Google Desktop Replacement: Cigarbox of Ideas

by T. R. Halvorson January 7, 2012

This is a cigarbox of ideas to replace Google Desktop.

Google discontinued Google Desktop. What to do?

You can still get it, if you like, at Old Apps.

The Home version of Copernic Desktop Search is free for certain uses. Free is nice. Copernic has been at this a long time. Worth a look.

dtSearch has a desktop edition. dtSearch sells components to application developers for implementing seaching in many different contexts. It is almost certain that you have used dtSearch back ends without knowing it.

The beta of x1 Pro 7 now searches, besides local resources, Facebook, Twitter, Gmail, and other webmail services. Hmmm. do I need that? Will feature creep weigh down the application? They claim it's fast, as fast as you type. Could be good for forensic searching of the dumb stuff people say in social media.

DocFetcher could do for me. I don't always go for the most full featured solution. Feature creep makes that over-featured. DocFetcher uses Lucene, which is a fine back end. nCodian uses Lucene in nCodian FreeText. If its implementation and front end are good, this could do. It's at SourceForge and it has a wiki. It is under active development, and it has a fair number of awards.

Another Lucene implementaton is MaxFisher. It implements through one's web browser, which many people might like. This one is in pre-alpha.

Because I sometimes go on scripting binges, I might try the technique developed by CheeserTheAverage.

Windows Search that ships with modern Windows versions is not so bad as its predecessors. For example, its Advanced Query Syntax  supports Boolean operators. It supports natural language searches. For encodians like me, the Windows Search index can be accessed programmatically using both managed and native code. Moth to a flame.

If you want to update Windows Search for XP machines, you can get Windows Search 4.0, KB940157. Check first to see whether updating has already installed it.

Keywords:

Filed Under: Cigarbox | Desktop Search

Searching for the golden mean between free form and heirarchy

by T. R. Halvorson January 6, 2012

How do I want to organize information?

Sometimes I go on binges of putting order to it. An example is folders and subfolders of emails. The trouble is, the more and more order I impose, the less usable the information sometimes becomes. I can develop such an involved hierarchy that the order makes the data irretrievable. Hierarchy is very good at getting information that, today, is not my priority out of my way. But it's not so good at letting me find it when it has become my priority. Then I wonder if life wouldn't be just as easy had I left the email in its pristine, chaotic state.

The trouble with chaos, though, is that it is like a teenager's bedroom. The data I want is in there, for sure, but where?

MemoryMate vintage manualThat's why I like free text information retrieval systems. Back in the DOS days, I said, "Computer crime is selling a computer without a copy of Memory Mate." What a beauty that was. Originally produced by Atlantic Coast plc (Brownbag Software), Broderbund sold it for awhile, and that's where I got it. Later Broderbund quit selling it, and Brownbag created a Windows version. See another blogger's summary here. The Windows version is okay, and I do use it, but am not as wowed by it as I was by the DOS version.

At nCodian, we've created a few free text retrieval applications for the Windows desktop from scratch. That has been some of the most fun I've had as an encodian, and it made me a better database searcher. In later years, however, we were reinventing too many wheels. We opted for using the Lucene.NET engine, and we placed standard Boolean syntax at the front end. That's our nCodian FreeText application. (It also has word, phrase, and Query by Form modes. A database can be searched in four modes from novice to expert.)

I use FreeText as the teenager's bedroom for such things as email, but a bedroom in which I can find what I am looking for. Free form text retrieval systems let me avoid bingeing on imposing hierarchies in the first instance, and then finding the hierarchies wanting in the last instance.

But sometimes I still go on email folder reorganizing binges. Still searching for the golden mean.

Perhaps the reason the oscillation between approaches keeps happening is that both approaches share one very important trait: They help you know by helping you ignore. It may seem counter intuitive, but knowledge is founded on ignorance. I'll see if I can explain that another day.

About the author

T. R. Halvorson

President of
Synoptic Text Information Services, Inc.

View T. R. Halvorson's profile on LinkedIn

Month List