Ben Summers’ blog

Searching for techniques in a changing world

A few days ago I wrote about using localStorage for messaging between browser windows. Within hours, readers had pointed out there was nothing new here, Google and Facebook had been doing this for years, and linked to:

It’s obviously an ‘old’ technique, but it’s new to me and a large number of others: ten thousand people read that blog post in 24 hours.

I think this has illustrated a problem in search: Given an objective, how do you find up-to-date information about techniques for achieving it in an ever-changing world?

I’d actually done quite a bit of research, reading specifications and searching for phrases like “inter-window messages”, “browser window message”, “browser window broadcast”, and so on, on Google and on Stack Overflow. I’d missed trying ‘tab’ instead of ‘window’, I suppose because the underlying browser object is called window and when I started doing this kind of thing, tabs weren’t invented.

But try the search yourself: inter window messaging

Google guesses, accurately, that “cross” is a good synonym for “inter” (but advertisers fail and attempt to sell you physical windows). And the results are all about communicating between browser windows — but only when you opened that window via JavaScript or put it in a frame. They’re good results, still relevant, but only a limited solution for sending messages between browser windows.


There seems to be a few problems here:

  1. Not enough people are writing about it.
  2. Not enough people are realising that the “Web Storage” API should really be called the “Web Storage And Messaging” API.
  3. The articles which do exist are drowned out by lots of high quality, useful and relevant articles about something very similar which uses exactly the same terms as the new techniques.
  4. Terminology changes over time: What an old-timer like myself calls a window, all the cool kids call a tab.

In a way, it’s a chicken and egg problem: (1) and (2) require (3) to be solved first, but you aren’t going to find the new articles when the older ones are so well regarded by search engines. Plus you have to help people like me by keeping synonyms up-to-date, to solve (4).

Problems solving the problems

There’s lots of very successful work being done with natural language processing to solve the problems around synonyms, and working out what my query actually meant.

But there does seem to be a more difficult problem. How do you let new information about changing techniques appear at the top of the search results? In this case, the useful articles:

  • do not have many links from the rest of the web,
  • are quite recent,
  • use exactly the same keywords as old and relevant articles.

However, any algorithm for pushing new content with those properties to the top is likely to be used by spammers, scammers and “SEO Experts”.

Everyone wants to get their pages at the top of the search results, and any algorithm which pushes new pages like these to the top is going to be abused by those who want to make search engines return rubbish for their own profit.

That said, maybe I’m reading too much into a case where I just failed to use a search engine to find something. But might search engines be holding us back by making it more difficult to share new techniques for solving old problems?



blog comments powered by Disqus


Hello, I’m Ben.

I’m the Technical Director of Haplo Services, an open source platform for information management.


About this blog


Twitter: @bensummers



Jobs at Haplo
Come and work with me!