I'm not impressed so far. A search for "linux find and replace multiple files" returned zero results. On google it returns 280,000. Further reducing the query to "find and replace multiple files" and "find replace multiple files" still return zero results.
The relevance is completely awful. I did a search in Cuil and ended up with a serp where 6 of 10 results were for the same site and the rest were comprised completely of ridiculous sites like one-page sales-pitches.
At the outset, I'm wary of any company whose primary claim is that they'll beat <current market leader>. If they're truly in a class apart, why would they need to rely on a stunt like that? "Look, we're related to <current market leader>: we'll be the ones to beat them. We'll do exactly what they do except better."
I think that's just a bad way to frame a company's mission. It looks arrogant and defensive. Show, don't tell.
I have a personal litmus test for new search engines - a vanity search on myself. I know my own content across the web better than anyone else's. By far, Google still rates far and above for returning the fullest and most pertinent set of results than any others. Cuil fails at this.
More generally, I compared common searches I do regularly relating to things like "javascript framework" and "css" and the results were poor.
Its not only about who has the biggest index. Its about retrieving the most relevant data from that index.
To be more crude, its not about the size of your <guess>, its about how you use it.
Google has had a lot of time and resource to index your content. So it is not a fair compariosn yet. Give Cuil time to develop/index and then see the results.
On a side note, Cuil takes forever to load compared to Google, and search results aside, that is a major issue they need to address asap.
I don't think we ought to excuse Cuil and provide them more time before we judge the search engine. They've had all the time they needed and put out a press release saying so.
Same litmus test, got nothing. Tried a couple of other topics I know well, results weren't particularly good. The presentation is moderately interesting though.
I didn't mean the layout so much as that the description of the hit doesn't necessarily come from the text of the page. For instance, the description of kde.org is pulled out of Wikipedia. I'm not sure that approach works, but I think it's an interesting path to flesh out.
It seems they are rather restrictive with their results. I tried it with "what is cuil?" and didn't get a single answer.
Also I don't really expect to get that much improvement from the typical one-shot-editbox searchengines anymore. There's only so much information you can get out of a handful of words (that is unless they start to actually understand the intend of questions and content of pages).
I suppose results could improve if the searchengines would enter a session or dialog after the first try. Some way to make clear that you're still searching the same stuff, only trying other words because the first result haven't been good enough. I think when I fire off several searches in a row far more than half the time the reason is that I'm trying to improve my results. Cuil's categories are a step in that direction.
My first search resulted in a 404 page several seconds later. Then.. it went totally down for a few minutes. Now, it redirects everything to http://www.cuil.com/info/unavailable.html
Interesting first impression! Google didn't have this trouble, but then back in 1999 we didn't have billions of Webheads flocking to whatever was new and shiny... that was the investors' job ;-)
these are bad:
- relevance of search results
- colors
- layout of search results
- specialisation for different types of search results is no where near where google is.
- some searches for things I know about returned 0 results.
Also... frames?
If they think google hasn't been updating their search engine for ages they are kidding themselves. Google has done very well to keep the design consistent for about a decade, but things are changing all the time.
Vanity searches can work well for a test but I do not think the index is big enough yet for a real comparison. I would say wait to see what happens once they are 1 week, 1 month, etc... out of the launch gates.
1) Make the search box on main page bigger and move it up a little bit.
2) I want larger navigation page numbers and icons. Center it once again. Maybe put page numbers on top as well.
3) Make safe search on/off a toggle switch. I don't want to go to another page and then hit save etc.
4) black/blue motif just hurts my eyes.
5) Ditto with the font
It's amazing how ingrained searching Google is for me. When I want to advance the search page, I instinctively move my cursor to the bottom right of my screen. Whenever I search on Yahoo or Live the presentation of the results just seem visually painful.
