At least if we were updated with the network issues, Seacom / ISP, I'm sure people would be more understanding
Last edited:
-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: How much are you willing to pay for an official Springbok shirt?
ADSL consumer packages - A DigiChilli survey
- Thread starter CambodiaDave
- Start date
At least if we were updated with the network issues, Seacom / ISP, I'm sure people would be more understanding
Shayd
Expert Member
Bandiwdth management on uncapped
I have an interesting idea, introduce client type shaping as they have in some overseas ISP's.
You define 3 service levels uh Gold,Silver and Bronze. Any requests sent to the server are handled on a client type basis so a Gold user would have their traffic prioritized then Silver then Bronze.
You can run a script that looks at the last 10 rolling days usage on all accounts, the server then separated the clients automatically into:
Gold, the top 3rd who use the service least
Silver, the middle third who use the service moderately
Bronze, the bottom third who use they service for major downloading.
The Gold users will be able to use the service for all time dependent applications and wouldn't be negatively affected by many users all trying to torrent at once.
Silver users would have a service that is mostly reliable and would only lag slightly in peak times.
Bronze users would have a service that would only work for HTTP, torrents and email etc. The services that aren't time dependent. This service would most likely be laggy as hell.
This way you would have a "natural" bandwidth ecosystem in which the less heavy users are still subsidizing the heavy user's but gain the benefit on a reliable consistent service.
Eg of 10 day rolling threshold: Bronze user stops downloading for a few days is automatically moved to silver or gold and will now be able to play games and use VOIP...etc
What do you guys think?
I have an interesting idea, introduce client type shaping as they have in some overseas ISP's.
You define 3 service levels uh Gold,Silver and Bronze. Any requests sent to the server are handled on a client type basis so a Gold user would have their traffic prioritized then Silver then Bronze.
You can run a script that looks at the last 10 rolling days usage on all accounts, the server then separated the clients automatically into:
Gold, the top 3rd who use the service least
Silver, the middle third who use the service moderately
Bronze, the bottom third who use they service for major downloading.
The Gold users will be able to use the service for all time dependent applications and wouldn't be negatively affected by many users all trying to torrent at once.
Silver users would have a service that is mostly reliable and would only lag slightly in peak times.
Bronze users would have a service that would only work for HTTP, torrents and email etc. The services that aren't time dependent. This service would most likely be laggy as hell.
This way you would have a "natural" bandwidth ecosystem in which the less heavy users are still subsidizing the heavy user's but gain the benefit on a reliable consistent service.
Eg of 10 day rolling threshold: Bronze user stops downloading for a few days is automatically moved to silver or gold and will now be able to play games and use VOIP...etc
What do you guys think?
Last edited:
What do you guys think?
In many ways the whole point of uncapped is to remove the stress associated with the internet. If a user decides today he feels like downloading the latest ubuntu, he shouldn't have to think, oh no, if i download this then i will drop from gold to bronze. It just makes using the net a constant decision game, which really hurts the experience.
The internet in this country is causing an increase in heartattacks because everyone blood pressure is always high. (thought i'd add a dose of drama to the discussion)
megas_alexandros
Active Member
wow its funny how if you stick a bunch of ppl in a box for a wile they cant imagine what could be out side of this "box" lol c'mon guys the way i see it digichilli is the first in a long line of providers that is actually trying to break this monopoly things will get better dont suggest making them worse ><
we want cheap uncapped that works
we want cheap uncapped that works
What about FTP? I'm a developer and also test on live sites.... but now I have to switch to my Afrihost account to even touch FTP.... and uploading 10-200MB each time gets expensive.
Skype doesn't seem to be a problem. Skype uses p2p to connect if it can't connect directly.
Last edited:
SizableHuggs
Active Member
- Joined
- Dec 2, 2009
- Messages
- 36
- Reaction score
- 0
Does no one else think they should focus on getting the network and current services/products working first, not mention their customer service before actually bothering about new porducts. I was in the DC trial and loved it, most of the reason i loved it was the constant feed back we got from dave and the also makhula on occasion it was awesome, and i thought to myslef these are guys we should support they really are trying.
SO then they launch, fine there are some hiccups i can understand that, i still signed up, thinking, its OK they will keep us informed and things will smooth out, INSTEAD communication started to dwindle, and then the network went down hill and now i am sure they are probably working frantically to get it sorted it out, and again having faith we will be kept in loop, NO i was wrong what we got instead is a call/survey about what products we want?!?!?!
Maybe i am just missing a vital piece of some puzzle here. But i definitely think delievering what people have paid for now is far more important than finding out what products they would like you to introduce in the future.
SO then they launch, fine there are some hiccups i can understand that, i still signed up, thinking, its OK they will keep us informed and things will smooth out, INSTEAD communication started to dwindle, and then the network went down hill and now i am sure they are probably working frantically to get it sorted it out, and again having faith we will be kept in loop, NO i was wrong what we got instead is a call/survey about what products we want?!?!?!
Maybe i am just missing a vital piece of some puzzle here. But i definitely think delievering what people have paid for now is far more important than finding out what products they would like you to introduce in the future.
Last edited:
LancelotSA
Banned
I partly agree. Not with the 256k throttle though. I don't think we should have anything throttled below 384k as if with these speeds we are treading outside of "broadband".
I have realised this month that I do not have a huge need for international. I have a 5Gb Afrihost account and I find I am not coming close to using it every month. As it stands I have not even used 1Gb and we are on the 13th already! Local data to be used on the news servers, however, is a different story. I have 30Gb available from my Telkom accounts and I am going to use this up soon.
We all know local is cheap so I still begrudge paying the same for an uncapped account when my usage will be totally different to those who smoke international data.
In that case webafrica may end up being a better bet for you. But we will have to wait and see, I know they promised very cheap local.
edit// but remember, part of the point of uncapped international is that you dont have to manage anything. No route sentry or the like needed.
Many people having been calling for uncapped local for eons. I honestly think if you told some people these international accounts are also uncapped local, they would begin to understand the real value in these R300 accounts.
Last edited:
LancelotSA
Banned
Yip, but the problem with the "R300 accounts" as they currently stand is that the speeds are shocking. I moved from IS to SAIX because of shocking speeds. No point in having an uncapped account that does the same for me as IS did.
I have also got used to having free NS access so also begrudge having a R300 account and then having to spend another R100 on NS access. Any service I take needs to put me in a better position financially. I really cannot see myself ever doing more than 50Gb per month as the internet currently stands.
Mcwidowmaker
Well-Known Member
- Joined
- Mar 22, 2005
- Messages
- 150
- Reaction score
- 16
Looking at it more from a business perspective, bandwidth in South africa is expensive. In order to have an internet business model that will work financially as well as for the different users, you need a variety of people using the service.
Offering uncapped only at good prices like this is not financially viable or user friendly. You end up with all the high traffic users contending for bandwidth at a 20:1 or higher ratio and cannot deliver on a good quality service and end up being flamed and loose your customer base.
In order to make the service financially viable and user friendly, you need 3 - 4 types of customers.
1. Low traffic users with a 3 - 20gb cap, that usually do not use their full trafic allowance, but use you because of good service and competitive pricing.
2. Bussiness users that qants quality of service and pays a premium for good bandwidth speed during business hours and that you give priority during these hours.
3. High cap and uncapped users that need lines with low contenttion ratios. These uncapped accounts, almost unshaped gives good speed all month round and they are expensive, usually around R2500 a month or so.
4. High download / uncapped accounts with a fair contention ratio that is shaped during business hours and higher speeds during the evenings and over weekends.
That way you optimally sell all your traffic and utilise your lines well. That way you can offer good pricing and make money.
Having only one of the 4 type of customer either stuff your service or gives you unecnomical line usage.
Your current business plan is not viable, because offering the uncapped only, you will draw all the high bandwidth usuage clients. These guys will have their torrent / news servers running 24/7, usually running at max speed, overutilising the bandwidth you have available and you will end up shaping them to hell and upsetting them.
The capped account users use their accounts while they awake and in bursts only, leaving plenty of bandwidth for the higher usage clients. The same with your business clients.
The lower cost shaped during the day account and running free at night will utilise the bandwidth at night and and weekends and these users are willing to buy the bandwidth at prices advertised.
The really high bandwidth users must pay more. Getting almost unshaped uncapped accounts take up huge system and line resources. Their traffic needs be prioritised and its costs money to the isp. These accounts must be charged accordingly and no way you can supply a 512 guaranteed line at R1000 a month.
Other accounts that can be considered is high cap local only accounts. Since these accounts ustilise very little of your costly international bandwidth, its not too expensive to deliver on these accounts.
Not sure what the interconnect fees are between the different isp's and utilising their news servers, but this should still be able to be fairly cheap.
Combining all these accounts and you have an economically viable business model. You cannot deliver the one without the other.
My 2c worth.
Offering uncapped only at good prices like this is not financially viable or user friendly. You end up with all the high traffic users contending for bandwidth at a 20:1 or higher ratio and cannot deliver on a good quality service and end up being flamed and loose your customer base.
In order to make the service financially viable and user friendly, you need 3 - 4 types of customers.
1. Low traffic users with a 3 - 20gb cap, that usually do not use their full trafic allowance, but use you because of good service and competitive pricing.
2. Bussiness users that qants quality of service and pays a premium for good bandwidth speed during business hours and that you give priority during these hours.
3. High cap and uncapped users that need lines with low contenttion ratios. These uncapped accounts, almost unshaped gives good speed all month round and they are expensive, usually around R2500 a month or so.
4. High download / uncapped accounts with a fair contention ratio that is shaped during business hours and higher speeds during the evenings and over weekends.
That way you optimally sell all your traffic and utilise your lines well. That way you can offer good pricing and make money.
Having only one of the 4 type of customer either stuff your service or gives you unecnomical line usage.
Your current business plan is not viable, because offering the uncapped only, you will draw all the high bandwidth usuage clients. These guys will have their torrent / news servers running 24/7, usually running at max speed, overutilising the bandwidth you have available and you will end up shaping them to hell and upsetting them.
The capped account users use their accounts while they awake and in bursts only, leaving plenty of bandwidth for the higher usage clients. The same with your business clients.
The lower cost shaped during the day account and running free at night will utilise the bandwidth at night and and weekends and these users are willing to buy the bandwidth at prices advertised.
The really high bandwidth users must pay more. Getting almost unshaped uncapped accounts take up huge system and line resources. Their traffic needs be prioritised and its costs money to the isp. These accounts must be charged accordingly and no way you can supply a 512 guaranteed line at R1000 a month.
Other accounts that can be considered is high cap local only accounts. Since these accounts ustilise very little of your costly international bandwidth, its not too expensive to deliver on these accounts.
Not sure what the interconnect fees are between the different isp's and utilising their news servers, but this should still be able to be fairly cheap.
Combining all these accounts and you have an economically viable business model. You cannot deliver the one without the other.
My 2c worth.
TelkomUseless
Honorary Master
- Joined
- Mar 13, 2006
- Messages
- 16,570
- Reaction score
- 11,208
I agree on the speed... IS seems like a speed deamon now.
TelkomUseless
Honorary Master
- Joined
- Mar 13, 2006
- Messages
- 16,570
- Reaction score
- 11,208
I agree 100%. I would love full 384 speed 24/7. 256k is uber slow.
But DC said min 256, so I accept that, but after hours... full speed please. If they can give full speed 24/7, YAH!
happychappy
Well-Known Member
- Joined
- Aug 23, 2009
- Messages
- 261
- Reaction score
- 0
Browsing only account- R100 unlimited browsing and e-mail no downloads
You can keep the other accounts like they are for now, but try and increase the speeds please.
Could be nice is we can see the lines reaching 100KB/s at the end of feb? That will lure more customers to digichilli and should bring prices down in the market. Besides you guys will be rolling in cash if you make less per account but a bundle on all of them???
uncapped local only account - R100 full line speed?
You can keep the other accounts like they are for now, but try and increase the speeds please.
Could be nice is we can see the lines reaching 100KB/s at the end of feb? That will lure more customers to digichilli and should bring prices down in the market. Besides you guys will be rolling in cash if you make less per account but a bundle on all of them???
uncapped local only account - R100 full line speed?
Lightscribe
Banned
Yep, burstable to double the speed you pay for.
Give the test account to someone here that will use it to report stats daily.
NO FULL SPEED!!! DOUBLE THE SPEED YOU PAY FOR, AS PER THEIR ORIGINAL STATEMENT.
PEOPLE HERE EXPERIENCED FULL SPEED, BUT IT IS NOT THE WAY IT WILL WORK.
TelkomUseless
Honorary Master
- Joined
- Mar 13, 2006
- Messages
- 16,570
- Reaction score
- 11,208
Relax dude.
Like I said.. they said 256k minimum. That is what I ask for. If it is possible afterhours.. 384kbs line speeds , not 4 meg line full speed. If you have 4 meg , then buy a another package
As Mcwidowmaker explained, you have to get a nice even spread of users to make money. Us high bandwidth users are expensive clients.
With that in mind, Digichilli should think about creating a "granny catcher" product. More grannies = more subsidizers.
R300 is still to much for grannies.
Dave i'm not sure what cards you are yet to pull, but from where i'm sitting there might be a need for some capped products. As detestable as they are, caps do allow you to go as low as you need to go to get the grannies. :wtf: Unless of course you have another way to get a sub R150-200 product onto the market.
Any ideas?
With that in mind, Digichilli should think about creating a "granny catcher" product. More grannies = more subsidizers.
R300 is still to much for grannies.
Dave i'm not sure what cards you are yet to pull, but from where i'm sitting there might be a need for some capped products. As detestable as they are, caps do allow you to go as low as you need to go to get the grannies. :wtf: Unless of course you have another way to get a sub R150-200 product onto the market.
Any ideas?
Lightscribe
Banned
No. No burstable to 384k speeds for the 256k product.
Burstable to double speed as advertised.
Nothing more. Nothing less.
No. No caps.
Everyone should stop trying to create new solutions/products.
The products are there. The specs are there.
The service simply needs to be tweaked to perform as advertised.
The suggestions are related to the existing products.
Stop trying to come up with new business models.