Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.
1 to 11 of 11
If you go to your user page and click on the "prefs" tab, there is an option to render math on all pages. As it mentions there, this is only a cookie, so you will have to do it for each browser / computer you use.
Should the default value of that preference be switched? In other words, are there more people who are annoyed by LaTeX rendering on the home page, or more people annoyed by it not rendering?
I prefer to have it render. Then again, I can also imagine an argument if LaTeX in titles rendered by default, then people would use more math in titles, making it more annoying to watch MathJax process the home page again (this was why rendering in titles was originally "turned off").
No! Do not make it render on the main page. The main page at Maths-SX is almost unreadable because of this "feature".
You forget that not all of your users live in California. Even on my high-speed internet connection in my office, MathJaX is painfully slow. When I look at the main page of Maths-SX or MO, I just want to get enough information to know whether or not to click on a question. So I don't want to have to wait for the maths to render (and remember that while it is processing, the maths is hidden from view).
We should think about speeding up MathJax generally. @Andrew, can you determine if it's the load time for the MathJax scripts, or something else? If it's the load time, we can put more effort into serving MathJax from a faster server.
Scott, if you tell me what to do then I'll run whatever tests you want.
Given that the problem also occurs on Maths-SX, I'd say that it's a general MathJaX problem than specific to MO. (I certainly notice the problem more on Maths-SX, but that might be due to other factors.)
(insert my standard rant about MathJaX being the wrong solution to the problem here)
@Andrew: it's been forever since you emphatically denounced some possibility I brought up. :-)
The question here is whether the default should be that the home page is rendered. There would still be the option to turn off the rendering through a cookie.
What Andrew Stacey said.
Regarding the question of default (admittedly I am biased, but I believe there is some objective component to it):
The likelihood that one has problems (slow or something else) with MathJax is I'd say typically higher when one visits the page from a computer/browser one does not frequently use for this purpose, say some computer in some internet cafe, library, whereever, and on which one could not or did not change the (system dependent) default.
For first time visitors -- who happen to have, but might not yet even know, a MathJax problem -- it could also be an advantage if the frontpage definitely works, with access to the FAQs and so on. As opposed to a situation where you visit a new site and the first thing that happens is something 'strange' you might likely not understand.
Thus, I think it is good that the default is, as it is now, 'off'.
Anton: What an_mo_user said.
Shogun70: I'm in Norway, using Firefox on Linux. Right now, I'm using FF4.0.1 at home; in my office I have FF3.something. I'm happy to run any diagnostic tests you want to help. If you want to contact me directly, you can find my email fairly easily: I'm in the mathematics department at NTNU (Trondheim). My homepage is linked from my MO user page.
(You should probably know at the outset that I think that MathJaX is an amazing piece of software, but I consider it to be the wrong solution to the problem of putting maths on the internet and that conversion to MathML server-side is the right solution. I think that MathJaX's role is to get everyone used to seeing maths on the internet in a decent form (ie not as images) as an easier route to getting MathML more widespread. With that in mind, what I regard as "painfully slow" might be "reasonable" for anyone else. I compare it with raw MathML which is, of course, much faster. But as any way of getting maths on the internet is better than none, and the general question of doing so is something that interests me greatly, I'm more than happy to help you improve MathJaX if I can.)
I'm convinced. Thanks Andrew and an_mo_user.