Thread: Hodor's Opinion Concerning Cross-Posting

Whoa! Who made my post a proper topic?! I'm kind of nervous now.

What works in a scientific field does not necessarily work in an open forum relying on the volunteer efforts of experts and enthusiasts.

Additionally, the original post in question is not an obscure scientific question posed to experts, but a general lack of understanding of fairly basic concepts.

There are many compelling reasons why cross-posting is considered poor etiquette. To me, the most cogent argument against cross-posting was by Salem (see post #14 here) - basically, a slippery-slope argument where tolerating such behavior could lead to an increase of that behavior, possibly causing a break-down of such helpful forums.

Moreover, this has been discussed here before: God hates cross-posters!

Originally Posted by Matticus

Additionally, the original post in question is not an obscure scientific question posed to experts, but a general lack of understanding of fairly basic concepts.

That's a fair enough observation. At first glance I don't think I agree with your statement but it's very late here in Australia and I will consider it properly before writing a considered response.

Originally Posted by Hodor

This is going to make me very unpopular. But I have to say it.

In my professional life (which has nothing to do with programming but has to do with science) it's very common that peers will ask many people the same question (or if not a question, seeking the same result), without letting the recipient know that Joe, Bob, Alice, Roger, Sarah, Jane and possibly more people have been emailed (or whatever) the exact same question, query, or request for advice. It's not just me who does it -- everybody does. If someone sends me something to review or asks a question, I would never be offended or think that my response was "less valuable" if I found out the person asked another person or another 100 people to do the same. I would not consider they were wasting my time; it's just a normal thing to do.

I know that cross-posting is considered (by some) bad "etiquette" but this is probably due to cross-posting on Usenet as either a way of causing a rift between two groups (newsgroups) of people, or because it's annoying to receive multiple alerts. This, however, doesn't really -- in my opinion -- reflect how things work in real life.

If the cross-posted message/whatever is an obvious (or not so obvious) troll or spam, then that's a different scenario. Asking more than one person or group the exact same question is not, however, intrinsically bad -- it's quite the opposite. In fact, even back in the Usenet days cross-posting wasn't considered as an automatic "bad thing".

Hodor

Edit:
It's lucky there is not a "dislike" button here!
Also, does posting code to, I dunno, codepad and asking the question on IRC also considered cross-posting? Of course not.

I agree. Assuming that it's not a straightforward question, you really need to get feedback from multiple sources. It's called doing research and not simply using one person's opinion as fact. I think cross-posting is mostly considered bad because it's usually some idiot student who has homework due tomorrow and is posting on 5 different sites to see who will help him/her first.

Discussion is not review. Review ≃ evaluation, discussion ≃ development.

Originally Posted by Hodor

it's very common that peers will ask many people the same question (or if not a question, seeking the same result), without letting the recipient know that Joe, Bob, Alice, Roger, Sarah, Jane and possibly more people have been emailed (or whatever) the exact same question, query, or request for advice. It's not just me who does it -- everybody does.

Just because everyone does it, does not make it sane.

The trouble with that approach to discussion or development is that the different solutions are never compared directly. Instead, the asker will use secondary, scientifically irrelevant, criteria to determine which one is the most plausible: "statistics" (even though you cannot make any kind of reliable statistical evaluation on a sample so small, with sampling biases dominating everything else), "standing" or "references" (because that person has gotten a Nobel price, they must be right), personal bias (I like that person, so they're most likely right), or a myriad of other reasons. But the solutions never directly come into competition with each other.

I've mentioned earlier that I'm just a burned out husk of a man. One reason for that is exactly the above behaviour. In real life, I do not have the kind of charisma or presence that gives my opinion weight. (That is also one reason I am so verbose: lacking truthiness, I try to give that weight using background information instead.) I do believe I can list at least a dozen occurrences where my proposed solution was ignored, with a faulty one chosen instead, until at some point where the faulty solution had to be discarded and my original solution adopted instead (but with absolutely zero credit to me, of course). I've probably garnered quite a few "hidden" enemies that way, too, by no act of my own.

I do believe that this answer of mine elsewhere pretty much illustrates my communications problems. (It is one answer of several, but one of the rare ones with zero votes. Yet, I cannot see what makes any of the others more preferable. Yes, it is absolutely write-only code, but the question explicitly states "I'm curious how few bit-wise operations could be taken".)

All this would be avoided by combining the discussions, allowing the proponents of solutions to point out flaws in others, and possibly come up with a better one altogether. That is, after all, the purpose of a discussion: development of ideas.

I hope that the questions Hodor has forwarded to multiple people have been more of the evaluation type, rather than the development type. I fully agree that peer review (evaluation) works, and I think it (and all sorts of evaluations) work best when the evaluators do not know who and what opinions the other evaluators have, nor how many others there are. But, again, evaluation ≠ discussion.

I for one am vehemently against cross-posting: it fractures the discussion, and promotes less than optimal answers, because the separate discussions lack access to all counter-arguments.

I would not mind it that much, if counterarguments in different discussions are pointed out as soon as they occur; it's just that it takes much more effort to follow a fractured discussion, and usually is not worth the extra effort.

Thus, my opinion: cross-posting ≡ wasteful suboptimal behaviour.

I'm a pretty jaded person, so my opinion is that people cross-post because they want to listen to the most helpful person. Like Nominal Animal was saying, this usually means the person who does the most work, who writes the most code, or who answers the shortest way. It's legitimately hard for me to curb my thinking if I dwell on it; aren't they just looking for the answer that helps them get through it with minimal personal effort? That is why cross-posting is bad.

Sometimes I even have a feeling like this on regular answers. If people read my post and all they come back with is something like, "I still don't understand," I wonder if they really cared what I said at all. Perhaps I linked to a manual page somewhere. They probably didn't read it. So, I post again, copy-pasting the sections of information I wanted them to read.

The problem with enforcing rules against cross-posting though is everyone has to check or it doesn't work. Variations of "Also posted here" are coming after I write something detailed lately, so maybe I should start. But there have been occasions where the poster wasn't getting really helpful responses elsewhere, in my opinion, like posts that were good, but didn't address the issue. Under those circumstances, I'm glad I posted something.