Not signed in (Sign In)

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.

    • CommentAuthorYemon Choi
    • CommentTimeAug 23rd 2011

    The question is here. As I said in the comments, I am not convinced that this deserves to be closed; my initial feeling was that it fits into a species of legitimate question, namely career advice or advice on good practice. However, I have not reflected on the question, so am willing to be swayed.

    (There also seems to be some ambiguity, as to whether one should interpret the question as normative or as descriptive.)


    It's a very interesting topic, but I just don't see awesome answers. Anecdotes will be lame --- can someone think of a reformulation, perhaps asking about studies, statistics, or essays written about this elsewhere?

    • CommentAuthorWillieWong
    • CommentTimeAug 23rd 2011 edited

    I cast the final close vote. I think in its current incarnation it is not a good question. There are probably as many ways of using the arXiv as there are mathematicians, and right now it is way too open-ended as a question. Reading a bunch of descriptions of what people do probably won't help OP much in figuring out how he should use the arXiv. Asking for "what other people do" is somewhat different from asking for advice.

    That said, I can be convinced that the question is useful if it were edited into a more focused question with a more specific goal in mind.

    For example: a possible reformulation is asking about the dos and don'ts, something like: "I am preparing a paper, and I plan to post it to arXiv. What are some of the common mistakes made by a person using the arXiv for the first time that I should make sure to avoid?"

    Then a reasonable answer could be: "Don't cross-post to every subject under the sun. That's considered spamming."

    or "Uploading TeX is better than uploading PDF is better than uploading PDF generated from Word."

    • CommentAuthorgrp
    • CommentTimeAug 23rd 2011

    I can imagine several questions that involve arXiv and might be suitable for MathOverflow and not meta.mathoverflow. All of them go away from what I perceive as the intent of the question: when should I publish on arXiv?

    The closest I can come is the following made up scenario, which should be read with less tongue in cheek than might appear: "I have a draft of a paper on numerical methods for simulating models that approach solutions to the Navier-Stokes equations. They are built upon work of Professor X, and seem to be a significant improvement.

    I showed it to my advisor, and asked him if it was suitable for publication on arXiv. He demurred, said that he wasn't familiar with the numerical analysis part, and suggested I ask the NA prof down the hall. He also suggested that I ask on MathOverflow.

    I went down the hall, and showed the NA prof the paper. He said he'd look it over, but it might take a few weeks as he was not familiar with the conditions surrounding Navier Stokes, and as for suitability for arXiv, he suggested asking Professor X or his students in the meantime. Also, he added, you might ask on MathOverflow.

    Unfortunately, Professor X passed away. I talked to two of his grad students who I could track down. Both of them said they would be happy to look at the paper, but since they had not submitted anything, they could not judge suitability for submission. Curiously, they had been cleaning the Professors office and found a Post-it on which was scribbled in X's handwriting the title of my paper, and the phrase 'arxiv? MathOverflow?' on it.

    Upon leaving the building I ran into a colleague of X. I asked him what he might have known about X and what he thought of my paper. He was very cordial, and said that he wasn't sure, but it sounded like my paper might be suitable, and anyway, why not ask on MathOverflow?

    So here I am. The link to the paper is here. What say you?"

    The points I wish to make to the above are the following: the motivation and the previous efforts are made clear. The question is specific and pertains to mathematics. By some standards this might be a good MathOverflow question. It certainly preempts the standard answer to ask ones advisors or colleagues.

    However, it it still a request to read a paper and judge it, which is not something to encourage at MathOverflow.

    Also, this story reflects the idea that there is perception that this is what MathOverflow is good for. Anything done to support that perception will require much more work to reverse or counter.

    If the community is ready for this sort of change, so be it. My gut says not yet.

    Gerhard "Ask Me About System Design" Paseman, 2011.08.23

    • CommentAuthorquid
    • CommentTimeAug 23rd 2011

    Somewhat related blog post

    The respective MO question got closed, and generated some discussion if I remember well.

    The current question is different, and certainly less controversial. I have nothing against it being open (for a while) and yes CW; if it draws too many answers and becomes noise one could still close it. However, I have no intention to 'fight' for it being reopened.

    Some quick answer to the current question, even if anecdotal, IMO could definitely be of immediate practical value to the questioner (who seems to be merely looking for some general opinions, or confirmation that the suggestion already received is in line with common practise) and perhaps some others. People could say what they personally do and why the do what they do. Or what they tend to do under which circumstances. (Just as Joe Silverman already did.) Everybody interested in advice could then judge for him/herself which described situation is closest to his/her own and act accordingly. In this sense to some extent this question seems better to me than if it would ask 'what should I do' because then noone knows the details of the situation.

    The question Willie sketches to me seems less useful as I believe the FAQs/documentation of arXiv is quite good and addresses such things.

    In the end, I vote to reopen this one: it's borderline, but it seems immediately useful, and completely harmless.

    P.S. in some sense the discussion becomes obsolete, as there is now an other answer in the comment and somehow I believe not too much more can be said. Perhaps this is a pragmatic solution.


    I can be convinced that the question is useful if it were edited into a more focused question with a more specific goal in mind.

    I think this is the only obstruction to it being a good MO question (but it's a serious obstruction). I don't think the question needs to be made more pertinent to MO, as Gerhard (sort of) suggested in a comment. Questions about dealing with refereeing, journals, departments, and job applications should be considered perfectly acceptable on a professional forum. However, such questions require extra care be taken to specify what the goal is and what constitutes an answer, so as to weed out the lame posts gunking up the thread.

    Joe Silverman suggested that the question shouldn't be closed, gave a good reason, and even gave what I assume is one of the most popular answers. But he didn't post it as an answer, even though he had the option to do so. This suggests that he felt (as I do) that something is amiss with the question. I want to say that the question simply casts too broad a net to be useful, but I can't think of a very good refinement. Perhaps something like "does posting an early draft on the arXiv generally increase or decrease the number of people who look at the final version?"


    By the way, I think the question should be CW, but only because it's effectively formulated as a poll. That is, the decision to make the question (and therefore the answers) CW makes the thread more useful. That is, the important thing is not so much the content of the posts (which anyone can generate), but the numbers next to those posts.

    The non-poll version of the question I suggested above ("how does an early draft on the arXiv affect the final readership?") is not a poll question and shouldn't be CW. In that case, the contents of the answers (arguments/data) are the important thing.

    The question is simply too vague, which by definition makes it bad. A specific question of the kind: "I have this paper, here are the circumstances that make me hesitate to post it on arxiv; what do you think/advise?" would be perfectly fine for MO. "What's the convention?" strikes me as inappropriate, especially because there is no evidence that such a convention exists. ("What's the convention?" questions could conceivably be acceptable in older areas where such a convention may actually exist.)
    • CommentAuthorquid
    • CommentTimeAug 23rd 2011 edited

    Thierry, it seems to me that 'convention' might simply be a non-optimal choice of word. What IMO is asked is: What is a typical time to post on the arXiv.

    With potential answers: considerably before submission, somewhat before submission, at submission, somewhat after submission, after acceptance, when the paper appears,...

    Or, still more explcitly: "I was told it is common to submit to the arXiv about at the same time when submitting the paper for publication. Is this true? (If not what else is a common practise)"

    • CommentAuthorNilima
    • CommentTimeAug 23rd 2011 edited
    EDIT: removed comment.
    • CommentAuthorgrp
    • CommentTimeAug 23rd 2011

    Nilima, the numerical analysis part was a nod to the need for more of it on MathOverflow, and the N-S was dredged up from my undergrad days in a previous millenium. Inasmuch as experience shapes what I write, everything else was pure invention on my part.

    I am sorry to hear Prof. Marsden has passed. Sorrier still that I did not make his acquaintance when I had the opportunity at Berkeley. At least I take care of his Vector Calculus text (with Tromba) and oh so infrequently refer to it.

    I invite you to contact me through email using the info on my registered MO user page. I think such detail as above is best reserved for email and outside of meta.

    Gerhard Paseman, 2011.08.23

    • CommentAuthorNilima
    • CommentTimeAug 23rd 2011
    Indeed, my previous comment was entirely off-topic, and I've edited it to near-0. Thanks for the gracious response, Gerhard.

    Just to be clear: I cast the last vote to reopen (effectively without using my moderator powers) and I made the post community wiki (using my moderator powers).

    I personally think this is an interesting, relevant, and important MO question. My recommendation for MO users with 100 points is to edit the question to include everything you ever wanted to know about the arxiv (but were afraid to ask).


    Incidentally, I find that this question isn't too different from one of mine [1] that was closed a few months ago. Just sayin'...


    • CommentAuthorquid
    • CommentTimeSep 23rd 2011

    This question was closed, too. It was only reopened after some discussion. Chances are you'd have gotten yours reopended as well, but you seemed a bit disinterested in your own question. Also, see my remarks in the meta-thread of your question ; in my opinion a main difference is that this question was a lot more focused than yours (but as I said right away in my opinion you also had bad luck). Even now, there is nothing stopping you from focusing your question with an edit, perhaps then it gets reopened.