What is this real-time search hype?

[pavlovapete]

Why all this hype about real-time search?

Goolge is pretty quick to index stuff - what is the reason people are talking about rt search?

For example, I'll bet this post will be indexed within minutes.

[tedster]

I think the big deal is Twitter, where the tweets flow by like rushing water, and some people feel a need to be right in that stream.

[pavlovapete]

thanks tedster

I don't understand the utility of searching a stream. Is this like some kind of self-refreshing AJAX news ticker?

Maybe I should join twitter to get a better idea of this kerfuffle.

(and this post isn't indexed yet which is a bit of a disappointment - c'mon G :)

[pavlovapete]

I stand corrected this post is indexed already.

Not bad G

[tedster]

If you haven't been following the growth of Twitter, then I understand why you are a bit perplexed. Suffice it to say that politicians and celebrities are using it, and sometimes news will actually break FIRST on Twitter. If you want to have a scoop on your site, following Twitter can be a valuable tool.

Also companies of many sizes are using Twitter to do real-time customer service. They intervene on complaints before they mushroom into a ground-swell of negative buzz that trashes their reputation.

When I attend various conferences, without Twitter I would miss knowing about real-time changes in meet-ups, new parties and so on. There is practical value (and poterntially $$$) in RT Search.

Yes, Google indexes things pretty fast - but is it fast enough to compete in this area? And even if it is, does the average person or company know that they are doing this?

[edited by: tedster at 6:27 am (utc) on June 17, 2009]

[Lexur]

is it fast enough to compete in this area? And even if it is, does the average person or company know that they are doing this?

No. Google isn't capable to compete there but Tweeter can't compete with the incredible efficiency of a microphone and a pair of speakers in a conference or in the hall of any hotel.

Regarding the average user, he don't cares where the information comes from but how good the results are. Google spider will visit the most updated sites a few times a day and it means you will find this post (news in online newspapers too) once you wake up today in the States (sunrising right now in Spain). I think a six hours gap is quick enough for average user because most times he will find what he was looking for.

[pavlovapete]

Well my vague thoughts about rt search is that the most computationally efficient place to do the search is in the website

database. So the best person to do a rt twitter search is someone who is connected to their db (inhouse of api)

However, when several sources/ sites are being searched then you need a crawler who is hitting the sites repeatedly in short

intervals. (Or is plugged in through a number of apis)

Google has obvious capacity here. They'd simply need to hammer the sources and voila - you have near rt across a number of sites.

Obviously I say near rt - unless you are plugged into the site database I can't see how else you could do real realtime.

I take your point about conference changes tedster - for me though this gets blurringly like subscribing to a feed or a thread.

<edit> spelling