geraldbaeck

Of course you are right, with such data you could build something like alexa. We are aware that this data is currently a tiny subset of the web and does not represent much, but it has potential if it grows.

Since we do not consider this data to be ours, we do not plan to charge anyone for the dumps.

Gerald, CTO Blippex

Yes, we will add metadata in the next dump, but currently the time_spent is all that matters.

Our plans are to add categories, the language and rough information about the content type (video, image, etc).

Gerald, CTO Blippex

Nice idea I think we should do that.

Gerald, CTO blippex

We have some rate limits at our API in place, but of course it not that difficult to change an IP-address. But most important we wrote some algorithms which checks submitted URLs and domains for suspicious or accelerating behaviour. If that happens we simply suspend that domain for some time. We are also planning to publish those suspensions.

-Gerald, CTO archify/Blippex

Differentiating the quality of a dwell would be nice, but that would mean to track search trails of our users, which is too much of a privacy issue. But we are thinking about semantically evaluating the DwellRank. For example a useful dwell would be for tutorials. But this is just an assumption, we simply need more data about that.

Gerald (Disclosure: archify/Blippex CTO)

Thanks. Yes we will implement a Firefox search engine addon soon.

I don't think that those "generic" term are the main advantage of Blippex. But if you search for example for "NSA" on Blippex (https://www.blippex.org/?q=NSA), we are assuming that you will get those articles about the NSA which is currently the most interesting or the most read. -Gerald (Disclosure: I am the CTO of archify/Blippex.)

Thanks, that is a very interesting input. We should think about running additional semantic analysis and relate them to the time spent on sites. We are very sure that our algorithm needs a lot of fine tuning and this could be a very important part of it.

-Gerald Disclosure: I am the CTO of archify/Blippex.

Could you please send me a link to the header modifying addon. So I can fix that.

gb@blippex.org

We are counting the per unique URL, which is currently not a big advantage for the popular sites, because they are hosting so many URLs. There no domain based factor in it right now.

-Gerald Disclosure: I am the CTO of archify/Blippex

I think you guesses are absolutely right. We never intended to compete against other engines in the field of "Answers". I think you will always get a better result for ""Population of America" if you search for ai at DuckDuckGo or Google. But in the other hand if you search for example for "NSA" on Blippex (https://www.blippex.org/?q=NSA), we are assuming that you will get those articles about the NSA which is currently the most interesting or the most read.

-Gerald Disclosure: I am the CTO of archify/Blippex.

Yes that could be a potential issue, expecially if you want to keep the data of the submitters as anonymous as possible. We are monitoring this closely, but honestly we don't have a single hammer solution against spam as always it needs my small steps. I will post some thoughts about it on our blog in the days.

-Gerald Disclosure: I am the CTO of archify/Blippex.

doesnt work with usernames only with pages like cocacolazero

how are they doing that?

Dropbox No Nees for USB drives anymore

Google Docs easy collaboration within our startup

and of course Twitter

The OS does not matter, because almost everything is running in my browser.

We decided to go for YC first and if they don't take us we will apply at other incubators.