# Cite documents rather than linking only

We should tell people to cite from documentation when they use link. I always look at old posts to learn something new, but some answerers give some link, but they are dead now (e.g., this). His first link seems to be broken.

Even currently there are more than 1000 answers which contains links and those links are broken. So if they cite what they want to tell then everyone can learn (say, 100 years later my granddaughter came to read one of my posts. If she sees my links are broken and I didn't cite anything then she, perhaps, can't find what she was looking for).

So, my suggestion is we should tell people to cite rather than only linking document.

– ACuriousMind Mod
Apr 8 at 12:56
• I don't really understand the question - as long as the link works, the answer is clearly more useful with the link than without it, so why remove it? You can of course ask the answerer to include the more helpful details directly in their answer or even propose an edit that does so, but then it is even more important that the link stays as attribution for the citation.
– ACuriousMind Mod
Apr 8 at 13:44
• To be fair, most people interested in physics will quickly come to recognize what Thorne and Wheeler's book is (and either Wikipedia or Amazon will get them there quickly). For journal articles, particularly since direct journal access is highly variable, I will add an actual reference (journal, volume, pages, year). Apr 8 at 14:19
• @ACuriousMind It is asking for Physics S.E. or its users to recommend a citation in addition to any link, not instead of one. Apr 21 at 1:57

The dead link in your particular example has been preserved by the Internet Archive. The Internet Archive is a fabulous project, and if you don't know about it you should spend some time getting acquainted.

However, looking at the link, it appears to be a student poster presentation from an AAPT meeting, whose "formal" citation is probably only an abstract in a meeting program from 2001. I made some conference posters around that time, which I may or may not have thought to put on my squiggle-username university student website, and which therefore may or may not still exist anywhere to be retrieved using a citation. (One of my posters from that era survived for a surprisingly long time as a letter-sized printout on my mother's refrigerator, but I think it was replaced by some mushroom people drawn by my own children.)

As a comment says: an answer which isn't useful without its link is not an answer, for exactly this reason.

Beware also that some links are better than others.

• For publications, a link to a digital object identifier may be more useful to users of this site, especially casual users, than a traditional citation. A person who doesn't know what to do with "Phys.Rev.Lett.867-5309(Jenny)" will almost certainly know to click on the underlined text, and the digital object identifier is supposedly stable even if the publisher updates their internal linking scheme.

• Links to the arXiv preprint server also likely to be stable long term (though if you notice a link to arXiv.org/pdf, please edit it to arXiv.org/abs, so save casual clickers the pdf auto-download).

• Wikipedia is great, but beware that your Wikipedia link won't say quite the same thing in five years as it does today.

• Youtube URLs seem to be mostly stable, but videos do sometimes disappear. Video links are also less accessible than text links, because they're mostly impossible to skim. We mostly don't have the problem of people trolling with links to irrelevant music videos, but there is always the risk.

• Links of the form school.edu/~squigglename, as in your example, are practically guaranteed to vanish when their maintainer graduates or moves to a new position. There are a few exceptions.

• For future visitors, if anyone wants to learn about the internet archive then they can watch it through YT. Apr 8 at 15:56
• The arxiv is now also rolling out DOIs. Apr 10 at 16:41