OK, so:
a) You would be fine if youtube content was actually held at CellC's data centre after the first download, but would prefer a dns and uri redirect to simply transparent access
You're mixing a CDN and a cache.
Dont care where a CDN is held, whether it's cellc or mweb or whoever - it's irrelevant. The point of a CDN is to provide a speed benefit by distributing load across many servers, ideally with more local server. Subscribers who can can stream internationally from youtube's existing CDN's have no incentive since there is no cap exclusion.
Why would anyone provide a cap benefit while reaping the OOB revenue costs.
The CDN - for the youtube example - would actually be managed by youtube's heartbeat and direction servers with cellc (or whoever) being added as a CDN and mirroring. There would be no dns URI redirections as such, in almost all video CDN's only the video is streamed from the CDN not the page content. Page content is sent from seperate servers on a different CDN.
b) You believe that CellC has n amount of subscribers times average subscriber connection speed of WAN bandwidth on tap (at no extra bursting cost).
They have it now and will have it in future, with WACS et al coming online I very much doubt this is an issue. In either way, as far as a CDN goes it's not as if it's cost effective (OOB) for anyone to stream to the degree that it would affect the network so I fail to see this argument point. Split the total youtube or whichever site you want's bandwith consumed by the number of subscribers.
Cell C and other mobile companies charge a heckuva lot more for OOB than it costs them to provision WAN.
I admit that caching would benefit the bottom line of the income statement of the provisioning company, my issue with it is the detrimental effects of caches if they are not implemented flawlessly and without affecting other services. If a cache is implemented the option should be available to choose to be on a
non-cached APN.
c) You believe that a caching connection would involve something else 'not-working' and if this was true that the user would have to do the switching
Have I got that right?
No. I'm saying that IF through some magical way the CDN / Caching actually did improve the speed to a few sites for people with bad speed it would not keep them on cell c (or whoever). IF they did get a speed benefit from the caching I can't see anyone holding onto their connection just for youtube (or whatever) and switching physical connections to browse and get email. I.e. from cell c to whatever.
SUMMARY:
I'm saying that for people who
a) do not have a problem with streaming a local CDN is irrelevant since there is no bandwith cost benefit for users.
b) have no speed (i.e. bad signal etc) the local CDN would not benefit them anyway and neither would caching - slow is slow.
c) have speed, the above (b) is a non-entity and most people would likely prefer to wait 200ms longer for a page and not have issues with caching systems being flaky or interfering.
Enough, I think my opinion has been laid out and this is turning circles and this is beginning to feel like a trollfest. I'm done unless there is actually something new to comment on.
Your question was if a caching system would be well received - I doubt it.
A CDN would be potentially beneficial, but since there is no local vs. international differentiation it is a question of speed benefit and if the mirroring is efficient.
A cache intercepts and potentially interferes. A CDN does not.
It's all a moot point anyway since it's doubtful customers would have any say in the implementation of a cache anyway. Likely the first we would hear is when things stop working and eventually someone admits to the issue.
Sign up with Google