I have an idea: it might be interesting to have a "pinned articles" section at the top of HN, similar to announcement threads in forums. It would only display for logged-in users (so it wouldn't bother anyone who was just anonymously browsing the site) and, once you were logged-in, you could simply close the div displaying them. The prime candidate for a pinned article would be the HN guidelines; this would probably be another.
Closing the div would actually just mark all the pinned items currently in it as done-with for you, so any newly pinned items would still appear. If you closed the div when you didn't mean to, you could also check the new "pinned" section on the top banner, which would display all pinned links regardless of done-with status.
It's quite annoying to see entire books submitted. It should be a requirement that you must be able to thoroughly discuss a submission. You can't discuss the entire contents of a lengthy text in a HN thread. It's like submitting a Google search or an index to pg's essays.[1] The discussion is always superficial.
If you were really interested in SICP, then you probably would have already found this link. It's the 2nd result on Google for "SICP". A blog post on the merits of SICP would be more useful.
I totally disagree. I consider links here recommendations for good reading. If the discussion is not as in-depth as you might like, think of it as a set of informal reviews and opinions, probably more reliable than the ones on Amazon.
"Ask HN" threads are much more appropriate for book recommendations. Imagine if every book in the threads below were submitted- we'd have a bunch of noise and no real discussion.
Submissions like this are only useful for people too unmotivated to Google for "SICP".
I actually never knew the link existed. As a Computer Science student, I can't describe how much I appreciate having this book available online. I simply can't afford to shell out $50 dollars for every intriguing and intellectually interesting book that is out there. The benefit I can gain from reading this book is immense, and I am very exciting to be able to work with it.
So yes, I would say it does have to get posted every week :-).
We get that the book is immensely valuable, and it's great that it's offered online. But this is something that anyone can find with ~5 min of Google searching.
Sorry :) I've seen terrible scanned versions of it posted before, but this particular source was never posted. The scanned versions are difficult to read and I figured others would enjoy seeing a clearer version.
This is the entire book, hosted by the publishers (MIT Press). I find it much more convenient to read it here than the scribd PDF or even the hard copy itself because when they reference earlier problems and sections in the book, they usually link to it for easy navigation.
