I like command line tools like traceroute, whois, groper, man, host, dig, gdict, etc.
A couple years back I’d written a tiny command line client that would let me search google (and a few other search engines) and output their results in plain text, ikiwiki format and org format.
I could also pipe those results into elinks or a speech synthesizer.
I loved it. I could search without leaving my editor or terminal session AND get it into a format I could easily use elsewhere, and track the keyword trails of my own searching, outside of google. (I don´t bookmark stuff, I use org-protocol to sweep useful stuff into emacs, and remember only the keywords I searched on, and the position of the results)
I didn´t want to have to get a license to search, so I had a buddy write a scraper. That stopped working a few weeks later, so I bit the bullet and wrote something that used their API.
That API got deprecated a few months later, and my search client stopped working a year ago. I stopped using it, reverting to the web for all my searches.
So I spent my spare time in the last two weeks writing new command line search client, using the newer json API that replaced the previous API, that had replaced my scraper.
I got THAT doing useful stuff again yesterday…
Only to find that google had deprecated THAT API… as of Nov 1st, 2010, and put a 100 query/day limit on the new one….
Today I heard facebook is trying to replace email entirely…
I am failing utterly at finding some form of commenting system for this blog that offers anonymity and privacy and local control of the comments…
I watch agast as myopenid tracks every website you login to… that´s a little more open about my ¨ID¨ than I would like.
Is Orwell spinning in his grave?
Hey, buddy, youse got a license for that question?
Meanwhile, all those other search tools I use daily have stayed the same.