If I write a reference (bug link, comment link or combo link) that doesn't exist, and it doesn't get linked because of it (well, it shouldn't get linked, see bug #146752), I might have mistyped. Hence I should have to confirm each of the links that is broken before the comment is added or the bug is posted. One likely spanner is referring to bugs on other bug systems, (eg redhat bug XXXX), but that's probably not a major problem given it's a confirmation. People shouldn't be doing this anyway because it can generate incorrect links. We should support inter-database links of course. And getting back on track (and this bug actually came out of a comment on one bug which made me realise I had to comment on another bug first and then that made me realise I had to file this and you would think I've had too much caffeine wouldn't you), errr where was I again? Oh that's right - fix it - it's broken.

>One likely spanner is referring to bugs on other bug systems And exactly because of this the functionality should IMO be configurable so that I can switch it off from params or something similar. Some installations use more external references than some; there's no sense in annoying people with needless confirmations on environments where "non-working links" are common.

(In reply to comment #1) > there's no sense in annoying people with > needless confirmations on environments where "non-working links" are common. I agree, such confirmation pages are a pain.