9 ms·
ok I'll go against the grain here but I personally found this book very tedious and of little value to me. But then I had been programming professionally many y
by plinkplonk 16y ago
ok I'll go against the grain here but I personally found this book very tedious and of little value to me. But then I had been programming professionally many years when I encountered this book so maybe I was not part of the intended audience.
(Then again I found "Godel, Escher, Bach", another book that is almost universally lauded hereabouts, to be frothy and pretentious. So maybe it is just that I am weird.)
- zwieback 16y agoI'm with you on Goedel, Escher, Bach. Also, thanks for "frothy and pretentious" - adding it to my verbal arsenal now... I like CC but also agree it's of less use to the seasoned programmer than to beginners. Then again, seasoning doesn't necessarily produce good habits.
- famousactress 16y agoMe too. I think the first couple chapters contained a few really valuable truths that resonated me way back when I read the thing.. the rest of book was bleh.
- DanielBMarkham 16y agoPlink, you should go over to Amazon and put in a review. There's like 150 reviews over there and none of them are less than 3 stars. Seriously. If you have a detailed criticism, it could help others not waste their time with the book. I find many computer books tedious. Computer literature, for some reason, has a much greater variation in styles than other forms of literature. Some books are basically just big lists, and folks love 'em. I read the book early in my career, so maybe your point about timing is a good one. Don't know. I do know that unless you're just being snarky, having an alternate point of view can be very useful for folks choosing what to spend their time on. I really regretted not being able to ding this book more. It'd be good to see a more detailed negative review. And GEB is on my short list :) I still have a dog-eared copy from 1995 or something.
- sruffell 16y agoEven as a seasoned developer I think the book has value when trying to make the case to less-seasoned colleagues that they might want to think more critically about how they are structuring their code.