Power to the people

[Prometheus]

Again I would raise the point of mathematical feasibility.

Even being outrageously generous by saying they will get 50% uptake in each area 'lighten' you land up with a CPE to head-end ratio of 25 to 1. So to service 1mil customers they need to deploy a network of roughly 40,000 head-end units and their associated back-haul links (excluding any dual-homed redundancy).

In comparison I would estimate Telkom's ADSL subscriber to DSLAM/back-haul (at the local exchanges) ratio is several 100 to 1, and their sizable technician force is barely able to service 200K customers adequetly.

And this huge infrastructure maintanace overhead is going to be funded by R200 per subscriber? Maybe I'm missing something, but it just doesn't make sense to me.

With a DSLAM having 20 ports they are in a better position than Telkom. In most other countries the DSLAM is funded by the line rental which is ~R100/month and you get the adsl "line" for free. The problem we have is that everyone can offer R200/month broadband but chooses to screw us over instead simply because everybody else is doing it. Once someone manages to change the mindset from non-competitive to competitive things will change. The only question that needs an answer is will GTS be the catalyst or not.

 

[DragonLogos]

In a rare move of (Dragon) activity, I checkd the mains power.... and it is still as per the energy crisis of thirty years ago - 49 cycles

 

[Prometheus]

In a rare move of (Dragon) activity, I checkd the mains power.... and it is still as per the energy crisis of thirty years ago - 49 cycles

Not likely to change either since there is no reason to adjust all the equipment in the entire country suddenly. They can easily save power by reducing the supply to 220V.

 

[Roman4604]

With a DSLAM having 20 ports they are in a better position than Telkom.

DSLAMs come in many shapes and sizes. Carrier grade stuff is usually chassis based housing multiple high density line cards. It not uncommon for one chassis to provide 3000+ ADSL ports on it.

But thats missing the point anyway. I was talking about the expense and maintanance of such a high number of distinct interlinked locations (or nodes) on the network. Things break (e.g. equip failure, lighting, vandilism etc.) and the more nodes you have them more expensive techies you need to maintain uptime & service levels. Also the more nodes you have the more expensive interconnection/back-haul links you going to need.

Just sound an extremely expensive way of doing things, in my view far more expensive than Telkom's infrastructure, as they will have way, way less than 40,000 (ADSL enabled) local exchanges (nodes) to support 1mil ADSL users.

 

[Prometheus]

Just sound an extremely expensive way of doing things, in my view far more expensive than Telkom's infrastructure, as they will have way, way less than 40,000 (ADSL enabled) local exchanges (nodes) to support 1mil ADSL users.

I don't think so. I can't see the difference in replacing a DSLAM in one of ten locations or in one of hundred. You still have to drive out to the location, take the old one out and put the new one in. With FTTH a viable alternative to DSL backhaul would not be that much of a problem. Here you only need one fibre link to every mini-sub instead of to every house. The main problem in this country, namely the loop is already there. The main problem I see yet again is the pricing. If it is comparable to that of iBurst we will yet again see slow uptake and rollout due to unaffordability and unavailability.

 

[Roman4604]

I don't think so. I can't see the difference in replacing a DSLAM in one of ten locations or in one of hundred. You still have to drive out to the location, take the old one out and put the new one in.

Yes at those number its sounds reasonable, but extrapolate to 40000+ locations. The the more drives you make & the more staff you need to make those drives, the more overheads you incur.

With FTTH a viable alternative to DSL backhaul would not be that much of a problem. Here you only need one fibre link to every mini-sub instead of to every house. The main problem in this country, namely the loop is already there.

Yes FTTH is as viable as copper these days from a hw point of view IF you are laying a brand new link. Its certainly not cost viable in comparison to existing links.

The main problem I see yet again is the pricing. If it is comparable to that of iBurst we will yet again see slow uptake and rollout due to unaffordability and unavailability.

OK lets forget Telkom and compare to iBurst. They have a subscriber to network node (tower) ratio of over 250 to 1 based on latests subscriber & tower figures. Oh and most of their tower backauls are wireless as well (very cost effective).

So again they have a vastly lower capital & overhead cost per subscriber and you want the BPL guys to come in with lower end user pricing.

Sorry, wet dreams I would say.

 

[Prometheus]

Yes at those number its sounds reasonable, but extrapolate to 40000+ locations. The the more drives you make & the more staff you need to make those drives, the more overheads you incur.

I don't see what you're getting at really. Even with say 100 DSLAMS housed in one box what are the chances really of two of them failing at the same time when there are 100x4,000 other DSLAMS that could fail. It's more likely that in the event of say lightning all of them would fail, but now you have 10 in the same location to replace instead of 100. Whether you have 4,000 or 40,000 locations you still have to go out to one of those locations and the average drive time is nearly the same. Lets just suppose 10 of them fail randomly and you now have 10 locations to drive to. You might have had 9. There's an even smaller chance that you might have had 8. As we go from 7 and downwards, well your chances of winning the lottery are becoming increasingly better in comparison.

Yes FTTH is as viable as copper these days from a hw point of view IF you are laying a brand new link. Its certainly not cost viable in comparison to existing links.

My point is they only need to lay one fibre link to each mini-sub instead of one or more copper links to each house. They are in a far better position than any telecoms company could ever dream of being.

OK lets forget Telkom and compare to iBurst. They have a subscriber to network node (tower) ratio of over 250 to 1 based on latests subscriber & tower figures. Oh and most of their tower backauls are wireless as well (very cost effective)

So again they have a vastly lower capital & overhead cost per subscriber and you want the BPL guys to come in with lower end user pricing.

My reason for comparing them to iBurst was not to compare hardware but rather to compare greed. In other countries iBurst could not be this lucrative as they would actually have to compete on price and charge in the region of R200-R300 for a full internet service. Yes I want them to come in with much better pricing. I want these guys to know that if they don't give at least 5GB for R300 they are just wasting our time because availability is not our main problem, price is, and we don't need more infrastructure for being ripped off.

 

[Roman4604]

I don't see what you're getting at really.

Come on, this easy to understand. It blatantly obvious that its more cost effective to service the same amount of customers with fewer sites/nodes/pops. Less travel, less staff, less access equip, less redundancy equip for them, less primary & redundant back-haul links, less routing equip, less enviromental equip (temprature, moisture etc.) less security, less power mgmt equip (UPS), less site rentals etc. etc.

My point is they only need to lay one fibre link to each mini-sub instead of one or more copper links to each house.

Thats not the comparison, the copper links are already there and were not discussing the last mile anyway. 40,000 back-haul links (& their redundant alternatives) are more expensive than 4000 of them, period. And their competitors will only require around 4000 pops or less to service the same amount of customers as them (i.e. 1mil customers at 250:1 cust to pop ratio) ... totally uncompetitive.

My reason for comparing them to iBurst was not to compare hardware but rather to compare greed.

Greed only comes into play if your overheads are low enough to actually make a profit. I dispute they have the ability to do this when stacked up against their competitors.

 

[Prometheus]

Come on, this easy to understand. It blatantly obvious that its more cost effective to service the same amount of customers with fewer sites/nodes/pops. Less travel, less staff, less access equip, less redundancy equip for them, less primary & redundant back-haul links, less routing equip, less enviromental equip (temprature, moisture etc.) less security, less power mgmt equip (UPS), less site rentals etc. etc.

A few things you left out. They are working with the municipality so they already have the sites secured and available to them. The last mile links are also supplied and maintained by the municipality.

Thats not the comparison, the copper links are already there and were not discussing the last mile anyway. 40,000 back-haul links (& their redundant alternatives) are more expensive than 4000 of them, period. And their competitors will only require around 4000 pops or less to service the same amount of customers as them (i.e. 1mil customers at 250:1 cust to pop ratio) ... totally uncompetitive.

Ok, how can you decide that you want to leave out the copper links? They are a central part of the cost. So you think 40,000 backhaul links and equipment are more expensive than 4,000 with thousands of kilometers redundant cabling that has to be supplied and maintained.

Greed only comes into play if your overheads are low enough to actually make a profit. I dispute they have the ability to do this when stacked up against their competitors.

I don't know if you are referring to iBurst or GTS here. iBurst chooses to milk a few customers instead of making money from many. A while ago someone involved with a big international company stated that they withdrew from providing R50 broadband to the masses not because it was technologically unfeasible but because it was legally impossible. Everyone charging a R300-R400 extra monthly "connection fee" are just milking people for what they are worth instead of providing real services. I ask again what makes South Africa so "special" that we can not have these services for R300pm when other countries can?

 

[Roman4604]

A few things you left out. They are working with the municipality so they already have the sites secured and available to them. The last mile links are also supplied and maintained by the municipality.

And you think they are charitable organisation, all efficeintly run so that they can dedicate focus onto non-core services, all with the same forward looking views on broadband provision? Right, couldnt think of a worse underlying infrastructure provider to have to deal with. Why do you think BPL is largely all talk and no delivery ... cos you can't deliver without them unfortunately.

Ok, how can you decide that you want to leave out the copper links?

Because the BPL providers will not get cheap access to it, almost all of it is owned by you know who.

I ask again what makes South Africa so "special" that we can not have these services for R300pm when other countries can?

I agree it can be done in the future, just dont think BPL is the technology that will.

If BPL ever becomes even modestly popular, and can deliver on the price & performance promises, you can use the biggest font to post "I told you so!".

 

[riaan_pta]

All very interesting - but when?

"Maguire promises an announcement on its website, goal.co.za, in this regard in the next few days." - 02/02/2007 - The bloody website is staler than last Easter’s cinnamon buns.

Sorry to sound like the stereotypical SA broadband pessimist again, but this announcement has been "coming" for the past 18 months.

Does anyone know what is really happening?

I am truely willing to cancel my Telkom ADSL service and hand-over my debit-order details if I only knew when!

 

[Prometheus]

And you think they are charitable organisation, all efficeintly run so that they can dedicate focus onto non-core services, all with the same forward looking views on broadband provision? Right, couldnt think of a worse underlying infrastructure provider to have to deal with. Why do you think BPL is largely all talk and no delivery ... cos you can't deliver without them unfortunately.

Never said they were a charity. It's either letting the infrastructure be used and getting something in return for it or nothing. I know which option I would choose.

Because the BPL providers will not get cheap access to it, almost all of it is owned by you know who.

Geez, were you in discussions with them? GTS was and think they are very receptive to the idea.

I agree it can be done in the future, just dont think BPL is the technology that will.

So you think everybody should sit on their asses and just wait for it to come along in the future. We've been doing that for over a decade already, this IS the future. I guess we should wait until 2011 to build stadiums since what they have now is just not financially feasible. I don't care if you think that BPL will never become modestly popular. At least someone in this country is doing something to bring broadband to more people. I just hope they don't kill it by overcharging for it like EVERYBODY else has done so far.

 

[Roman4604]

Geez, were you in discussions with them? GTS was and think they are very receptive to the idea.

Yes at their standard, highly profitable rates. You should know how Telkom operates by now.

Otherwise will be waiting with baited breath for your "Told you so!" post.

 

[redheadfan]

iBurst

Bear in mind these guys started up with no existing infrastructure and next to no money. They've been raising funds as they go along and most of their advertising is carried by their dealers. What this means is that say about 50% of the current customers have had to deal with technical problems and poor service and the majority of the country still cannot access their service. The current business model of iBurst seems to be: Milk to rich and the desperate and slowly grow the network. The biggest problem is that Telkom, Sentech or Neotel might easily eclipse iBurst's technology. So in 2 years time iBurst might have completed their network but they won't have any new customers as the competitive will have far better technology. Also bear in mind with iBurst you have to either sign a long term contract or pay for an overpriced modem.

GTS on the other hand, has more capital behind them and will potentially raise significantly more capital in future. The are also integrating themselves into an existing infrastructure. They are conducting extensive trials before launching the product. They also envisage multiple service delivery (Internet, Phone, Security and TV).

The reason why the rest of the world is so far ahead of us is because of BIG companies with BIG money to throw into the game.

I am most certainly holding thumbs for GTS and Neotel. I can tell you that I'm not keeping my Telkom line a day longer than I have to.

 

[Prometheus]

The reason why the rest of the world is so far ahead of us is because of BIG companies with BIG money to throw into the game.

The way I see it is not that the other countries are ahead as that would imply a valid reason but that we are behind not because of too little big companies as we have plenty of those but because they don't have to compete. The only question that has not been answered is will GTS choose to compete or will they simply be another leech. I suggest that if they charge us more than adsl prices for the bandwidth and more than R100 (what the phone line should be) for the connection we ensure they do not get any customers.

 

[feo]

It seems that the norm here in SA is that even if theres more than one company offering the same service as another company then the attitude is 'we do you a favour by offering a service and if you dont like it bugger off'. Thats why even if GTS does make it into the market eventually I dont see them dropping prices to get market share. Like Prometheus said just another leech added to mix of money sucking cash cows.

This whole high speed internet thing in South Africa is in a very very bad place and we desperately need a company like GTS to actually deliver on their big ass promises. Im also praying that Neotel isn't a flop and becomes Telkom 2. Its sad when you see countries like India with a strong presence on the international web scene when we here in SA hardly make a dent.

 

[JTech]

So...

I have been in contact with the manufacturer of the DS2 system in Spain. The DS2 system GOAL are using is not what I am currently researching. A whole bunch of the usual tripe by MyADSL know-it-alls was posted in here and I got myself confused I admit that.

The DS2 system is RF based, although some of it ends up as modulation on the mains in general it will travel through the wires regardless of whether there is power or not, even through the earth conductor and apparently through steel guttering I have been told. This is why it WON'T travel through the substation transformer.

The systems I work with, use different techniques. We actively modulate the mains sinusoid at the 11kV side and it goes through substation transformers readily and for quite some distance. That's ripple control. The other spin we do on this is also rightfully called a power line modem, yielding bi-directional communications, but nowhere near the speeds of DS2. These signals also propagate readily through transformers and for some distance but if there's no power you are buggered, the modem cannot connect.

My work has nothing to do with BPL, and they better see to it that their **** doesn't interfere with mine.

So next time, please use the proper terminlology and stop confusing people

 

[JTech]

On the subject of cost....

I have found out what the modems cost for this system, as well as the costs of the provisioning equipment, all in US dollars. It's dirt cheap equipment, and if GOAL thinks they're going to come in and fleece the SA consumer just like all the others (Sentech, iBurst, etc..) they must think again.

 

[Roman4604]

It's dirt cheap equipment

Just remember the cost of the the actual access CPE and head-end equipemnt represents a miniscule portion of the costs incured in providing a high-performance widely-avaliable broadband service. Read my previous post re. financial viability problems I have with distribution & backbone portion of the whole BPL architecture, fundimentally due to the very limited short range of local loop.

Interestingly the very same reason why Telkom has so stubbornly hung on to ADSL1 technology for so long, when the norm internationally has been ADSL2 & 2+. Its all about the infrastructural cost (per subscriber) you inccur in using a technology with a shorter local loop range. Obviously overseas they have far higher urban population density concentration which makes short range local loop far more feasible.

Unfortunately its not all about the technology, you have to understand the economics behind it also.

 

[feo]

You left out the part where you actually tell us what the cost is. So how much is it? Or is it confidential?