Hacker News new | ask | show | jobs
by lambda 1052 days ago
And even when it doesn't block you completely, it delays website loading, makes you jump through frustrating captchas, etc.

It's probably third in the list of frustrating web behaviors in the past couple of years (behind GDPR popups and registration/paywalls that seem to have gotten much worse recently).

And somehow there are some sites that I get CF delay walls on every time I visit.

This feature is utterly broken for a good web experience; it pushes users away from sites which use it.

Every time that "checking your browser" page comes up for a legitimate user should be considered a failure. Sure, it can maybe happen a few times in a thousand, but the feature is utterly broken if it comes up every time I visit the same site from the same browser not in private mode.

2 comments

It's worth noting that the websites that you are visiting chose Cloudflare, and have enabled the features that irritate you. They have browser integrity enabled, have bot protection enabled, maybe turned the security level up (gitlab famously is a nuisance because they lean heavily on Cloudflare for protection). Sometimes they've wholly barred VPNs or entire geographic areas! And that is entirely the decision of website operators, and note that they did all of this before Cloudflare came along.

Cloudflare's customers are website operators, not you the end user. Those website operators seem pretty pleased with the service, so clearly they are doing a good job for the people who they are building it for.

And every Cloudflare customer is a company I won't do business with (unless there is absolutely no way around it)

Cloudflare is running the single biggest, most blatant man-in-the-middle attack in history, and far too many people are happy about it

Agreed. The same goes for the 3rd party "data privacy" popups which simply hide a long list of opt-outs several layers deep in a Vendors list. I refuse to use such sites and I let them know by email.
In what way is it an attack? (I know what a mitm is, I'm not asking you to explain that - I'm pretty conversant in the concept of a proxy, I'm asking you to explain why it's an attack specifically)
they block and/or slowdown vast swaths of the internet

if that's not an "attack", I don't know what is

I don't get it. They offer a service that that people choose to sign up for and take active steps to use. I don't see how that's an attack. Honestly I'm still trying to understand who is being attacked.

Like is it an attack on the site owner - are you saying cloudflare is extorting them or something? That seems unlikely but I agree that would be a form of attack... it also doesn't seem to be what you're saying.

Is it an attack on the user of the website because the website owner successfully denies visitors it does not want? Does that mean that login credentials are a form of attack too? Would an on-prem load balancer or WAF that dropped all traffic from a region or matching patterns still be an attack?

It just doesn't make sense that it's an attack.

They block and/or slowdown over 20% of the internet

How can you not see that as anything but an "attack"?

And this is precisely why I don't bother reporting Cloudflare's failures to site operators anymore. I used to do it, when it was pretty infrequent. Site operators were usually concerned that something was blocking customers, but most were clueless about what was causing it or how to fix it.

Eventually I gave up. I don't even bother with their captchas or other stupid human tricks anymore. Whenever Cloudflare gets between me and the site I'm trying to use, I move on and shop somewhere else. Life's too short for this.

Why not take a screenshot of the CF error and send it to the website owner? It would freak me out if I thought a significant number of my website's users were being blocked by CF.
I’ve done this before, and the response is always “this is the first time I’ve seen this” and “you must be a bot operator”.
+1 to anybody who creates a site to name and shame CF customers who block legitimate traffic. For a few months now I've been taking screenshots every time this happens, but with no end goal. Complaining to the individual site owners feels like a lifetime commitment, and there are virtually none I need that badly.
I have done that numerous times. Even sent a screen recording of the Cloudflare spinner of hell. The response is always the same: you must be running some shady software on your machine.

Cloudflare is acting as judge and executioner, and site owners never accept that the product may be faulty.

They will just tell you to use unmodified Chrome.

And soon with Web Integrity API they may start telling you to use Chrome on Windows or MacOS, rendering Linux completely unusable.

The work needed to maybe get it past outsourced customer support is not at all worth the effort for any site I don’t actually need to use
How do you send them that when you can't access the contact form and/or contact information on their site because Cloudflare blocks it? (assuming a normal visitor, not someone who knows about whois etc.)
Send it to the domain contact from WHOIS information.