And Hetzner is a no-nonsense real company that's been around since 1997.
Hetzner machines have dedicated 1Gbps connections. Arc.io promises nothing.
Arc.io is a faceless company with no track record, no history, and doesn't say how long they've been around.
Who would you choose?
The crypto/distributed space has an issue with vaporware and not living up to claims. Arc.io has a lot to prove, in other words.
I'm hoping users will think twice before using infrastructure that's not ready, or up to par, or not likely to have longevity.
If you're going to sell high-performance then you should at least make your own website perform well.
If your website performs well then I'm more likely to believe your story.
(*I'm not assuming Arc.io does this, just that "fast" and "reliable" are table-stakes properties of CDNs.)
2. It also puts me, the end user, in an odd position because I don't control the content that is being stored and served from my machine.
Putting those two things together I find myself worrying about the following scenario: I visit site example.com that serves content that is legal in my jurisdiction. example.com is using arc.io and while I'm browsing example.com some content that is illegal in my jurisdiction is served to someone. This puts me in a position where illegal content was stored and served from my machine.
I think that puts me in a dangerous position just for surfing example.com.
If I signed up for Arc Rewards then I might get paid for storing and serving illegal content.
We endeavor to create a communal web where everyone cooperates and benefits, and we want to make that as easy as just visiting your favorite website. Everyone shares a little surplus bandwidth to help each other load the web faster, sustain their favorite sites, and earn rewards. Everyone wins.
That's a different web than the current web. Right now, with the current web, no one shares and nothing is shared with you. The web isn't communal; it's adversarial. Ads interrupt you and trackers stalk you. We want to change that.
It will take time. Just like how the once frightful notions of getting into a stranger's car (Uber) or sleeping in a stranger's house (Airbnb) are now banal. These things take time.
But, of course, if preferred, an opt out is two clicks away. Sites with Arc must display Arc's widget in the lower left so visitors can learn about Arc and easily opt out (or back in).
> I think that puts me in a dangerous position just for surfing example.com.
When one visits nytimes.com, they don't know what files will be downloaded and cached on their computer before they hit [enter]. They trust nytimes.com not to hand them inappropriate content (porn, violence, etc).
Arc has that same responsibility -- to only cache appropriate, legal content. It's our responsibility to get that right. And we will.
- Sites, and their content, are reviewed and vetted before they get access to Arc's CDN.
- Automated checks are run on assets for appropriateness, a la Google's SafeSearch Cloud Vision API. Inappropriate content never enters the network.
Hope that helps.
Also, and wholly unrelated: big fan of Movie Code. =]
One one site that had a few hundred unique visitors per day, and the P2P CDN "saved" me kilobytes per month in bandwidth costs. Maybe a higher traffic site would see more benefit (especially if it isn't focused at tech people who have a nearly 100% install rate for ad blockers).
>Users receive 10% in user rewards for sharing bandwidth with the network.
Does this mean the users who are visiting the site with the Arc widget (acting as seeds for the p2p cdn)? Is Arc going to send me a check in the mail, and if so, what registration do I need to do to get that money?
If most users don't sign up to be paid, what happens to their "share" of that 10%? Is it redistributed to other users, or kept by Arc?
I'm not sure what other optimizations they're doing or what guarantees it has. Wish they had a page with more complex assets being delivered via their CDN to test performance. Their own site is being delivered this way, but I don't think that's the use-case. There are many ways to host simple static sites free/cheap -- the big CDN costs often come from media like videos or images, which in addition to being delivered also need to also have very low latency to "feel" fast. (Not sure a jpeg fragmented and delivered over WebRTC is going to outperform a CDN delivering it over a single HTTP request regardless of how fast it is. Adaptive video is also not going to be straightforward.)
Price doesn't seem appealing to me given how it works and how little is explained. But it's a cool concept, and I'm all for new ways to decentralize content and/or save on delivery costs.
Now the site is down...which makes me think this cdn doesn't work so well.
I currently pay ~500/mo for a cdn so I am a real potential customer.
Out of curiosity, is it profitable in the end? 2k new users per day seems like a lot.
CDN Bandwidth $0.02 per GB
Egress (worldwide) (76% cheaper than AWS1)
Usage is billed monthly, on the first day of each calendar month.
Additionally, only peer-to-peer egress bandwidth is billed; all peer-to-peer ingress bandwidth is free. For example, if a 100kB image is loaded from Arc's peer-to-peer network, only 100kB of bandwidth is billed, not 200kB (100kB of egress + 100kB of ingress).
That makes no sense and this whole idea sounds like a pitch I would see for an ICO.
As long as it delivers better performance, it might be worth it.
For comparison, Bootstrap is 180KB, Elm is 29KB and Vue.js is 100KB.
It might not be much for sites otherwise dealing with large multimedia assets, but for anyone going for lean / fast web development, this seems to be a no-go.
How does this affect GDPR? Are we using & monetizing customers' bandwidth without their permission?
not covered by General Data Protection Regulation.