MetaFind – The Road Ahead

At some point, you have to stop adding features and ship what you have. But we’re not out of ideas, not by a long shot. Here are a few of the things we have in mind for upcoming versions:

Blocklists – this is a feature we wanted to include in this version, and in fact, it’s not that difficult to allow you to have the ability to save a URL to a list, and then prevent it from being shown in search results again. However, there are a variety of existing blocklists available, and we felt that the ability to import those was too valuable to be left out. That makes both the research and development tasks significantly more time-consuming, and forced us to postpone this feature.

Bookmarks – currently, MetaFind supports adding Bookmarks only for Internet Explorer and Opera. In general, managing Bookmarks for multiple browsers is a fairly complex task, so such a degree that Bookmark programs have become their own software category. We are committed, however, to including the ability to add Bookmarks to all browsers supported by MetaFind.

Filtered views – by applying a filter to existing Blended MetaSearch results, you will be able to restrict the results to those matching criteria not part of the original search. This will effectively turn one set of results into many. For example, you could tell MetaFind to show only the results that came from a certain group of domains, or had certain words in the title or description. Apply or remove as many filters as you like – each filtered view is in essence a new set of search results – without having to run a new search first. Combined with the Fetch Text command, this will give you the ability to perform complex searches within existing results, without having to wait for the search engines every time you want to adjust your criteria.

Image MetaSearch – applying the same techniques as Blended MetaSearch, this tool will combine the results of multiple image searches to produce a list of descriptions and thumbnails ranked by the consensus of the engines used. ImageFetch will then be supported – like TextFetch, this will visit selected source pages and automatically retrieve additional and/or larger images.

Faster display rendering for Blended MetaSearch results in Standard Mode – using a Rich Text control to display search results provides a great deal of flexibility when it comes to things like text searching and selecting multiple results. However, it also tends to be a bit slow to fill and resize when there are more than a few hundred results displayed. We’re exploring several alternative approaches, and will be changing this part of MetaFind once we’ve found a way to deliver increased performance without sacrificing flexibility.

More customization options for results in Browser mode – for example, optional search term highlighting.

Related/Suggested search terms – these will be extracted, combined, and available for an immediate one-click search, or for addition to the search target list for future searches.

Toolbars – browser toolbars for Internet Explorer and FireFox are currently being developed. Toolbars for other browsers will also be developed when those browsers support them.

Batch Processing – MetaFind Scripting offers a great deal of power and flexibility, but it is, after all, Scripting. Batches would allow you to enter search terms and groups, add them to a list, and then have a Script created for you.

Multi-Domain MetaSearch – another new type of search, this would overcome a common limitation in Advanced Searches. Currently, search engines that support a Domain restriction allow only a single Domain to be specified. By running multiple searches – one for each Domain – and blending the results, this search would allow you to create Search Groups that consisted of web sites without their own search engines, and even supported advanced searches of those sites.

Direct integration of your choice of McAfee SiteAdvisor, Norton SafeWeb, and Web of Trust into search results. Obviously, this is a feature that would require the cooperation of those three companies, and we have not in fact approached them with this idea yet. However, there’s not a great deal of technical difficulty here, rather, it’s an issue of having enough MetaFind users to make it worth their while.

DiffChecker – compare saved Blended MetaSearch results and find out who’s new, who’s gone, and who’s changed ranks.

MyMetaFind.com – registered MetaFind users will be able to quickly log into MyMetaFind.com, download and install a “Mini-Find” component, and be searching with MetaFind in ToolBar mode in just a few seconds. From any browser, on any PC, anywhere. All the power of MetaFind, with the convenience of a simple browser search. On-demand downloading will send you only the data needed for the search you’re running, and you will even be able to save your preferences at the web site, and have those applied wherever you use MetaFind. When you’re finished searching, it will even uninstall itself unless you decide to leave it in place.

Automated Real-time Parser Library updating – the design of MetaFind already separates the search engine result parser from the main program. As a result, adapting to changes in search result formats does not require a whole new version of the program to be downloaded and installed. The next step is to provide for automatic checking for parser library updates, which will be quickly downloaded and installed in the background.

Proxy support, and automatic proxy rotation.

Enhanced Text Fetch – the ability to spider entire web sites found in search results, then store and index the text would essentially give you your own specialized search engine.

Script Validator – at the moment, the Script Engine is robust, but unforgiving. The Validator would thoroughly inspect your Script and let you know about problems before you find out the hard way.

Separate engine for Scripting – the component MetaFind uses for all internet access is ideal for most users because it allows seamless integration of both result download and direct browsing. However, script operation needs no direct browsing, and will offer greater speed, flexibility, and reliability when driven by its own engine, which can be powered by a much less demanding component.

Expression evaluator for Scripting – although you can already extend MetaFind’s scripting functionality by calling external programs, you only get access to parsed results and ranks, and have to write actual Windows programs to use this capability. We are developing an expression parser that also provide the ability to perform a wide variety of logic, string, numerical, file, and web navigation operations within the Script Engine itself. These operations can be performed on the result and rank lists, but also on the original page content.


The MetaFind Development Team