The name probably caused them more trouble than it should have. It's the same criticism being thrown at Bing (the idea that Google is a verb now, and you need a name that works well for that).
I can't just tell my friends "hey, try out Cuil", I have to spend a few seconds explaining how to spell it. That's no way to get the word out.
The technology is good, the site is very fast and relevant. But there's a lot in a name.
When 'CUIL' first came out their crawler was all over my servers, sometimes to the point of making them inaccessible for users, I remember having a devil of a time getting rid of them because they came from lots of different IP blocks.
Apparently their bot is called twiceler, maybe they should name it 'CuilBot' or something to that effect.
Judging by your response, maybe they shouldn't name it Cuilbot.
I wonder if their overzealous bots and subsequent blocking are reasons why they launched with such terrible results. I heard several reasons for that catastrophic failure:
* Overstressed subject-specific servers failed, leading to
porn on hamburger searches and that kind of thing
* Crashing, random result variations due to overload
* Excessive media hype exacerbated the load issues
making Cuil the neighborhood store stocking the hard to find antique and offbeat books.
Cuil's design is too slick to pass as a neighborhood book store. It needs more rough edges and needs too appear old and established. Instead it feels young and fresh and naive.
when i looked at this, i wasn't excited or envious of all the geek gear, i just thought it represented a massive waste of resources. they took their shot, they flopped...any time/mindshare leftover for google alternatives now belong to bing and players to be named later (blekko etc). to continue to pour money into hardware and development just seems like a massive waste. just give the money to an inner city school...actually put that wealth to some positive use for humanity
The name is silly, but who cares. The search results just aren't very good. My impression is that the ranking algorithm is passable, but that they just don't have nearly the depth of Google. Less popular but highly relevant pages that come up in google are completely missing from cuil.
I'm still awaiting the day Cuil does phrase queries. I know, most casual search users don't do quoted-phrase queries -- but Google and Bing do, and I occasionally use them, and Cuil has no chance of being my default search engine if it doesn't take the trouble to support them, too.
I'm sorry, but 7 petabytes for the largest search index in the world? I'm hoping the reporter misquoted the developers here. If we conservatively assume servers with half a terabyte of hard disk space, that's 14,000 servers for the web index. Google was estimated to have 450,000 servers in 2006, and presumably has many more than that now. So less than 5% of their server stack would be required to store that size of web index. Considering that web search is still Google's core product, this seams easy to believe. And much more if you include image caching, maps, videos....
17 comments
[ 3.0 ms ] story [ 55.4 ms ] threadI can't just tell my friends "hey, try out Cuil", I have to spend a few seconds explaining how to spell it. That's no way to get the word out.
The technology is good, the site is very fast and relevant. But there's a lot in a name.
As for CUIL, their crawler was such a nuisance at some point that I blocked their complete network. I doubt I'm alone in that.
And if those 'twiceler' guys don't start playing nice I'll do the same to them.
That clinches it then. One more agent for the ban list.
Apropos, I think that plenty of bad guys are masquerading as twiceler as well.
Mozilla/5.0 (Twiceler-0.9 http://www.cuil.com/twiceler/robot.html
When 'CUIL' first came out their crawler was all over my servers, sometimes to the point of making them inaccessible for users, I remember having a devil of a time getting rid of them because they came from lots of different IP blocks.
Apparently their bot is called twiceler, maybe they should name it 'CuilBot' or something to that effect.
I wonder if their overzealous bots and subsequent blocking are reasons why they launched with such terrible results. I heard several reasons for that catastrophic failure:
* Overstressed subject-specific servers failed, leading to porn on hamburger searches and that kind of thing
* Crashing, random result variations due to overload
* Excessive media hype exacerbated the load issues
Another one to add to the list?
Cuil's design is too slick to pass as a neighborhood book store. It needs more rough edges and needs too appear old and established. Instead it feels young and fresh and naive.
Their wooden pallet/cables hanging over table system seems like every Sysadmin darkest nightmares...