- Friday, 27 November 2009
- INQ Mobile
- RSS
The dross always floats to the top in IT companies - A tall Scandinavian
Gmail goes IMAP
EVERYONE'S FAVOURITE data-mining conglomerate and incidental search engine provider, Google has managed to get another one-up on other webmail providers.
Not content with increasing the storage capacity of Gmail accounts to a whopping 4 gigs, the outfit is also busy rolling out IMAP access to the unwashed masses.
This is unlikely to be of interest to many Gmail users, who are content to venture no further than Gmail's moderately friendly web user interface. But this feature will be a total godsend to anyone who prefers to have access to their email from alternate mail clients such as Thunderbird. These will now be able to access their entire Gmail mailbox from multiple clients in a far more convenient way.
This means that Google will have less opportunity to serve ads to those people who've been champing at the bit for such a feature. It's possible that Google's execs feel that once users have a Gmail account they'll end up on their websites often enough that it's worth the extra cost of the providing the additional service.
But the smart money is on the idea that the more comprehensive Google's enormous email database is, the more value it can extract from statistical analysis of it.
Whatever the motivation, the extra convenience of IMAP access is likely to win over a few more converts to Gmail. And probably spawn a few more paranoid conspiracy theories to boot. ยต
Share this:
ShareBusiness Intelligence Software
Customer Relationship Management (CRM)
Enterprise Accounting Software
Enterprise Resource Planning (ERP)
Enterprise Systems Management
To Niki Mistry (previous commenter):

Niki you missed a point. The article talks about IMAP4 protocol, not POP3 protocol.

They are two different (very different) protocols to access your email.
Sergey
"But the smart money is on the idea that the more comprehensive Google's enormous email database is, the more value it can extract from statistical analysis of it."

Trust me there is a no difference if you have 100 000 or 100 000 000 users. If they behave more or less similary. Obviously if you want location(or gender) specific info(like what people in Washington chat about) you need more than 50 people from Washington to make usable info, but offering mail only to analyze mails is not smart.
as google allowed you to access gmail / googlemail by pop/smtp for ages.
For anyone who uses gmail from more than one computer this is great! We will be able to use other mail clients to view our mail while LEAVING THE MAIL on the mailbox. 

Yes one could leave the mail in the mailbox with pop, HOWEVER the benefit with IMAP is it will sync whatever is done in the email client with gmails. IE delete, read etc. 

So for someone like me who gets a ton of email a day I could read it at home using outlook, thunderbird, evolution, ..., and then when I got on gmail email that I have read would still show up read!
Good. Now I don't have to use Gmail's obnoxious and inferior search system for their mail. I can search using substring matching instead of whole-string matching. It's amazing that Google doesn't even provide adequate search capabilities, or even the basic ability to sort messages by fields, in their email system.
Zoidberg,

I work in forecasting and data pattern analysis and I can say you are wrong.

A data pattern can emerge with 100,000 users and the same pattern can exist with 100,000,000 users. In this case the size of the pool may not matter.

You fail to realise that the additional 900,000 can introduce new patterns which you couldn't find without them being part of the pool.

Data Monkey
"You fail to realise that the additional 900,000 can introduce new patterns which you couldn't find without them being part of the pool. "
Well they CAN. I just said that it is unlikely that it will. If you can provide me with a link or something i would be happy to admit that I was wrong. You said that you work with forecast... IMHO people are easier to analyze than forecast so that is why I belive that 100 000 accounts will give you more or less the same amount of info as 5 000 000. 
very simple eg:
you play with a "broken" coin . In the first 3000 experiments you got 1600 heads. It is pretty safe to assume that p(head) is in very small interval around 53.3%. You don't have to repeat the test 15 000 times.