Big player already do that.Why should I trust a small party too?
Not being cynical, just wanted to see if the payoff is worth the compromise.
I'm the developer of the extension.
Totally valid point, we thought a lot about privacy, and our approach to managing the issue is two fold
1. Your browsing history isn't identifiable on our databases since we use a random hash from your credentials to track your history. 2. We have absolutely no interest in using the data, even non-identifiable data for marketing purposes 3. We leave the final control in your hands, if you feel your privacy is breached, you can clear all your data from our servers and uninstall the extension.
Here's our privacy policy for reference - https://getfetch.net/privacy.html
How do we know that you in fact have properly randomized the hashes, really aren't using the data, and most importantly, how do we know you really are clearing the data from your servers when we ask you to? Are you going to make the source code available?
If its running on your servers, you have to plan to monetize right? How do you plan on doing that?
Edit: I should note it looks quite cool, and I am probably going to try it, but was wondering about the above points.
One additional question. Is it possible to ban specific sites from showing up? Like work related intranet things?
1. Fetch indexes page content as well as the titles and URLs of your history, making it easy to find links 2. Fetch ranks search results by engagement which means the most useful links usually surface on top when you search.
How do you handle page versioning? Do you take a new snapshot every time I visit - and do you keep the most recent version or can I look at/search through previous versions?
Ranking pages by how much time you spend on them seems great from a productivity standpoint. From a search standpoint, often the pages I want to go back and find ex post facto are the ones that I only looked at for a few seconds. So maybe it's still helpful? Yet to be seen!
A little feedback on the page copy. Any time someone claims they have "proprietary algorithms" it sets off alarms. It's empty jargon, and sounds scammy. Your headline "Say Goodbye to Bookmarking Webpages" is good because it speaks to people who already use bookmarking services and find them lacking. But "Imagine never having to re-google what you had googled before" reaches a broader audience and also speaks to the actual problem they have.
We take a snapshot every time you visit the page. We don't version the page yet. Do you have a specific use case for a feature like that?
There's a very strong correlation between time spent and usefulness of the article. We index all articles, even the ones you spend less time on, so when you search, a few page scrolls get you to the link you need. That said, time spent is a heuristic, its useful in a lot of cases, but not all, we'll have to see.
"Proprietary algorithms", we have a patent pending on how we sort and filter data
We'll think of a catchphrase around ""Imagine never having to re-google what you had googled before"
Hope you enjoy using the extension. You can send us feedback from the plugin directly, or mail us at dev@tryheld.com
> Imagine never having to re-google what you had googled before :)
and im still looking for that app/extension that will allow me to do that with firefox..
But some people don't care about their privacy any longer, because, well we probably have none at this point anyway.
So in for a penny, in for a pound.
Edit:
Consider putting something about privacy right at the top of the page, if such a feature exists. e.g all bookmarks are encrypted, and only the user has the key, or whatever.
If you want security-conscious folks to use your service (that is, hackers from HN), please provide more information about how you actually plan to use the data and more robust language about security and transparency. There's no way in hell I'd let a third-party like yours access the entirety of my browser history without seeing some evidence of vigorous privacy protection first.
Anyone know of a similar local effort? (I spent 98% of my time on my own two machines anyway...)
However the main challenge is that much of the web is now SSL, so a light caching proxy isn't feasible unless it does MiTM which I don't want.
My current idea is to log DNS lookups but that doesn't, of course, provide page content data.
Also something I've learned building a sass business: you need customer feedback. People who pay you money are properly motivated to give you the feedback you need.
