13
$\begingroup$

One very enjoyable feature of mathoverflow is that the math just works. You enter it using the usual LaTeX, and then the jsMath magic does the rest. One of the most frequent activities for many mathematicians is checking abstracts on arXiv. So you open an abstract and you see all these dollar signs which you have to mentally decipher.

Question: Is there a way to set up jsMath, Firefox, etc., so that you can see formulas when viewing abstracts on arxiv.org?

This may be a meta question, but who reads meta threads, right? And there is a high concentration of experts here who may be able to answer this. That would improve quality of life for many mathematicians, I am sure.

Notes on the comments:

  1. Yes, people should make an effort and write abstracts on arXiv without using math. symbols, preferably. But the easiest thing to do, really, is to reuse your article abstract for the arXiv abstract, and that's what people do. And you can easily use some math symbols in your article abstract.
  2. Is it a question for arXiv people and not for MO? Maybe. But this is a very active community of people who encounter this problem on a daily basis. I hoping there is someone clever enough that solved this problem for themselves already.
Improve this question
$\endgroup$
17
  • 5
    $\begingroup$ You should contact the arXiv people, really. $\endgroup$
    – Mariano Suárez-Álvarez
    Commented Mar 29, 2010 at 1:25
  • 2
    $\begingroup$ Well... hopefully, the author of a math paper speaks math-ese :P $\endgroup$
    – Mariano Suárez-Álvarez
    Commented Mar 29, 2010 at 2:31
  • 3
    $\begingroup$ This is a very borderline case in my opinion, but I'm voting to close as 'off topic' on the basis that this question should be addressed to the arXiv directly. (Addressing it here will serve no purpose as far as I can tell.) I would be fine with a variation on the following instead. Is there a user-side method to render latex formulas on arbitrary websites? $\endgroup$
    – François G. Dorais
    Commented Mar 29, 2010 at 3:11
  • 3
    $\begingroup$ Francois -- this post may be off-topic, but it may be helpful for the community, since it may attract the attention of those in charge of the archive. On the other hand I do wish there were an alternative to jsmath given the amount of bugs. $\endgroup$
    – algori
    Commented Mar 29, 2010 at 4:17
  • 7
    $\begingroup$ FWIW I've noticed that most LaTeX in ArXiv abstracts tends to have undefined macros in! That makes the problem much harder. You typically get things like "Consider the curve \$ \C:y^2=x^3+1 \$ . In \cite{BSD} it was conjectured that...". You're never going to be rendering this in your browser ;-) (unless your browser reads the preamble of the TeX source before processing the abstract!) $\endgroup$
    – Kevin Buzzard
    Commented Mar 29, 2010 at 6:54

1 Answer 1

Reset to default
5
$\begingroup$

You may be interested in looking at the following greasemonkey script:

http://www.gold-saucer.org/mathml/greasemonkey/

It doesn't look as though it's been worked on recently but that doesn't stop it being usable. The basic idea is to overlay a simplified latex->MathML converter on top of a webpage. Since it is a greasemonkey script, it is entirely user-controlled and can be applied to any webpage, arxiv or otherwise.

Of course, there is always the problem of non-standard macros in abstracts and I completely agree with the sentiment that abstracts should be 100% legible without requiring extra parsing.

(In addition to posting this answer, I am add my vote to the list of "to close" as I agree with the various sentiments on that line in the comment thread. That I'm answering the question is not paradoxical since this isn't actually an answer to the question that was asked. I'm not convinced that I would want to see the question to which this is the answer also posted on MO, but think that the easiest way to prevent it being asked is to answer it here. I'm community-wikifying this answer not because I want others to be able to edit it, but to underline my opinion by forgoing any reputation for this answer (not that I assume it would actually get any!).)

Improve this answer
$\endgroup$

Not the answer you're looking for? Browse other questions tagged .