Benjamin Franklin and me
My first job was one of "consultant". In those times, consulting meant
reformulating a problem the client knew he had, and listing fairly
obvious solutions to throw at it (with more consulting hours).
The client would then either
- entirely ignore our advice and pretend the problem never existed
- pick up the cheapest solution and cut the recommended budget by half
- come up with a grand, very expensive plan to address not just this
specific problem, but any one of its kind, for now and the rest of
Consulting firms are said to hope and push for solution #3, but that's
only partially true. As human beings, we like to help, solve problems
and build things that last.
Yet it's true that if we can't implement a good solution, we'd rather
bill a lot of hours for a bad solution than not participate in it at
all. (Consultants are very good at claiming success as theirs while
blaming others for disasters.)
After a few years of this, I chose a different line of work and went
on to build custom systems and solutions (i.e., without the consulting
part whenever possible): I had come to the conclusion that smart
companies are able to guide themselves, while not so smart ones don't
benefit from external suggestions very much, or listen to the wrong
I was pleased with this diagnostic... and then I came upon this
phrase: "Wise men don't need advice. Fools won't take it."
It's usually attributed to Benjamin
although it may have come from an earlier English proverb. I find it
quite true; I kind of wish it hadn't taken me so long to rediscover
Do I Even Need Google?
I used to think my life depended on Google -- specifically, Google
Search. Now I'm not so sure. I'm not talking about using an alternate
search engine such as Bing or Duck Duck Go -- I'm talking about not
using a search engine at all. Here's why.
I mostly search for three kinds of things:
- answers to programming questions
- answers to general knowledge questions
- things to buy
Not long ago, those three categories were populated by millions of
webpages, out of which Google would extract the best and most relevant
information. Then consolidation happened:
- I only shop on Amazon
- Wikipedia has answers about everything
- more importantly (because as an engineer, this is my livelihood),
every programming question is answered on StackOverflow
So, at least in theory, were Google Search not around anymore, I could
search each "category killer" separately.
It's true that search engines for specific web applications are very
bad (Amazon's A9 being the worst of the worst); they're so bad that
we will continue to use Google to search inside one website
(using site:) instead of going to the website and search from
there. But they're not so bad that we couldn't use them if
Google were suddenly unavailable.
Of course, like everybody else, I still use Google Search every day;
but for the first time I realize I could do without it. I used to be
afraid -- and now I'm not. This is quite novel.
Unmoderating Hacker News Titles
Hacker News is an extremely
addictive social news website that lists stories of interest to
On HN, people submit stories with a link and a title; stories
that are voted up make it the front page where they enjoy
a lot of traffic.
Titles are first written by the person submitting a story;
but then moderators sometimes change them, with no way
to know what the original title was.
Users have complained about this many times, but so far to no
avail. So I made a little bookmarklet:
that displays the original titles next to the new ones; here's how the
home page looks like after clicking on it:
- only titles that have been substantially modified are shown
(for example, if a moderator only removed a dot at the end,
- it doesn't interfere with the original site in any way; if you click
on the bookmarklet many times you don't create a load on Hacker News
(only on my server!)
The less good:
- the main limitation is that it currently only works with Firefox
because Hacker News is in SSL and the bookmarklet isn't; only Firefox
allows mixed content by default (see security.mixed_content in about:config) and I'm not sure I want to buy an SSL certificate
just for this
- of course, since I have to store original titles and only started to
do so yesterday, it only works with current stories, not on old ones
- it doesn't work on comments page yet, but I'm working on it (on a
comments page there is just one story title)
To try, you just need to drag this link:
to your bookmark bar, and then click on it when you're on Hacker News,
and see what happens!