Review: Online databases let you structure and share your data

Dabble DB

Editor's note: Dabble DB is no longer in operation.

Smallthought Systems Inc. obviously understood a few basics when it developed Dabble DB: first, that databases are not spreadsheets, and second, that if it's too difficult for a user to create multiple tables, then the users might as well just code their own from the outset. In fact -- and I say this as a longtime MySQL fan -- using Dabble can be cooler than coding your own.

Many database applications can import an Excel worksheet or CSV file into a database table. Where Dabble DB shines is in importing multiple tables and then setting up relations among fields in different tables after the import. After less than half an hour (including watching an 8-minute demo video), I had imported my data and automatically set up relationships between what were separate Excel sheets.

I can't imagine the process being simpler. All I had to do was go into one record, select the configuration option for the field I wanted to link, and choose "list of entries." This created many-to-many mapping. (There's another option, "link to entry," if each record only needs to map to a one entry in another table.) That change in one record redesigned my entire database structure and automatically mapped all my existing data.

Cleanup on 250 records was minimal. In fact, Dabble DB did some cleanup for me, since (in a fit of laziness) I'd made some entries single product records in the product spreadsheet, even though they contained multiple products -- for example, one entry read: Ajax13, Google Docs & Spreadsheets, ThinkFree Office Online, Zoho Office Suite. By telling Dabble to split the list by comma, it automatically created four separate entries.

Adding and editing single forms was truly elegant. For example, if you create a "selection" field that has too many choices for check boxes or a drop-down list, Dabble automatically creates an AJAX-based search box. The user just has to start typing in the search term, and matches will start to appear

Other pluses: Dabble made it easy to drag and drop fields from my table to create a Web-based data-entry form. And the site allows scheduling automated imports to update data from a Web site or RSS feed.

Default viewing options include charts and maps as well as text, although the mapping isn't as sophisticated as Google Maps. For example, when I imported a spreadsheet of 2007's Computerworld Best Places to Work list, it correctly tallied companies by state when the headquarters field included the state (e.g., Coral Gables, Fla.), but not when the city stood alone (e.g., San Diego -- although clicking to see the Google Map option correctly found San Diego in California).

Dabble DB isn't perfect. There are robust grouping and sorting options when viewing within the application itself, giving the application many capabilities of traditional databases. However, once you export a "view" to an external Web site, that view is static and no longer sortable by column header. This is a serious limitation for those seeking to create even moderately interactive data applications on their own Web sites.

Another limit: no scripting capabilities, such as automatically e-mailing someone when a record is updated.

And if you use the free service, you'd better not be playing with private data. With a free account, you can limit your account access to enter/edit/delete data by password, but the data is available for public viewing under the Creative Commons license. Anyone can view your data and the underlying application structure. You have to have a paid account (plans start at $10 per month) if you don't want your data publicly accessible.

Ultimately, I found Dabble DB a useful and cool tool for some purposes, but like any low-cost Web-based application, it has limits, particularly if you need scripting or want a robust public application to embed in a Web site. However, if you want a group to view and interact with data within Dabble DB itself, it's a decent option -- and a free one, if you don't mind all that data being public.