711.jpg?1563828683){kind=avatar}
What would be more interesting is to proportion this by ISP including # of customers (and / or transactional volume).
I suspect cellular edge services contribute heavily to this piss poor stat; and the reason for edge speeds is as a direct result of cellular operator greed and ICASA back-boneless.
I suspect cellular edge services contribute heavily to this piss poor stat; and the reason for edge speeds is as a direct result of cellular operator greed and ICASA back-boneless.
Last edited:
I'm with Bern on this. While the Internet quality in South Africa has plenty of space for improvement, I'm a little skeptical of these stats.
You can't reliably collect these kinds of stats without large volumes of end user tests or probes. Has anyone on this forum participated in this research?
You can't reliably collect these kinds of stats without large volumes of end user tests or probes. Has anyone on this forum participated in this research?
People (understandably) want cheap data. Cell C for example is aggressively pursuing this need/want/desire to get more traction in the market. But is is it sustainable? I don't think so. This puts pressure on the others to do the same so we end up with a situation where the business model starts to fail and where investment in upgrades and new roll-outs become unattractive. And so the domino effect is set in motion. Contention goes up. Service levels come down. Why the graph shown earlier is in steep decline after a steep incline, I believe.
The same goes for wired communications. More people then also go wireless putting even more pressure on the wired business model. Companies are not sure whether to break through the "wall" by larger more aggressive investments (risk!) or to take a more conservative approach and try to better optimise what is already there with throttle tactics and connection dumping.
The same goes for wired communications. More people then also go wireless putting even more pressure on the wired business model. Companies are not sure whether to break through the "wall" by larger more aggressive investments (risk!) or to take a more conservative approach and try to better optimise what is already there with throttle tactics and connection dumping.
Rocket-Boy
Honorary Master
Im interested to see the results from project Bismark. It is a test of adsl clients where you have a router connected to your network that measures stats and performance of your line over a period of a few months.
Granted that would be adsl only and not a true reflection of all internet connections.
These stats are absolutely CORRECT for Bryanston, 10ms, 3.1 MB download and 0.38MB upload. Tested this morning after reading the article. In addition, the line is very unstable so it is not possible to watch You tube or any other TV live...
grim
Expert Member
While I understand that we have far more distance to cover, it's ridiculous that Telkom expects us to pay R140/m for line rental for what amounts to two pieces of copper running into your house where you can get a cellphone with R135 airtime on a top up contract for R135/m. There is no value in having a landline for the average South African nor having ADSL unless you are gaming or moving large amounts of data every month, 3G beats ADSL to sh!!t in value for low bandwidth users, which make up the majority of internet users in SA. The main reason that BB is so successful in SA is that for R60/m you get unlimited internet use on your mobile device and most of those users use less than 500MB a month, ADSL has no value in SA for users that use so little.
Sodan
Expert Member
Don't worry, guys, the DoC will save us by commissioning yet another study on broadband in SA.
Yeah, I know, that would be funny if it wasn't so true.
Yeah, I know, that would be funny if it wasn't so true.
Bern
Expert Member
Seriously? "personally" explained? I don't accept the positive results either without methodology explained, surely you agree without explaining HOW the tests were done the results are pretty meaningless?!? All I would like to know is where this data comes from, ISPs, user tests or what? Remember there are a lot of GPRS and EDGE users out there that can sway stats if it is per number of users. If it goes on data volume those numbers would be much higher.
Maybe they could at least discriminate fixed line from mobile/wireless.
Project Bismark? Sounds interesting - any details you can add? Do they just pull SNMP to get the modem stats?
3.1Mb down is not 0.33Mb down.