search

betterscientificsoftware / bssw.io

Better Scientific Software Homepage
https://bssw.io
Other
139 stars 88 forks source link

Numerical recipe books on line #1340

Open markcmiller86 opened 2 years ago

markcmiller86 commented 2 years ago

http://numerical.recipes/oldverswitcher.html

bernhold commented 2 years ago

Hmmm, I'm of two minds about this as CC material. With the caveat that I haven't looked at the new (2007) C++ book, I think it would be safe to say that the earlier books in the series were widely viewed as both the most useful and the most harmful books for developers in their day. The algorithms were often simplistic, and as time goes on, less and less likely to be what you'd actually want to implement on today's hardware. And the code that came with them was almost certainly the most widely pirated code in scientific software, back in the day. Even now, it doesn't look like they've opened up the code.

markcmiller86 commented 2 years ago

it would be safe to say that the earlier books in the series were widely viewed as both the most useful and the most harmful books for developers in their day.

Is this two-truths perspective somehow unique to this material? I don't think so. I think the same can be said of a great number of resources and practices.

as time goes on, less and less likely to be what you'd actually want to implement on today's hardware.

Isn't this material good introductory material for people just learning about numerical algorithms?

And the code that came with them was almost certainly the most widely pirated code in scientific software, back in the day. Even now, it doesn't look like they've opened up the code.

😕

markcmiller86 commented 7 months ago

So, I would like to resurrect this issue. I got some confused faces on it and I honestly do not know why.

Hmmm, I'm of two minds about this as CC material. With the caveat that I haven't looked at the new (2007) C++ book, I think it would be safe to say that the earlier books in the series were widely viewed as both the most useful and the most harmful books for developers in their day.

Harmful...how? I used material I learned from it with great success in my Master's thesis

The algorithms were often simplistic, and as time goes on, less and less likely to be what you'd actually want to implement on today's hardware.

Speaking as one who often develops "simplistic" material (e.g. hello-numerical-world), especially for the purposes of introducing others to new concepts in the discipline, I am bemused why we would cite simplicity as a criteria for declining to curate.

And the code that came with them was almost certainly the most widely pirated code in scientific software

Hmnmm...wide pirating is a) not really our concern or b) an acknowledgement that the code was found to be highly useful.

back in the day. Even now, it doesn't look like they've opened up the code.

The article would be about the books, not the code.

markcmiller86 commented 1 week ago

I am happy to do this one for November.