Search Engines: Plumbing Becomes a Thing Again

August 10, 2020

Two search related items.

The first is Hndex. If you want to locate articles posted to HackerNews, a tech-oriented headline aggregation site, you have an option. This is an example of what might be labeled a “site specific search” solution: One site, search it. Navigate to and plug in a search term. We entered a query for “enterprise search” and retrieved on point results. The comments are available; however, these are not indexed. Click the “cached” button, and you can view the original article. Click the “comments” button and you can view the comments. HackerNews provides its own search service, which is weirdly located at the bottom of the page. DarkCyber will reserve further comments until we have experimented with the system for a few days.

The second is Infinity Search, another metasearch engine positioned as a free Web search system. DarkCyber finds metasearch engines interesting, but these often pretend to be running their own crawlers. To Infinity Search’s credit the company states:

When you search for something on our site, we take the results from other search engines and our own indexes, organize it, and display it directly to you without logging any information about you.

Metasearch systems have to deduplicate results lists and find a way to remain in the good graces of companies running primary Web crawlers. Disclaimer: My son worked for Vivisimo (now the heart and soul of one of IBM’s marketing confections. He has moved to other adventures, but I remember our talks about the issues metasearch presents. For example, latency, screwed up query interpolation, and wonky deduplication which deduplicates useful results out of the results list. I think Vivisimo lives on in, but I am not a fan of metasearch systems which recycle others’ indexes and remain vulnerable to partners who pull out of deals, thus putting a dent in results.

Stephen E Arnold, August 10, 2020


Comments are closed.

  • Archives

  • Recent Posts

  • Meta