Never post a web link that requires readers to click on it to find out if they want to click on it.
Ha. Depends on your goals.
If you are Buzzfeed or otherwise a click-farmer it's "Make each web link seem like it leads to life-revolutionizing information, with only tangential regard as actual content".
Or tab-exploders like TvTropes or even Lesswrong at times- "Give links obscure and cool words which just barely hint at something novel and unusual without revealing it, and include as many of them as possible per paragraph"
Never post a web link that requires readers to click on it to find out if they want to click on it.
Ha. Depends on your goals.
The other side of that coin is "Never click on a link unless you already know that you will be glad to have clicked on it." Recognising clickbait is as valuable a skill as recognising spam.
"7 secret signs of clickbait you won't believe! Discovered by a mom! Spammers hate her!"
This is the first post of the 2015 repository rerun, which appears to be a good idea. The motivation for this rerun is that while the 12 repositories (go look them up, they're awesome!) exist and people might look them up, few new comments are posted there. In effect, there might be useful stuff that should go in those repositories, but is never posted due to low expected value and no feedback. With the rerun, attention is shifted to one topic per month. This might allow us to have a lively discussion on the topic at hand and gather new content for the repository.
The first repository to be rerun is the Boring Advice Repository, because of... on a whim.
Enter original motivation (by Qiaochu_Yuan):
The Boring Advice Repository is filled with lots of diverse advice, I've summarized some of it in a comment below.
So what should go here? To go with Qiaochu_Yuan again (adding emphasis):
I don't know if you should post new advice here or in the original repository. Perhaps search the old repository with ctrl+f (when on windows) and if you don't get results, post it here.