Google's www search is successful for three reasons. The size of their database. The efficiency of their servers. "Page Rank"[tm].

A product that indexes an intranet can take advantage of the technology they've already developed, to deliver (2) and (3), but the intranet would have to be rather large for Page Rank to be noticeable, I'd think.

A product that indexes your hard drive would be entirely new. They'd have to build it from scratch, and the only existing technology they'd be able to leverage is pretty banal stuff - indexing different file types. Files on your hard drive aren't linked together, so Page Rank wouldn't work. The application (and its indexes) would have to be optimised far more for size than for speed - I imagine that's the opposite of what google do now.

And, of course, Microsoft are allegedly going to make the filesystem of the next Windows version a database, which (I assume) could have full-text indexing plugged in in the same way you would any other RDBMS. Now that -would- be taking Google down the same track as Netscape.

So really, Google have nothing to leverage in creating a localhost-searcher except their name, and a product-lifespan that's likely to be over before the technology matures. Don't you think it's better to keep doing what they know they can do, and what they've proven they can do?

Perhaps Dave should look at Apple again. Introduced with OS 8.5 was Sherlock, which let you not only search the Internet from a desktop application but also allows you to index the contents of your HD (like the text within Word documents) and search it. Granted, indexing a HD is a PITA and takes too long, but if you've never used Sherlock 2 to search a HD full of data, you're missing out on perhaps one of the most-overlooked technologies in Mac OS 9 and Mac OS X.

Perhaps Dave should look at Apple again. Introduced with OS 8.5 was Sherlock, which let you not only search the Internet from a desktop application but also allows you to index the contents of your HD (like the text within Word documents) and search it. Granted, indexing a HD is a PITA and takes too long, but if you've never used Sherlock 2 to search a HD full of data, you're missing out on perhaps one of the most-overlooked technologies in Mac OS 9 and Mac OS X.

Perhaps Dave should look at Apple again. Introduced with OS 8.5 was Sherlock, which let you not only search the Internet from a desktop application but also allows you to index the contents of your HD (like the text within Word documents) and search it. Granted, indexing a HD is a PITA and takes too long, but if you've never used Sherlock 2 to search a HD full of data, you're missing out on perhaps one of the most-overlooked technologies in Mac OS 9 and Mac OS X.

Perhaps Dave should look at Apple again. Introduced with OS 8.5 was Sherlock, which let you not only search the Internet from a desktop application but also allows you to index the contents of your HD (like the text with Word documents) and search it. Granted, indexing a HD is a PITA and takes too long, but if you've never used Sherlock 2 to search a HD full of data, you're missing out on perhaps one of the most-overlooked technologies in Mac OS 9 and Mac OS X.

"All our products are powered by our patented DexTM technology, which automatically creates a master cross-referenced index of all of the information on a user's PC and across the enterprise - including Internet references."

They're selling to both desktop and enterprise users; I think they have some performance issues, but the ideas are interesting.

The larger question of selling through the front or back door (enterprise vs. personal corporate customer) is actually an incredibly interesting and often difficult one; and one we've struggled with at length with the product I work on. One one hand, many people sitting in cube farms at larger corporations don't have a lot of discretionary spending power, and face IT-imposed obstacles like locked-down desktops, etc.

The IT/CIO level usually holds the pursestrings, but their needs are often quite different from the average user.

To me, Google's approach to the enterprise market makes a lot of sense; it's simple to deploy, and is more likely to succeed -- be more profitable -- than a desktop version ever could.

"All our products are powered by our patented DexTM technology, which automatically creates a master cross-referenced index of all of the information on a user's PC and across the enterprise - including Internet references."

They're selling to both desktop and enterprise users; I think they have some performance issues, but the ideas are interesting.

The larger question of selling through the front or back door (enterprise vs. personal corporate customer) is actually an incredibly interesting and often difficult one; and one we've struggled with at length with the product I work on. One one hand, many people sitting in cube farms at larger corporations don't have a lot of discretionary spending power, and face IT-imposed obstacles like locked-down desktops, etc.

The IT/CIO level usually holds the pursestrings, but their needs are often quite different from the average user.

To me, Google's approach to the enterprise market makes a lot of sense; it's simple to deploy, and is more likely to succeed -- be more profitable -- than a desktop version ever could.

>>What I think Dave is talking about (help me out here) is Google's power turned towards indexing files on a local hard drive.

Yes, that's what I want. I have all this data on my hard disk, and ironically it's harder to find stuff there than on the Internet at large.

My dream product is an integration of Google's www search with search of my local resources. When I search for something on Google, the results should be intermixed.

Now I know this is a huge technical project, but I am so in awe of Google engineering that I think they could pull it off. It's the next thing after they put a Web Services interface on the centralized Google so they can index news sites in real time without polling. I still have to write that up, but I've been banging that drum for a few years already.

Anyway I'm rambling. Thanks for the great coverage of Google Appliance.

>>What I think Dave is talking about (help me out here) is Google's power turned towards indexing files on a local hard drive.

Yes, that's what I want. I have all this data on my hard disk, and ironically it's harder to find stuff there than on the Internet at large.

My dream product is an integration of Google's www search with search of my local resources. When I search for something on Google, the results should be intermixed.

Now I know this is a huge technical project, but I am so in awe of Google engineering that I think they could pull it off. It's the next thing after they put a Web Services interface on the centralized Google so they can index news sites in real time without polling. I still have to write that up, but I've been banging that drum for a few years already.

Anyway I'm rambling. Thanks for the great coverage of Google Appliance.