Live data from Hacker News

Ask HN: What do you think about my search engine?

news.ycombinator.com

1–6 of 6 posts

Ask HN: What do you think about my search engine?

#1
Tomorrowbook is a web search engine that displays results as an infinite list of favicons by default. However, the admin and domain owner have the ability to upload custom images/logos for domains and/or sub-pages "results".

Examples of search results for "Obama" and "News":

https://docs.google.com/file/d/0B6jnvIhOPJO7aWVNMDUwX3pGeGM/edit?usp=sharing https://docs.google.com/file/d/0B6jnvIhOPJO7cTk2N0NHWmtpUzQ/edit?usp=sharing

The goal was to offer a more visual display of results for children, elderly and mobile devices. Originally it used the Google API, but the number of queries on the first day catapulted it into the pay category - so it was rewritten using the Bing API which at that time was free with unlimited queries (at least I never saw query limit at that time - I do not know about now).

I have written about tomorrowbook in HN comments with somewhat favorable responses, so I am posting about it, because even though it folded after a year, with HN's insight I can still learn and grow from the experience. Maybe even get some motivation. I learned a lot already, for example my logo was just a bad idea period; after seeing The Social Network I thought when I got 220,000 queries on the 4th day after launching I would get someones attention - I did not; and I will not even mention the disaster of a press release.

Re: Ask HN: What do you think about my search engine?

#4
I'm lost as to what's going on here. I clicked your links and I read your description but I don't get how they are tied together. For example, how does your opening sentence have anything to do with the Google Docs links?

"Tomorrowbook is a web search engine that displays results as an infinite list of favicons by default."

Link #1 - to a jpeg

You then say "The goal was to offer a more visual display of results for children, elderly and mobile devices." Too vague - are you trying to offer a more visual display for the favicon next to the search results? Is that your "differentiator" I assume? I'm just so lost here.

EDIT: No one wants a "web search engine that displays results as an infinite list of favicons by default." People might want to view the favicon next to the search results but no one wants an "infinite list of favicons by default."

Re: Ask HN: What do you think about my search engine?

#5

I'm lost as to what's going on here. I clicked your links and I read your description but I don't get how they are tied together. For example, how does your opening sentence have anything to do with the Google Docs links? "Tomorrowbook is a web search engine that displays results as an infinite list of favicons by default." Link #1 - to a jpeg You then say "The goal was to offer a more visual display of results for c…

>I clicked your links and I read your description but I don't get how they are tied together. For example, how does your opening sentence have anything to do with the Google Docs links?

I would have posted a link to the live search engine, but as the OP says it "folded" after a year, so I did the next best thing and gave links to 2 screen shots of example search results pages. Sorry, I just do not understand why the description of my search engine leaves you lost, especially to the point you feel the description and actual screen shots are not tied together - the only difference between the description and screen shots is that because they are screen shots you do not have the benefit of continuing to scroll down and have infinite results populate, also most of the results are not favicons but custom uploaded logos (but I mention that function in the OP).

>are you trying to offer a more visual display for the favicon next to the search results? Is that your "differentiator" I assume? I'm just so lost here.

As the links show, my results display the favicon only - no text based results.

>EDIT: No one wants a "web search engine that displays results as an infinite list of favicons by default." People might want to view the favicon next to the search results but no one wants an "infinite list of favicons by default."

I asked for insight, but ouch. Did I shut it down for lack of traction? Yes, but starting from day 1 I had over 10,000 queries/day so I would not say "no one wants".

Besides I can build a search engine right now that people do want, that does not mean anyone will jump ship from Google for it. For example, I know from posts on HN that people have been waiting for years for Google to permit infinite results (see: http://news.ycombinator.com/item?id=4875463) instead of x results per page, so in minutes I could create a Google API search engine that provides identical results to Google and offers infinite results on the first page - but do you know how many people will use it? None, and for anyone who thinks it is as easy as building something people want, then test your theory now and build the infinite Google search engine that people on HN have posted about and see if building something people want actually gets people using your engine instead of Google.

Disclosure: I know you commented on the post I linked above about infinite scrolling, specifically that you do not like infinite scrolling for web search but you like it for image searches - a reply comment to yours was another HN user saying they prefer the infinate scroll feature of DuckDuckGo but find the actual Google results to be better quality.

Re: Ask HN: What do you think about my search engine?

#6
What is missing in the UI is that there is no context for why a result is relevant. There are multiple Yahoo, Facebook, CNN logos on the result page - how do I know which one I would be more interested in, without clicking on it?

Given you want to make search visually appealing, consider displaying thumbnail of a prominent image or video from the target page in the results page. However, although this will improve the UI, you will take a hit on performance unless you are crawling top news sources and caching image thumbnails yourself.