ettubrute
Expert Member
AntiThesis said:
What an optimist! 
-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: How often do you receive spam calls?
ICASA ADSL Draft Regulations
- Thread starter rpm
- Start date
What an optimist! 
AntiThesis
Executive Member
Ok ok... I'll unbend a little.
The completion of this process is imminent.
*rolls around chuckling hysterically*
The completion of this process is imminent.
*rolls around chuckling hysterically*
bekdik
Honorary Master
ernstn said:
Telkom's tariffs have to go through ICASA ...
Arthur said:
Amen brutha! Amen! Unfortunately things are not that simple for our poor minister who seems to be caught in a time warp and thinks socialist style controll is good for the country. She says all the right things ala USSR style propaganda but does nothing concrete or simply moves ever so slowly on something that really needs to be moving quite quickly.
I agree with you though. Would be nice if the SAfrican telecoms industry were left to sort itself out via competition.
ernstn
Expert Member
When - Not according to ICASA if it is SAIX.bekdik said:Telkom's tariffs have to go through ICASA ...
bekdik
Honorary Master
ernstn said:
What's SAIX's company registration number?
WarpedShadow
Member
After reading the Draft Regulations I am a little confused of the boundries between the "network operators" (NO) [which I assume are Telkom and SNO only] and "Service Providers" (SP) [who I assume include ISPs].
Who "enforces" the cap? the NOs or the SP? I thought cap would be an SP issue since they wold be buying bandwidth on wholesale, but the regulations states that teh NO should inform subscribers about the cap?
Also I see that they do not want to allow the NO to charge for different port prioritization, but does this mean that SP can't either?
Who "enforces" the cap? the NOs or the SP? I thought cap would be an SP issue since they wold be buying bandwidth on wholesale, but the regulations states that teh NO should inform subscribers about the cap?
Also I see that they do not want to allow the NO to charge for different port prioritization, but does this mean that SP can't either?
bekdik
Honorary Master
ic said:
Bingo - you get the prize!
antowan said:I agree with you though. Would be nice if the SAfrican telecoms industry were left to sort itself out via competition.
LOL - That's excellent incentive for the DoC: Quick inform GOVT that they don't need to do ANYTHING, just let competition sort out the ICT sector for them.
It looks like when they TRY to get involved, the process is so slow that it gets nowhere.
Got this from an ISP discussion I have been following and the source will remain unknown for the purposes of this discussion. Please tell me what you think.
I must admit that I was taken aback at how different the views of the ISPs are to that of the consumers. IMO the differences between what the ISPs and consumers differ much more than they would have in a normal open market where Telkom would be faced with proper competition. Naturally one cannot expect the interests of business to beperfectly in line with customer interests but in SA they are totally out of whack because of the current status quo. It is IMO so bad that certain people in the business community root more for Telkom than the consumer purely from a profit protection environment. I encourage people truly interested in the views of business to find the major associations and sign up for their mailing lists and newsletters.
I can however understand why they say and do what they do but it is just another force we need to push through then whenever their priorities differ abnormally from that of consumers.
I just so wish the government would wake up and open the market to more than just an SNO and get the market functioning normally. Customers are not king at the moment and that is the way it should be!
PLEASE NOTE: I DID NOT WRITE THIS!
Should the source of this want to identify him or herself, they are free to do so.
--------------------------------------------------------------
Problem 1:
2.1 and 2.2 entrench the requirement to purchase a Telkom PSTN service in order to obtain ADSL.
Why?
Why should people pay for a PSTN service if they do not require one? They should have the option to obtain telephony services from a VoIP provider or simply not to obtain telephony services where they do not require them.
Solution:
Amend to remove "monthly telephone line rental" and associated wording.
Problem 2:
2.2 is ambiguous, referring to an undefined term "connectivity charges" (do they mean the equally ambiguous "access charge"?) Why?
It doesn't clarify which charges apply at time of installation vs those that are recurring.
Solution:
Define various cost elements and charges unambiguously, clearly indicating which are once-off and which are ongoing charges. Make use of defined terms in wording.
Problem 3:
2.3 does not define the "cost of providing connectivity of such services."
It is therefore open to abuse in the form of outrageously high installation fees.
Why?
If Telkom face potential loss of monthly revenue, they may be incentivised to increase the installation fee.
Solution:
Restrict the scope. eg. "cost of installing local loop and configuring DSLAM port." (Possibly include cost of DSLAM port, however, this is more likely a rental cost since the port is recovered if the user cancels.)
Problem 4:
2.4 specifically forces ISPs to purchase services from Telkom that are probably in the VANS arena and therefore services they should be able to provide.
Why?
This is illegal and anti-competitive.
Solution:
Define elements of the ADSL network, separating IP bandwidth from inter-exchange capacity and empower ISPs to purchase either or both from Telkom at their discretion.
Problem 5:
2.4 makes no attempt to define 'wholesale rate' or limit the extent to which Telkom can inflate it.
Why?
Telkom are still able to resell via their retail arm TelkomInternet. As long as they are, ISPs have little or no negotiating power for the 'bandwidth'
invisaged.
Solution:
Link 'wholesale rate' to cost or establish a framework for fair pricing.
Problem 6:
3.4 allows for potentially unlimited cost of service and high service charges to consumers.
Why?
Local bandwidth has a cost, albeit lower than international. Neither Telkom nor ISPs can be expected to provide a service below cost. International bandwidth has a signficant cost.
Solution:
Define a reasonable local cap (higher than the international one) based on analysis of the costs and consider how 10GB can be made as affordable as 3GB currently is.
Problem 7:
3.6 is inaccessible information
Why?
Obtaining copies of white papers that are subject to constant change is impractical and costly.
Solution:
Telkom web-site should have a facility to check if an exchange is ADSL-enabled and, if not, when Telkom plans to enable it.
Problem 8:
4 is open to way too much interpretation.
Why?
Which ITU recommendations?
Solution:
ICASA must quote the relevent current ITU recommendations in the regulation.
Problem 9:
5 is so ambiguous, conflicts with 3.2, and could push the costs of service up significantly. It also fails to effect the intended result.
Why?
It doesn't prevent prioritising ports; it simply says that one may not charge more for unprioritised service. It conflicts with 3.2 in that 3.2 allows for surcharges whereas 5 doesn't.
Problem 10:
6 doesn't take cogniscance of the fact that contention can appear in different parts of the network and varies with sales. It is also useless to the consumer.
Why?
Inter-exchange links may be contended at different ratios to local bandwidth and international bandwidth and it is no use if only the best of these is quoted. The contention ratio is also useless to the consumer since better contention ratios will be achievable in areas with different customer profiles and volume and can therefore not be used to assess or enforce quality of service received.
Solution:
If kept in place add the word "regularly" before 'publish' and change 'contention ratio' to "contention ratios for various elements of the network." Better yet, replace the whole section with proper quality identifiers. Eg. operator to publish average latency, packet loss and jitter stats achieved on its network over the past three months.
Problem 11:
7 does not provide for urgent maintenance and is impractical.
Why?
Contacting all the customers is too burdensome.
Solution:
Keep the first sentance in 7, but amend the rest. Require publishing on web-site of planned maintenance.
Problem 12:
8(v) is impractical as other unpredictable factors may influence the achieved rates.
Why?
If this were as simple as theoretical maximum, users would have referred to IEEE or similar speficiations.
Solution:
Eliminate.
Problem 13:
8(vi) is useless
Why?
It doesn't refer to where the file is being downloaded from. There are no guarantees made, so the term 'guaranteed' is ambiguous in this context. If the bandwidth were guaranteed, the time would be obvious (3MB = 24Mb + IP overhead ~2.7% + ATM overhead ~11%).
Solution:
Eliminate.
Problem 14:
8(vii) Restricts ADSL to being used as a single bearer for IP service and limits sharing of the bearer link.
Why?
The current PPPoE allows multiple users on a LAN to share a single ADSL access but each purchase their own account from the ISP of their choice.
Connection sharing isn't limited to a single ISP. PPPoE also allows for use of multiple service providers, enabling your ISP to serve you at the same time as your Video-on-demand service provider.
Solution:
Either eliminate on the basis that it is unnecessary if one has eliminated the regular disconnections, or allow users to chose between RFC1483 configuration (leased-line like) or RFC2516 PPPoE configuration (dial-up like).
Problem 15:
8(ix) limits upstream rates to 384kb/s
Why?
There is no good reason. Upstream rate should be as high as user can afford to pay for.
Solution:
Eliminate.
-----------------------
I must admit that I was taken aback at how different the views of the ISPs are to that of the consumers. IMO the differences between what the ISPs and consumers differ much more than they would have in a normal open market where Telkom would be faced with proper competition. Naturally one cannot expect the interests of business to beperfectly in line with customer interests but in SA they are totally out of whack because of the current status quo. It is IMO so bad that certain people in the business community root more for Telkom than the consumer purely from a profit protection environment. I encourage people truly interested in the views of business to find the major associations and sign up for their mailing lists and newsletters.
I can however understand why they say and do what they do but it is just another force we need to push through then whenever their priorities differ abnormally from that of consumers.
I just so wish the government would wake up and open the market to more than just an SNO and get the market functioning normally. Customers are not king at the moment and that is the way it should be!
PLEASE NOTE: I DID NOT WRITE THIS!
Should the source of this want to identify him or herself, they are free to do so.
--------------------------------------------------------------
Problem 1:
2.1 and 2.2 entrench the requirement to purchase a Telkom PSTN service in order to obtain ADSL.
Why?
Why should people pay for a PSTN service if they do not require one? They should have the option to obtain telephony services from a VoIP provider or simply not to obtain telephony services where they do not require them.
Solution:
Amend to remove "monthly telephone line rental" and associated wording.
Problem 2:
2.2 is ambiguous, referring to an undefined term "connectivity charges" (do they mean the equally ambiguous "access charge"?) Why?
It doesn't clarify which charges apply at time of installation vs those that are recurring.
Solution:
Define various cost elements and charges unambiguously, clearly indicating which are once-off and which are ongoing charges. Make use of defined terms in wording.
Problem 3:
2.3 does not define the "cost of providing connectivity of such services."
It is therefore open to abuse in the form of outrageously high installation fees.
Why?
If Telkom face potential loss of monthly revenue, they may be incentivised to increase the installation fee.
Solution:
Restrict the scope. eg. "cost of installing local loop and configuring DSLAM port." (Possibly include cost of DSLAM port, however, this is more likely a rental cost since the port is recovered if the user cancels.)
Problem 4:
2.4 specifically forces ISPs to purchase services from Telkom that are probably in the VANS arena and therefore services they should be able to provide.
Why?
This is illegal and anti-competitive.
Solution:
Define elements of the ADSL network, separating IP bandwidth from inter-exchange capacity and empower ISPs to purchase either or both from Telkom at their discretion.
Problem 5:
2.4 makes no attempt to define 'wholesale rate' or limit the extent to which Telkom can inflate it.
Why?
Telkom are still able to resell via their retail arm TelkomInternet. As long as they are, ISPs have little or no negotiating power for the 'bandwidth'
invisaged.
Solution:
Link 'wholesale rate' to cost or establish a framework for fair pricing.
Problem 6:
3.4 allows for potentially unlimited cost of service and high service charges to consumers.
Why?
Local bandwidth has a cost, albeit lower than international. Neither Telkom nor ISPs can be expected to provide a service below cost. International bandwidth has a signficant cost.
Solution:
Define a reasonable local cap (higher than the international one) based on analysis of the costs and consider how 10GB can be made as affordable as 3GB currently is.
Problem 7:
3.6 is inaccessible information
Why?
Obtaining copies of white papers that are subject to constant change is impractical and costly.
Solution:
Telkom web-site should have a facility to check if an exchange is ADSL-enabled and, if not, when Telkom plans to enable it.
Problem 8:
4 is open to way too much interpretation.
Why?
Which ITU recommendations?
Solution:
ICASA must quote the relevent current ITU recommendations in the regulation.
Problem 9:
5 is so ambiguous, conflicts with 3.2, and could push the costs of service up significantly. It also fails to effect the intended result.
Why?
It doesn't prevent prioritising ports; it simply says that one may not charge more for unprioritised service. It conflicts with 3.2 in that 3.2 allows for surcharges whereas 5 doesn't.
Problem 10:
6 doesn't take cogniscance of the fact that contention can appear in different parts of the network and varies with sales. It is also useless to the consumer.
Why?
Inter-exchange links may be contended at different ratios to local bandwidth and international bandwidth and it is no use if only the best of these is quoted. The contention ratio is also useless to the consumer since better contention ratios will be achievable in areas with different customer profiles and volume and can therefore not be used to assess or enforce quality of service received.
Solution:
If kept in place add the word "regularly" before 'publish' and change 'contention ratio' to "contention ratios for various elements of the network." Better yet, replace the whole section with proper quality identifiers. Eg. operator to publish average latency, packet loss and jitter stats achieved on its network over the past three months.
Problem 11:
7 does not provide for urgent maintenance and is impractical.
Why?
Contacting all the customers is too burdensome.
Solution:
Keep the first sentance in 7, but amend the rest. Require publishing on web-site of planned maintenance.
Problem 12:
8(v) is impractical as other unpredictable factors may influence the achieved rates.
Why?
If this were as simple as theoretical maximum, users would have referred to IEEE or similar speficiations.
Solution:
Eliminate.
Problem 13:
8(vi) is useless
Why?
It doesn't refer to where the file is being downloaded from. There are no guarantees made, so the term 'guaranteed' is ambiguous in this context. If the bandwidth were guaranteed, the time would be obvious (3MB = 24Mb + IP overhead ~2.7% + ATM overhead ~11%).
Solution:
Eliminate.
Problem 14:
8(vii) Restricts ADSL to being used as a single bearer for IP service and limits sharing of the bearer link.
Why?
The current PPPoE allows multiple users on a LAN to share a single ADSL access but each purchase their own account from the ISP of their choice.
Connection sharing isn't limited to a single ISP. PPPoE also allows for use of multiple service providers, enabling your ISP to serve you at the same time as your Video-on-demand service provider.
Solution:
Either eliminate on the basis that it is unnecessary if one has eliminated the regular disconnections, or allow users to chose between RFC1483 configuration (leased-line like) or RFC2516 PPPoE configuration (dial-up like).
Problem 15:
8(ix) limits upstream rates to 384kb/s
Why?
There is no good reason. Upstream rate should be as high as user can afford to pay for.
Solution:
Eliminate.
-----------------------
Last edited:
PostmanPot
Honorary Master
i'll believe it when i see it.
MyDraadloos
Expert Member
antowan said:
Who holds the leash on the flearidden mangy mutt ?
Don't think Gov and Telkom would like to see their profits eroded.
A step
I tend to agree with people saying "Will believe it when I see it". It is great to know that the regulator is trying something, but I have reservations that all the nice stuff outlined there will happen quickly
Only competition will make things right ... but while there is none all we can do is hope that the regulator has some kind of power to enforce the regulations
I tend to agree with people saying "Will believe it when I see it". It is great to know that the regulator is trying something, but I have reservations that all the nice stuff outlined there will happen quickly
Only competition will make things right ... but while there is none all we can do is hope that the regulator has some kind of power to enforce the regulations