Warning, the first search tends to be slow, which makes one wonder how the thing works, but give it time and you'll see HN search results on the right. Successive searches are faster.

The first search for me also had some screwy CSS. Mine looked really funky on the first go, which made me skeptic and I came here and read your note. Subsequent searches run fast and don't have CSS issues.

Love it.

Cool cool cool.

I never looked into Webmynd until seeing this. Thanks for adding HackerNews to the list of sources and kudos on the effective marketing :)

Thanks :-)

What other sources would you like to see? If you find yourself ever typing into a search box on some site that isn't Google (or one of the other large search engines) let us know and we'll add it.

I find myself doing a lot of site searches for LifeHacker

The pirate bay would be a good choice :D

This is a very important step towards what Search will be (somewhat social). If there has been a HN thread about what I am looking for, I will check it before I check the Google results.

We launched a similar plugin with Webmynd for Fluther and so far our users have been very receptive.

I have to say it's a brilliant way for them to market their service, and everybody wins.

Does anyone else think it would be helpful if Google integrated options like this into their search?

Google have started to add personalization on the left-hand side with SearchWiki. I think a problem with it is the effect is not obvious - sure the ranking might be different for you if you do the same search again, but how will it effect different searches? By doing much more obvious personalization which is applicable for every search you do thereafter, I think WebMynd gives more value for the (small) effort to configure it.

Also, whatever they might like to experiment with, I think Google will be reluctant to try radically different designs or really mess with the right-hand side of the page. As a public company they need to optimize for ad-revenue.

Those of us at WebMynd do not. :)

Yesterday I installed Feedly and it looked great while using the dashboard and as a start page. But when I went to Google to search something I was surprised because feedly inserted search results (from my feeds) just above google search results.

I then went to Feedly website, searching for info whether they are storing my searches somewhere, how long they keep it, etc. I was unsuccessful so I uninstalled it right away.

Don't get me wrong here, I find this functionality very useful. I also understand it's great business to gather data on people searches and some people really don't care. Just I am not one of them.

So can someone please tell me if the searches I do are sent to Webmynd, stored, sold, hacked or not?

What WebMynd does is covered in our FAQ:

http://www.webmynd.com/html/faq.html

Depending on the settings, WebMynd does collect some search and browsing information. We never reveal this information to third parties. The information is not exposed to other users (even in aggregate / anonymized form).

WebMynd includes the ability to save and index (on our servers) your browsing history so you can find things again faster. This is disabled by default in the Hacker News Search version. You can turn on recording of history and select history search on Google with a couple of clicks if you would like to. More information on that here:

http://www.webmynd.com/html/help.html

We do gather data of how many searches are performed, how users interact with our UI and which sources are generating click-throughs all to try and make the service better. I think this is standard practice for search engines / aggregators and most users expect that.

Thanks for the info.

Hi EastSmith,

Here is the feedly privacy information: http://edwink.devhd.com/2008/06/18/feedly-and-privacy/

There is a section on search:

Question: Does feedly search keeps track of my search history?

Answer: The answer is no. The integration of the search result is perform directly on the client. Feedly does not track at all what keywords people search. The search integration component only tracks at the global level what percentage of users disable the search feature so that we get a sense of if it is valuable or not.

Good job guys. I still prefer your "my top sites" search result though. That thing is just absolute magic :-)

1a. Does HN have any kind of hidden search engine?

1b. If so, could OpenSearch be implemented over it, enabling a real HN search extension?

This is good, but sometimes freezes my firefox, going to disable it anyway.

Good stuff, Brilliant job

Effective marketing at its best!!!

Very nice. Its like <search term> site:news.ycombinator.com but orders of magnitudes more polished and better designed.

How does this differentiate from using 'google site:ycombinator.com [search this]'? I've never had a problem searching for what I wanted using google's site option. i.e. I have 334 HN threads saved in my profile; googling 'site:ycombinator.com mllk' results in 326 entries found. I haven't used the application yet so, of course this is my prebias speaking, but I never really saw a point for redoing a site search besides the idea for using a different index.

Here are a few of ways adding the site-specific search on the right is helpful:

1) If you frequently find yourself referring to sites like Amazon or Wikipedia, or indeed Hacker News now, to find information on particular topics. WebMynd saves you a step. You don't need to navigate there or craft your query in a particular way.

2) If you don't realize that a particular site or source could have useful information on a particular topic. With this toolbar you don't have to remember or think to check, you will be presented results from your favourite sources for each query you do.

3) It increases the amount of information on the page and so increases the chance you'll find exactly what you want.

Bear in mind that while the Hacker News search is currently powered by Google's site specific search, most of the other sources WebMynd provides use the sources' own search engines. Which means they may include results which Google does not readily surface.

I hope you do get a chance to try it out!

I haven't tried WebMynd yet (will do so). But, from the FAQ,

"Some people call the WebMynd browsing history a DVR for the web - it saves and records the pages you see online so you can find what you have seen before faster:"

This sounds very similar to what InfoAxe is trying to do. Has anyone here tried InfoAxe? How different are these services?

I get it; it acts like a client-side filter/index, my fault for jumping to conclusions.