Couldn't agree more. I agree with these ideas 100%, but it's getting to the point where they're all beginning to feel cliched.
You need to say the same thing over and over to refine your own understanding and presentation. You need to do this not just once, but periodically thereafter, because the thing you believed in 2007 will be different when you restate it in 2010. Experience has taught you things since then.
You need to say the same thing over and over because students don't learn it all the first time. They, too, grow in experience over time. When I reread the things that I read twenty years ago, I see completely different nuances than I did before.
But, most importantly, you need to say the same thing over and over because most people didn't hear you the first time, and even if they did, the world keeps making more people. If nobody ever reposts or retweets the classics, they rapidly become forgotten. (And then they aren't classics anymore: The reason classics are classics is that people can't stop quoting them.)
We've "all" heard that NDAs are worthless and that execution matters... and yet prospective clients keep pulling out those crazy NDAs. We've all heard that agile development is better than waterfall... except for the thousands of programmers who are working on waterfall projects right now. We all know that there is No Silver Bullet, and yet silver bullets continue to appear and continue to be touted, almost 25 years after the original essay.
It's okay that Spolsky himself no longer feels called to the vocation of patiently teaching the same things, over and over, to new generations of programmers and entrepreneurs. That job is not for everyone, and he's got a few other things he could be doing. ;) And it's also true that the web does a miserable job of segregating the newbs from the old hands, such that it's hard to avoid hearing lessons repeated more times than you really want to. But however much Spolsky may not need to write the 18,001st restatement of No Silver Bullet, there are always going to be people who need to read it. We're never going to finish the job of teaching people. They aren't born knowing.
Well said. A corollary is that Joel was unfairly labeled a "blogger" even though he never really blogged.
A blog is a reverse chronological list of short posts. Readers often assume a kind of exponential decay of usefulness, where old blog posts are far less interesting and relevant than new ones. Consequently, tech bloggers keep posting the same content, as if to say pay attention to this, it's still important.
What we really need is a new label for content like Joel's, one that doesn't carry any negative stigma for text written long ago. I have no suggestions for what that might look like, but I have to imagine there's something better than the status quo, which is:
(1) People keep posting the same stuff, and (2) I send around a giant, growing list of old articles to every newly minted CS graduate I meet
But blogs, including Joel's, display the "updates" page as if it were the main website. There are sometimes other views, like by tag, or on Joel's by topic in the sidebar, but they're positioned as clearly secondary, just pointers into the primarily-chronological blog. They aren't the default display, and posts are still prominently formatted with the date, as if they were journal entries--- on Joel's blog, the posting date is even on the same line as the title, in font that's just as big.
Collection of essays?
I see what you did there.
...
But, most importantly, you need to say the same thing over and over because most people didn't hear you the first time, and even if they did, the world keeps making more people. If nobody ever reposts or retweets the classics, they rapidly become forgotten."
I don't disagree with you. However, there is such a thing as overdoing it. And I feel that it's just as dangerous to do that as it is to not repeat yourself enough. After all, in saying something too many times, you're diluting your message.
Plus, posting these things on Hacker News is preaching to the choir. The companies that use NDAs probably aren't going to read hacker news. Neither are the "waterfall" developers.
I read Hacker News out of habit now, but it's not nearly as enjoyable as it once was. Perhaps I'm just getting old, but I remember the intellectual stimulation of scouring the web for great blog posts around 2002-2004. You know, before people worried about ensuring the content was search-engine optimized, or if they'd be able to sustain the Slashdot (or I suppose Digg/Reddit) effect. There was something special there and people seem to have lost it as they rushed to implement Google Analytics and ensure that their readership kept growing.
you can consume a blogpost in about a minute, maybe two. during an hour lunch break you can probably finish 20-30 posts. I found that the internet really isn't that big. There's only so many ways to way NoSQL > Sql , or scaffolding is the second coming of Christ, or how to give your VC the perfect handjob.
Even with a billion or two people available to you on the net, there just isn't all that much unique stuff. Ya, you've got 5 reasons your startup failed, but they always boil down to "My cofounder didn't work as hard/wasn't as smart as me/I should have listened to PG."
The internet itself has basically revealed itself to be 1) A great way to buy books, 2) a replacement for your encyclopedia and 3) porn.
The concept that the internet has "more" ideas is bunk. There are just more people.