51–60 of 127 posts
250ms to Facebook and that 'sucks', glad to see New Zealand's opinion on our internet connectivity is validated! I routinely see 270-300ms pings to facebook. Anything on Amazon's CDN is at 150ms that as is CloudFlare. Akamai does have local nodes, as does Google but the lack of competition on the cable makes it slow. Guess I should move to Kenya?
Earlier quoted context omitted.
And sniff/proxy _everything_ for economic gain. Your company's motives are not pure and altruistic. Don't pretend they are.
Your company's motives are not pure and altruistic. Huh? Let's be clear: CloudFlare is a for-profit company and hence wants to take in more money than it spends. We do that by getting people to pay us. Our revenue is from web site owners paying us. It's pretty clear what our business model is: pay is a little for bandwidth and hardware is possible, get web site owners to pay us.
That is suprising to me. I lived in Tanzania for a year, and while the internet connection there was not perfect by any stretch of immagination, I had way less ping than this (100 ms to London, where my company's servers are, 120-150 to Boston). For reference, that is: aerial to my ISP two kilometers away, then fiber to Dar Es Salaam, then Seacom to Mombassa, then Seacom again to Marseille, then terrestrial fiber to…
the eassy link covers much of the same route: http://www.submarinecablemap.com/#/submarine-cable/eastern-a...
Earlier quoted context omitted.
Your company's motives are not pure and altruistic. Huh? Let's be clear: CloudFlare is a for-profit company and hence wants to take in more money than it spends. We do that by getting people to pay us. Our revenue is from web site owners paying us. It's pretty clear what our business model is: pay is a little for bandwidth and hardware is possible, get web site owners to pay us.
You might want to ask your founders for a copy of their original pitch deck.
250ms to Facebook and that 'sucks', glad to see New Zealand's opinion on our internet connectivity is validated! I routinely see 270-300ms pings to facebook. Anything on Amazon's CDN is at 150ms that as is CloudFlare. Akamai does have local nodes, as does Google but the lack of competition on the cable makes it slow. Guess I should move to Kenya?
I really hope that something like ppsp / swarm / ipfs gets integrated into browsers. I'm pretty sure almost every non-user-content file I download every day has been downloaded somewhere in my town. I shouldn't have to go all the way to the US to get it.
In one of Google's lobby there is a 3-D animation of a spinning globe, with spikes that appear wherever and whenever searches are performed, with the height of the spikes being proportional to the rate of query submissions. The recruiter who showed me this was very proud of google's accomplishment and indeed it was impressive to see the spikes erupting from the east and west coast of the US, Europe, Japan and India.…
I'm in Africa and... I'm sorry, but we have nameservers already. That's not really a problem. Are you really talking about DNS servers here? I wasn't sure. The issues which prevent those spikes you'd like to see on the maps at Google are education (primarily literacy, especially English literacy), access to devices, and culture. Access to devices is being solved by cheap smartphones becoming obtainable to some degree…
One of the things that hurts progress in my country isn't really a lack of education as much as it is ignorance in the strict sense. It's funny, because a couple of years ago, a trans-Pacific cable was laid close to the island, but the decision was made to not invest in tapping into it due to the cost--even nowadays, people are interested in improving connections, but it isn't a priority. They really haven't experienced the internet, at least not in it's full capacity, so they don't know what they are missing. They know about it from what they see on T.V. ("viral" review shows targeted at baby-boomers who still watch T.V., I suppose) and social networking, so simply, they don't know what they're missing. I recall when I finish high school and moved back here to the US for college--I basically missed the dotcom-boom days of the late 90's and early to mid 2000's, the internet was not just a convenience, it completely changed my living habits.
But the so-called "other three billion" really can't know this because they haven't experienced this, so not only do you face real obstacles, but you face lethargy as well. At least for Palau, I feel like that is one of the main reasons internet access is pretty subpar there today.
[*] I do admit that this might not be relevant in parts of Africa. For one, Palau is a pretty small place (no more than twenty-thousand people), so attitudes are probably much more homogeneous (with that attitude towards the urgency of improving internet infrastructure) compared to hosts of nations with millions of people in sub-Saharan Africa.
Earlier quoted context omitted.
Your company's motives are not pure and altruistic. Huh? Let's be clear: CloudFlare is a for-profit company and hence wants to take in more money than it spends. We do that by getting people to pay us. Our revenue is from web site owners paying us. It's pretty clear what our business model is: pay is a little for bandwidth and hardware is possible, get web site owners to pay us.
That's a nice straight forward business model most people can get behind. The problem comes later, as it does when any company gets big enough. Few years down the line, the board of directors has a look over the company and see's a big opportunity: all that data you're flowing around can be turned into money with some very clever thinking at 'no real loss to anyone'. People don't like to leave money on the table if t…
When they pay for our service they are trusting us with their DNS and HTTP traffic. If CloudFlare were to start messing with their data it breaks that trust and we lose our customers. We'd end up biting the hand that feeds us.
In one of Google's lobby there is a 3-D animation of a spinning globe, with spikes that appear wherever and whenever searches are performed, with the height of the spikes being proportional to the rate of query submissions. The recruiter who showed me this was very proud of google's accomplishment and indeed it was impressive to see the spikes erupting from the east and west coast of the US, Europe, Japan and India.…
I'm in Africa and... I'm sorry, but we have nameservers already. That's not really a problem. Are you really talking about DNS servers here? I wasn't sure. The issues which prevent those spikes you'd like to see on the maps at Google are education (primarily literacy, especially English literacy), access to devices, and culture. Access to devices is being solved by cheap smartphones becoming obtainable to some degree…
Earlier quoted context omitted.
I'm in Africa and... I'm sorry, but we have nameservers already. That's not really a problem. Are you really talking about DNS servers here? I wasn't sure. The issues which prevent those spikes you'd like to see on the maps at Google are education (primarily literacy, especially English literacy), access to devices, and culture. Access to devices is being solved by cheap smartphones becoming obtainable to some degree…
Yes, the issues are substantially bigger. South Africa can't even provide reliable electricity in the cities, never mind elsewhere. People living there tell me it's not a problem. If they're not bothered by that, I doubt they give a damn about google searches. If memory serves me correctly that's been going on for about 10 years now.
There's the electricity issue (we call the rolling blackouts "load-shedding") which affects the cellular base-stations and telco exchanges as much as anything, but also another more obvious issue for content delivery: the local loop.
For the past 15+ years (as long as I've had Internet access), the incumbent telco (TelkomSA) has not been forthcoming with upgrading or maintaining the ADSL infrastructure. Finally, there seems to be traction with FTTx providers, but this seems limited to affluent pockets of a few thousand people in the country. All the new fibre offerings serve the same locations.
While it's hopeful this will improve the situation in the medium-term, the backhaul from ISP to eyeball is oversubscribed at least 30:1 for ADSL, with equipment being replaced by (IMO) inferior network equipment. Additionally, there are reports of automated line-conditioning software which should improve local loop connectivity, but I've seen DSL sync issues only since the software has been implemented.
The (ex-public, privatised under Telkom) Exchanges have been neglected and have recently been divided into three groups: Earners, Maintainers and "Lost Causes". The latter two will not receive upgrades from ADSL1 and the Earners will likely migrate to newer technologies. This is largely due to copper-theft and vandalism of exchanges/cellular base-stations (which are oversubscribed enough that wireless technologies routinely sustain 3000-20000ms latencies). The oligopoly running cellular networks have no incentive to improve this, since the Regulatory Authorities don't really seem to push competition as well as we'd like.
TL;DR: There are issues, but rose-tinted glasses make me hopeful in the medium term. Claims are that wireless will fix everything seem bleak, since base stations become oversubscribed faster than they can be built/upgraded (and the backhaul from them is measured in megabits/s, rather than gigabits/s charged at ~US$0.15/MB).
Bummer about all those IPv4 traceroutes. I'd like to see the output of IPv6 traceroutes or mtr output.
From what I've seen running an HE tunnel through London, Youtube and Facebook make up the bulk of IPv6 traffic (as expected). At least those networks are the only thing routinely dedicating 10%+ of the total inbound traffic to IPv6 according to my home network flow analysis (I think it's about 18% IPv6 usage on a monthly basis).