I just posted a new answer on the question on dynkin diagrams and was asked:
Is it OK for me to include [the diagrams] into the paper on this subject?
The answer is, of course, "Of course", and I directed the person to the link at the foot of the page.
Reading the links myself, it seems that we need to decide what "proper attribution" is for such things. I would hazard a guess that we'll actually want to define a few different levels of attribution.
- Very minor stuff, such as what package to use
- Stuff that actually involves putting in some code, but is "low level"
- Really obvious stuff, like diagrams and complicated layouts
I'm tempted to say that for (1), attribution isn't required, though a nod to tex.SX in the acknowledgements would be nice, whilst for (3) I'd go for something more formal. Of course, the real question is where to draw the boundaries!
As well as credit-where-credit's due, there's also the principle of attribution-as-advertising, so although we don't want to cripple someone's paper by requiring every last hack to be properly attributed, we'd still like a mention.
Thoughts? Suggestions on format, citation type?