New incognito search engine

[Stosskraft]

Register your own domain with Dyn and run your own email server. I'm very happy with mine.

I suppose I should run through the setup process once more, documenting it step-by-step.

Sounds like a good idea. I am getting a website made for my small business so I will look into it.

[SteveRiley]

Just wish I could pull myself away from gmail.

Register your own domain with Dyn and run your own email server. I'm very happy with mine.

I suppose I should run through the setup process once more, documenting it step-by-step.

[bra|10n]

Startpage privacy policy gets even better.
Thursday, July 5th, 2012

Startpage Stops Recording Browser and Platform Information Entirely.

Great news! As of this week, Startpage will no longer record browser type and platform information (also known as the "user agent") of our users. This change has gone live on both Startpage and our sister search engine Ixquick (www.Ixquick.com), and translation of the text into non-English languages is underway.

https://startpage.com/eng/press/star...ts-better.html

[Stosskraft]

Another privacy enhanced search is ixquick's startpage: https://startpage.com/

Choose your poison, but since it's mostly a matter of trust, you should probably pick one of the more established players.

Of course, practically any search engine is likely to be more private than google.

I have switched to them also. really happy with the searches. Just wish I could pull myself away from gmail.

[bra|10n]

SteveRiley for your response also in representing KFN. I too received an email pointing out a post had indeed been made. Remarkable all 'round...

[SteveRiley]

Robert, on behalf of all of us at KFN, I wish to extend my gratitude to you. It's truly rare that an organzation would choose to comment about their policies on forums such as ours. I commend you for that.

While investigating EuroPrise, I noticed that you renewed your certification after the first one expired. I also noticed that EuroPrise publishes the summary findings -- which I had planned to ask whether you'd be willing to share but now don't need to.

I am considering de-Goolefying at least part of my online existence, so I'll take note of your forthcoming email service when it's ready.

Dear Forum Members,

As the CEO of Ixquick and Startpage.com I was asked by 'Dennis' to contribute some background information about our search engines privacy practices to this forum.

Recognizing the importance of search engine privacy as early as 2006, we were the first search engine to start protecting this important civil right.
Protecting the privacy of our users has been our full focus ever since. More information can be found here: https://ixquick.com/eng/protect-privacy.html

We initially focused on logging of IP addresses and ID-cookies. In 2006 we started deleting all IP addresses completely from our log files. Initially we kept IP addresses for 48 hours to be able to block automatic programs that abused our site by 'scraping' the search results. This increased server loads and costs and increased response times. In 2009 - after we were able to deploy more advanced abuse detection systems - we stopped logging IP's altogether and since then only keep it in cache for a few minutes. https://ixquick.com/eng/press/pr-zeroip.html There is no legal obligation for search engines to store any data whatsoever by the way. This is different for ISP's and Telco's.

We also stopped the use of identifiable cookies very early and now offer a completely cookie free service. We offer a a settings-specific URL on our settings page.
If people rather want to use an 'non-id cookie' to set their preferences in - it's also possible. https://ixquick.com/do/preferences.p...age_ui=english

Then we continued with further privacy enhancing features.
While not storing IP's and getting rid of cookies offered the possibility to search in privacy - clicking on a search result - including the sponsored links - would still take the user outside our website and privacy protection and revealing his/her identity.
To facilitate also being able to review a found web site's content in privacy we started offering a "proxy" service in 2010: https://ixquick.com/proxy/eng/help.html
Clicking on the word ' Proxy' underneath a search result opens the website through our servers - keeping the users unknown to those website.

The difference between Ixquick and Startpage is the source of the search results. Startpage offers Google syndicated feed search results for which we have to pay.
Ixquick offers a combination of various search engines sources like Yahoo - Gigablast - open directory etc. The combined services delivered well over 2 million searches per day in March of this year.

Trust is everything in privacy. To facilitate this we have undergone a very stringent certification process by Europrise - an EU funded privacy certification entity from Germany. https://www.european-privacy-seal.eu...elease-en.html This process has taken us close to a year and went into unbelievable detail. It includes random checks and updates.

In 2011 we decided to change to default HTTPS to prevent ISP's from snooping. Over the last 2 years we have been working very hard to offer a truly private Web mail program. It will have various forms of mail encryption and many privacy features. We hope to release it by the end of this year.

Hope this information will contribute to a better understanding of our position in the privacy world to your forum members.

Sincerely,

Robert Beens
CEO Ixquick.com / Startpage.com

[GreyGeek]

.... Now, I too am suspicious about just how much anonymity one can achieve.

In my opinion, very little. That's why I make no attempt to conceal my identity when I post. My handle is easily connected to my real name and address. IF my political opinions were a threat to my personal freedom and/or safety in a country where freedom of speech and political association are written into the 1st Amendment then the ball game is over. If I lived under a dictatorship, or PC run amoke, and I could be imprisoned for expressing views contrary to group think, then I would try to anonymize my internet connections, but I know that such attempts may keep Joe or Sally Sixpack from learning who I am but it wouldn't in any way hinder a government supported group of hackers from determining who I am. They could, for example, intercept packets from both sides of an anonymizer and compare the input packets with the output packets. It wouldn't take much to determine the IP of input packets that compared in all respects, except the output integer, to determine the IP of the sender.

IF I wanted secrecy I would use a 1024 or 2048 or 4096 by encryption key, encrypt all my messages, and give the public key to those whom I wanted to read my messages. But, how long could I keep my key secret if I am tied to a chair and being hit with clubs or had electric wires attached to body parts? Probably not long at all unless the revelation would cause someone else to lose their life.

[Ole Juul]

As far as several hundred people using the same IP as you are, they really don't.

I figured they'd have a way to get back to me, otherwise things could get pretty lonely here. So really, those IP reporting sites that are (in my case) off by a hundred miles (and several IPs) are really just deceiving people by making it look like that's all there is to it.

When an IP packet leaves your router, heading for your ISP's gateway, each packet has your IP address, the one your ISP assigned to you (static or dynamic) beginning at bit offset 96, for 32 bits. Network Address Translation protocols (NAT) can change that IP address in route. In the case of the search engine in question that is not necessary, since they are not acting as a 3rd party intermediary, like an anonymous server is. But, anonymous servers do use NAT to switch the "source IP address" information to another IP address associated with THEIR server. They take your packets, and who knows how many from how many other users, and replace the source IP address info with their IP address. Meanwhile, using higher level info in the packet itself, or TCP./UDP port translation, they keep translation tables which link your real IP address to the returning packets usiing that higher level information. So, in the anonymizing server's RAM, or on their HD, is a table linking some integer value with your real IP address. That integer value was packaged with the outgoing IP packet containing the anonymizer's IP address. To that Anonymizer that Integer value is YOUR IP address.

There are other methods to do this as well.

Thanks for the explanation GG. I realize now that a single IP is not the defining identity it is often made out to be. Of course there have to be other records generated along the way. Now, I too am suspicious about just how much anonymity one can achieve.

[SteveRiley]

IANA handed out the last of its available blocks to regional numbering authorities about a year ago. ARIN, the numbering authority for the Americas, has exhausted its block. APNIC, the authority for Asia, has changed its allocation policies and is restricting what remains. RIPE, the authority for Europe, will likely run out in 2013.

Why aren't we seeing an explosion of IPv6? Pure economics. While IPv6 support is appearing in many products, it's still cheaper to apply that stupid band-aid called NAT. This will, of course, come to bite us all in the ass of complexity soon, because NAT is dumb. Oh, and it appears there's money to be made selling unused IPv4, too.

[GreyGeek]

So, according to all the prognostications, we were supposed to run out of IPv4 addresses sometime last fall. What happened? It seems like VERY FEW sites, or hardware, is ready for IPv6. I know my wifi isn't, nor is my ISP.

[SteveRiley]

When IPv6 takes over all they'll do is keep a look up table on their servers with your IP address and the IP address they give your outgoing packets. No TCP/UDP or port translation will be necessary.

The original design of IPv6 used MAC addresses to generate IP addresses, thus guaranteeing uniqueness and eliminating the need for NAT. Eventually it became apparent that this approach results in a loss of privacy, which RFC 4941 seeks to solve and has been implemented in all modern IPv6 stacks.

(I, uh, would like to call your attention to the authors. I'm not one, but I have been affiliated with the employer of one. )

[GreyGeek]

.....
Also, I always get stumped on this IP thing for identification. I don't doubt that an ISP can identify individual users, it's just a matter of how. With several hundred people using the same IP as I am, the mechanism isn't clear to me.

All this search engine is PROMISING to do is delete the link between users IDs and their IP addresses. Because of the current situation with ALL legal entities I don't trust any of them. As far as several hundred people using the same IP as you are, they really don't. When an IP packet leaves your router, heading for your ISP's gateway, each packet has your IP address, the one your ISP assigned to you (static or dynamic) beginning at bit offset 96, for 32 bits. Network Address Translation protocols (NAT) can change that IP address in route. In the case of the search engine in question that is not necessary, since they are not acting as a 3rd party intermediary, like an anonymous server is. But, anonymous servers do use NAT to switch the "source IP address" information to another IP address associated with THEIR server. They take your packets, and who knows how many from how many other users, and replace the source IP address info with their IP address. Meanwhile, using higher level info in the packet itself, or TCP./UDP port translation, they keep translation tables which link your real IP address to the returning packets usiing that higher level information. So, in the anonymizing server's RAM, or on their HD, is a table linking some integer value with your real IP address. That integer value was packaged with the outgoing IP packet containing the anonymizer's IP address. To that Anonymizer that Integer value is YOUR IP address.

There are other methods to do this as well.

When IPv6 takes over all they'll do is keep a look up table on their servers with your IP address and the IP address they give your outgoing packets. No TCP/UDP or port translation will be necessary.

[Ole Juul]

While I'm not concerned with "law" enforcement, I asked the question regarding the exposure time of 48 hours... is it still considered NOT private after the time has elapsed?

I'm guessing that if they don't have it, then it's not private. In fact it isn't anything. Anyway, what did they say?

[kubicle]

is a plain admission that lxquick DOES store personal data of its users.

Logically, no. If someone says "I don't sell cars", it doesn't logically follow that he/she has cars to sell (he/she might have, or might not).

IP address is probably logged for abuse/DOS attack protection purposes, but if they delete the logs after 48 hours (as they say), it isn't much of a privacy issue. Of course it's ultimately a game of trust, whether you believe they do what they say they do.

But like I said before, anyone is probably more private than google that we know gathers everything, stores it indefinitely and has the power (by the extensive number of services they provide) to combine data imaginatively from different sources.