Microsoft Offers A Peek At New Search Engine

ObsessiveMathsFreak writes "The Inquirer is reporting that Microsoft is offering a preview of its new search technology. The search engine preview has a minimalist interface, similar to Google. Microsoft claims over one billion web pages searched, but admits the fact that searching is a little slow. This technology hasn't yet been incorporated into MSN Search, though the site claims it eventually will be. In related news, the Financial Times is reporting that Microsoft are to improve the regular MSN Search site by removing paid advertisements from regular internet searches, a move that will cost them 'tens of millions of dollars.' Are the Search Engine Wars finally upon us?"

Search for Linux

[grandmofftarkin]

Redhat, SUSE, Mandrake, etc. don't turn up on the first page. WTF?

Oh I get it. Microsoft don't want their competitors turning up in search results.

Hmmm ... I think I'll stick to google.

Re:Search for Linux

[SimplexO]

A search for 'search' returns a list of websites that all contain the word search in their domain name, none of them are any of the good ones - Google, AV, Yahoo, and (even!) MSN.

Hrm...

Shhh... secret Microsoft code!

[AKAImBatman]

function search(var terms)
{
return screen_scrap_google(terms);
}

On a more serious note, this engine is not all that great. For example, if I do a search for my name I primarily see forum posts. While that's great, it's not very useful. Listing software I've written, articles that mention me, etc. are far more useful results, especially to a future employer. Google nails this because the more interesting stuff always has back-links to it. Sorry Microsoft, you need Google's technology to compete with Google.

Their indexing is waaaay out of date

[Anonymous Coward]

Searched for my domain, found results pointing to web pages that I haven't had online since November of 2002.

Boycott

[OYAHHH]

Anybody know what the search engine spider calls itself.

I make sure that msnbot is not allowed to traverse my web site via the robots.txt file. I'd like to do the same with this robot.

BTW, I've noticed no appreciable decline in web hits at all.

Re:"fisherman"

[bje2]

i randomly searched for "baseball" on both... Microsoft's #1 results was the mlb.com website (which is what i would expect)... Google's #1 result was baseball-almanac.com doesn't really mean anything, just thought that was interesting...

Same algorithm

[baywulf]

Would this imply that Microsoft is using the same algorithm as Google?

to answer their question

[CoffeeJedi]

What do we have under the hood? A brand new algorithmic search engine -- built from the ground up -- on Microsoft technology. Give it a spin and tell us what you think.

This site will be available for a short time. After we have time to incorporate your feedback and add some features it'll be back, new and improved.

what do i think? i think you should stop having the marketing-drones right your copy, when Google says things like "give it a spin" or "it'll be back", it seems genuine, when Microsoft says it, it sounds forced and derivative of the original

Re:Search for Linux

[awhelan]

MSN was famous for this. It is probably a coincidence since I doubt they'd dare do it again, but a few years ago Microsoft was accused of intentionally blocking certian linux websites from search results.

Re:Playing catch up

[strictnein]

Very interesting: Drilling down using NetScan and you actually get a page that links to Google Groups.

MESSAGE HEADER

The message header for this article is not available.

MESSAGE BODY

The message body for this article is not available.

Message-ID: 409e3b22$0$29341$[e-mail]

Sear ch for this message at Google Groups

Kind of suprsising

Re:Search for Linux

[betelgeuse-4]

Search for "microsoft" and only 1 result is returned, compared to Google's 103 million.

MSN Bots = Annoying

[Bruha]

I'm getting more hits than most of my regular visitors in my logs.. even google does not check my site on a daily basis but these seem to do it on a semi hourly basis.

Re:Search for Linux

[WebCrapper]

The sad part is, MSFT is just using the Slashdot crowd as real life load testing...

MSN Bot Inefficiency

[Anonymous Coward]

MSN Bot is highly inefficient at indexing large dynamic sites.

One of my sites [futplex.com] is a dense, heavily cross-linked statistics search engine. Over the last few months I've noticed MSN Bot making a spirited attempt to pull the entire site. Since the number of unique pages (without a custom search) exceeds 100,000 it was generating enormous amounts of traffic [futplex.com] until I banned it a few days ago.

Considering that the site gets very little user traffic and many of its pages are similar, I wouldn't expect anything like the level of effort expended by MSN Bot. By comparison, Google only indexes the first few pages.

they need to stop. just stop

[scaaven]

I don't care if it is being /.'ed, it's still slow as Bush. Also seems like there's not alot going on under the hood (hey,also like bush). but seriously.

MS has alot of really dumb execs who get their panties in a wad if they don't feel they're "on top." All this search engine can do is hurt them. Time and money will be wasted in a cheap imitation of google that nobody will use except the newbies who can't change the default page of IE. and why? because their egos are as bloated as their software.

Now you see it, now you don't.

[mccalli]

Did a search for my domain, Eruvia. It told me there were fifteen results, then gave me the first ten and and a next button.

Clicked next - suddenly, "Sorry, there a no results for 'eruvia'". Pardon? Try searching again from scratch, and once again suddenly Eruvia has disappeared. Can't get my original 15 results back at all.

I'm putting this down to extremely ungraceful load handling and the ongoing Slashdotting. What's this running on anyway? Netcraft says IIS/6.0 on Linux [netcraft.com], so it's another Akamai job, but the hardware behind it doesn't seem up to scratch yet.

Cheers,
Ian

Search for Open Source returns NOTHING!!

[Monoman]

http://techpreview.search.msn.com/results.aspx?q=o pen+source&FORM=SMCRT

Re:Search for Linux

[iocat]

It's not just wikipedia...

Sorry, no results were found containing "rabbit"

(Google found 6.8 million, fyi).

At what point do the search engine creators just decide that their engine isn't ready for prime time? I mean, a rabbit is a fairly common thing.

Interesting results for "Google"

[TopherTG]

Try searching for Google. I half expected it to not show up at all, but it comes up as the first link. The second link is an antique site, which the contains no reference to Google or even the text "Google".

Re:Search for Linux

[sponge_absorbent]

hmmm, it didnt seem to find any results for "microsoft sucks", despite the fact www.microsoftsucks.com exists... how strange

First impressions -- can't find much

[Mad Bad Rabbit]

Even ignoring the slowness, the results are extremely lacking
compared to Google. Here's a few quick searches I tried:

gethostbyname - no matches (Google=161000)
starbucks - 1 match (Google=1500000)
spiderman - 1 match (Google=2550000)
cassini - no matches (Google=941000)
gucci - 1 match (Google=3340000)
"garner state park" - failure (Google=3190)

Looks like it will find roughly /2 million times fewer/
matches for a given keyword of Google. I don't think it's
quite ready to enter ship-test just yet.

Re:Boycott

[OYAHHH]

I,

Think you hit the nail on the head with the word "why".

What is the point of MS providing a search facility via their websites and eventually tied directly into their OS. So they can make more money. Yeah a secondary concern might be to help their customers have a better experience, but believe me, it's to make money.

Has MS been particularly honest about their search results in the past. I've done the "Linux" search on MSN before and quite frankly it was appalling to me that a company could be so blatent in their disregard for truthfullness.

Thus, "Why" should I provide any opportunity for MS to make any additional money. I certainly don't feel the need to line Bill Gates pockets anymore.

And, I certainly don't feel the need to give sustenance to someone who has made it their life's calling to squash anything that is non-MS.

Call me vindictive and I'll wear it like a crown if you like.

To me, collaborating with MS in any manner is sorta like having the person who shot and killed your mother over for dinner. Ain't gonna happen under my command.

By disallowing the msnbot I effectively, in a small way, reduce MS' ability to rule the world completely.

And as I noted in my first post, before I disallowed msnbot from my site I only got a few redirects off of MS sites daily anyway.

If somebody wants to find my site they can readily find me on any of the other search engines. My site's consistently in the top five returns for the pertinent subject matter on any search engine. Except MSN, of course, hee! hee! hee!

It's a bug

[uss_valiant]

First, I did a search for "nei" (because it's part of my domain name). The search engine presented several pages of matching results.
Then, the search engine had an error during a request and I did a new search for the same word, "nei".
And the search engine answered with:
Sorry, no results were found containing "nei"
So it's a bug, because I got thousands of results when I searched for "wikipedia".

Re:MSN Bots = Annoying

[underpar]

For a while I was wondering if maybe the poor little MSN bot on my site was lost or something. It stayed for hours, but MSN still has a month old version of my site cached. Google bots are sooo much smarter.

try xfree86

[cyfer2000]

and the search engine said

You have entered a search term that is likely to return adult content.

Warning: If you are under 18 or live in an area where it is illegal for you to view explicit content, please revise your search.

More FUD... oe something like that

[poptones]

MS kicked Netscape's ass because Netscape fell into a spiral of devolution while IE became a much better product. In the fight between Netscape 3 and IE4, it was no contest because IE was, at the time, simply the better product. Remember, this was when Netscape was just sure it owned the web and, at the time, it was even thumbing its nose at the W3C. This is perhaps the most perfect example of Microsoft's "embrace and extend" philosophy. Once MS becomes dominant in a market it's very easy for themn to retain that dominance - but attaining dominance in any particular field is NOT a given even for MS.

So compare then to now: you can't even get decent fucking search results of Microsoft's own support site from Microsoft itself. They can't even properly track their own content - how on earth can anyone trust them to track everyone elses? I work tech support a few days a week and I don't even think about using Microsoft's "search the knowledge base" page - it's often laughably short on search results even for well known things like "xp rpc exploit" and "download ie6."

When I can get proper tech support info on Microsoft's own products without having to go to Google and type site:microsoft.com THEN I'll start to believe this is like Netscape vs. Microsoft.

Feedback

[flossie]

"Please provide your feedback in the form below. Thank you for helping us to improve MSN Search. Although we won't respond directly to your comments, we will use them to improve our service."

Are they being choosy in deciding from whom they will accept feedback? When I click the "submit" button (using Firefox on Linux) nothing seems to happen except that an error message appears in the Javascript console: "Error: changeImage is not defined".

Re:Microsoft's gonna lose this one

[glinden]

I don't think it's true that "Google won the mindshare a long time ago." As of Jan 2004, Google has less than 40% [searchenginewatch.com] of the search market, nearly tied with MSN and Yahoo.

Unfortunately, all Microsoft has to do is to catch Google. If the quality is essentially indistinguishable from Google, most people will use MSN Search, since MSN Search will be the default in IE (and probably MS OFfice and WinXP soon).

As long as Google keeps innovating [blogspot.com] and stays ahead, they'll do fine. But, if they trip, Microsoft will catch up and trample over them, just like they did to Netscape.

Re:Search for Linux

[ad0gg]

Its not the domain name, its the text between the anchor tag that makes a difference. You can see it with the "French Defeats" search and other humor searches where the search term isn't even the domain name or the page itself.

MSN beta search works little like google's page rank where inbound links are counted toward search placement, but it looks like msn places lot more emphasis on it. Problem with this, it breeds link farms and spamming the results.

Poor search focus, or strange ordering of results

[Simon Brooke]

I have a Java class, ShuffleWidget [weft.co.uk], which I often use when assessing search engines because its name is unique and so all the results tend to come from my site (or places that refer to my site). Plugging this into the new MSN search [msn.com] does, indeed, bring back pages from my site, but, interestingly, not the ShuffleWidget page itself, despite the fact that according to my logs, "msnbot/0.11 (+http://search.msn.com/msnbot.htm)" [msn.com] has scanned the page on an almost daily basis throughout June. Google by contrast brings the ShuffleWidget [google.com] page back as it's second hit.

It seems they've some work to do in ordering their search results.

Re:Search for Linux

[Bozdune]

Good point. However, a lot of things that Slashdot trolls have been complaining about since '98 continue to be true today. How about the XP dispatcher? Still can't get smooth performance without a second CPU. Corbato et al solved this problem in 1961. There is no excuse.

MSN Search Bombs

[CommandoB]

Microsoft's new search engine seems to duplicate "Google bombs".

A search for "miserable failure" [msn.com] or "weapons of mass destruction" [msn.com] yields the familiar results.

Microsoft's new search technology, however, has brought fame to latest most "talentless hack" [msn.com].

So, critics of google bombing won't find any relief from Microsoft...

QGI? It's the quorumgroup.. quorumgroup.. quoru..

[phazethru]

I tried typing in 'QGI', the acronmy for an open source project I started/dropped a few years ago, as a test. The first 5 ranked pages come up and point to www.quorumgroup.ca. The first ranked page was index.html, and the next four were single clicks off of index.html. Seems a bit redundant. Maybe I'm just picky, but I think the point of a search engine is variety.

No cache mode ... also can't trust Microsoft

[minairia]

There's no cache option (yet). That's the feature I love most about Google, how I can get the information I want matter if a site has been taken down or is on some balky, slow server somewhere.

On an aside, the ulitimate combination would be in Google would buy Archive.org and you would be able to get a historical cache of every site on the web from the very beginning.

Also, I will find if very very hard to ever trust Microsoft to give me real, unbiased, un-pre-purchased search results. Google is equally a stockowner owned, megacorp, but (so far) they have shown a spirit of remaining honest and aboveboard. Microsoft definitely does not have this kind of rep ...