SwiperProxy: Fast open-source web proxy
swiperproxy.github.io
swiperproxy.github.io
I went with this approach because I wanted to use it with Opera Mini and it had no support for normal proxy protocol.
Why I wanted to send all my traffic through yet different proxy? Simply, I wanted to change the web app I was using to find bus connections so it had bookmarks and could remember where is my home, so I don't have to type it over and over again.
So I made this XML transformation language that behaved quite like jQuery so I could move elements on the site and have some state between requests.
https://github.com/mikz/rProxy/blob/master/app/plugins/isis_... https://github.com/mikz/rProxy/blob/master/app/plugins/dpp/a...
A detailed bullet-point list of features (available an obvious click or two from the front page - I appreciate you might not want that sort of detail on the front page in order to keep it clean) would be useful for assessing it though, i.e. does it support SNI (so it could proxy for several HTTPS end-points using a single IPv4 address), IPv6, and so forth.
Also, benchmarks against using nginx and other common options as a proxy would be handy, though I get the impression from the blurb on the front page that ease of installation/configuration/maintenance is more of a priority than speed (not that this is a problem - getting that side of things right is a most laudable goal)
Writing up a feature list and running benchmarks are some good ideas, thanks. :)
Edit: I should probably mention, the live demo is heavily ratelimited as an easy way of conserving bandwidth. This way, it is more than fast enough for browsing pretty much anything, but makes it a less interesting for mass downloading large files and streaming video. It can still do that if you don't rate limit, obviously.
https://swiperproxy.github.io/documentation/sys-shorewall.ht...
Unless I'm missing something that's not a possible comparison. Nginx is a reverse proxy, but this is a forward proxy.
I'll look again later...
Which I'm totally fine with, running these things for a lot of people doesn't come cheap and if that's your business model then hey, go for it. I don't particularly want to use them though, and the only way that's avoidable is going self-hosted.
With self-hosted proxies your list of available options shortens dramatically. Sure, there's all kinds of VPN's, SOCKS proxies, hell, even Tor, but all of those require you to have people configure things on their machine. The first test it needed to pass, was sending it to my grandparents and see if they could use it without any explanation whatsoever.
And then, why stop there. Lets make it as easy for the sysadmin to deploy as well. Hence why it immediately runs as a fully functional webserver. There's a good few advanced things you can do with it, yes, and all of them are either already documented or being documented right now. But at the end of the day, any novice sysadmin should be able to clone and have a functional proxy within five minutes out of the box.
These two things are something I haven't seen any other proxy do so far and that's the use case: a very simple web proxy for both the end user and the person running it. At this point I feel it strikes a very nice balance between simplicity and having plenty of features and customization.
I'm not trying to be ornery, and I'm not proxy illiterate, I worked on Squid for a decade, and have built my own toy proxies to test ideas. I'm trying to convince you to make your website more clear about why someone would use your software.
You can't just call it a "proxy" and expect understanding. Even when talking to someone who has deployed 1000+ proxies, you're talking past me. There are a lot of proxies that do many things. What is yours for? What do you want people to do with it? There are caching proxies to make slow links seem faster and use less bandwidth, there are security-oriented proxies that anonymize to one degree or another, there are proxies to allow users to safely tunnel through a firewall in a locked down corporate environment, there are reverse proxies for scaling and accelerating websites, there are proxies for content filtering and access logging used in schools and libraries and such, there are proxies that simulate different network conditions to aid developers, there are proxies for sniffing user data illicitly, there are proxies to allow people in other countries to watch YouTube or Netflix. There are all sorts of proxies. What is yours for?
Using the DNS on AWS returns the correct answer.
http://dnsviz.net/d/donttrustrobots.nl/dnssec/
Edit: I've removed dnssec from that domain in the mean time. Should be reachable on any dnssec validating resolver as well when it has propagated.
Of course it's entirely possible for whoever runs a proxy to sniff all of your data; that's unavoidable. There's a big warning by default on the index page for this exact reason. I'm not overly concerned about phishing though. If it happens, it happens, whether or not I publish proxy software.
And really, even if I did have the proxy modify the DOM to generate huge warnings, what's the point? Anyone who wants to use this as a phishing tool can just as easily remove it. That, and the average phishing victim usually doesn't look at warnings, but will click past them.
There are already better phishing methods in place., and phishing is going to be a problem regardless of this software's existence and regardless of what it does.