• ## Discussion Feed

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

• CommentAuthorHarry Gindi
• CommentTimeApr 8th 2010 edited

To improve readability, whenever I see LaTeX in titles, I remove it and replace it with unicode characters because it doesn't display properly on the front page, and having jsMath run on the front page significantly slows down the site (this was tried a while ago). If the expression is complicated, I often add in words to compensate for the missing symbols. I was wondering if something like the following could be added to the FAQ (I'm not going to claim that this is worded well):

Please try to avoid using LaTeX in the titles, since it does not render on the front page, and this makes it harder to read. If you do put LaTeX in the title, someone may edit it out to improve readability, so please don't be surprised if this happens.

Is there any opposition to something like this, or even the entire practice of editing LaTeX out of titles?

1.

My only issue with Unicode is that I don't know how it affects searchability. Whatever you replace LaTeX with, it should be searchable.

2.
I am actually annoyed that jsMath isn't on front page, and came to meta to post about it, and saw this post instead.
So, I am much against editing out latex in titles, it is easier to read and can accomplish math faster/slimmer than english.

I suggest having a button for "(re)process math" like in posts. I would always press it, as I assume others (with a fast computer maybe) will too.
• CommentAuthorHarry Gindi
• CommentTimeApr 9th 2010 edited

@Dror: This was already tried, and it made the site very aggravating to use. In particular, I have a very fast computer, but the problem is that while jsMath renders stuff, you can't click anything. It's also quite slow.

3.

I would also like to say that my computer is not all that quick - and I prefer to have something which looks a bit less slick but which loads much quicker and doesn't have this lag time that fpqc mentions. [Yes, I have drunk from the web-accessibility Kool-Aid; how can you tell? ;) ]

4.
Is it possible to do some prerendering on the server? Say, it stores the html code for jsMath fonts and also for image fonts, and on load the javascript to display them is tiny compared to jsMath - just see if user has jsMath fonts, if he does show first html code, otherwise, show image fonts code.

Wikipedia takes the root of having the server do all the work with a tex renderer. I assume this would be too much for MO servers. But prerendering html code just on titles might be fine, since titles are short and have a very small bit of latex.
• CommentAuthorHarry Gindi
• CommentTimeApr 10th 2010 edited

Having the jsMath fonts helps less than you would think. Also, jsMath in the titles looks bad because it's underlined, but it breaks the underline into two pieces of different vertical alignment.

5.
That can easily be fixed by applying a css style to the "typeset" class, making links inside have no underline. I say "typeset" after looking up the html in a specific question. There might be a few other relevant classes.
• CommentAuthorjonas
• CommentTimeApr 18th 2010

I don't see a problem with the formulas not rendering. Can't mathematicians read most of these simple LaTeX formula sources fluently? In fact, what's the easiest way to disable jsmath on pages other than the front page, for using in the case when some jsmath does not render well in my browser, so I can read the source instead?
6.

@jonas: you can prevent jsMath from rendering the page by clicking the jsMath button in the lower right-hand corner of your browser, clicking "Options", and unchecking the "Enable tex2math plug-in". This doesn't technically disable jsMath entirely, it just makes it so that it doesn't try to convert stuff between dollar signs to rendered math, but that's effectively the same thing on MO.

• CommentAuthorjonas
• CommentTimeApr 19th 2010

To Anton Geraschenko: thanks, that works.
• CommentAuthorMariano
• CommentTimeApr 23rd 2010

Harry, there is really very little point in replacing «Is $f`_*\mathcal O_X = \mathcal O_Y$?» by «Is the direct image of the structure sheaf on X isomorphic to the structure sheaf on Y when X->Y is flat and proper between smooth schemes over an algebraically closed field?». Such a loooooong title simply does not help: anyone on the site will be able to parse the original one and understand it with less effort than what is required to read your variant!

• CommentAuthorHarry Gindi
• CommentTimeApr 23rd 2010 edited

I dunno, the original title wasn't descriptive at all, and I didn't want to leave out any details! I'd rather not be responsible for ruining the title by making it less descriptive, so I erred on the side of caution (although yes, a bit excessively).

• CommentAuthorMariano
• CommentTimeApr 23rd 2010

You could have erred on the side of not doing the change, too...

• CommentAuthorHarry Gindi
• CommentTimeApr 23rd 2010 edited

Or removing the \mathcal-s and the \$-s. Either way, the original title could have meant anything, but now it will show up in searches.

7.

I actually agree with both Mariano and Harry. I think the original title wasn't descriptive enough (in general, I think people tend to pick pretty bad titles), but the new title is too wordy (making it hard for people to quickly make sense of it to decide whether to follow up). I propose the alternative "Is a flat proper morphism with connected fibers Stein? (i.e. is f_*(O_X)=O_Y)" It leaves out the bit about smoothness and being over an algebraically closed field, but it communicates the gist of the question. It also introduces the keyword that people would (or should) actually search for: Stein.

• CommentAuthorrwbarton
• CommentTimeApr 23rd 2010

I agree that the original title was not good. I always feel a bit cheated when the title forms a complete question but it is not the one the poster was asking (in this case because there were about eight conditions left out).

• CommentAuthorMariano
• CommentTimeApr 23rd 2010

Such a title would have had the added plus of informing me that that is the meaning of Stein! :)

• CommentAuthorHarry Gindi
• CommentTimeApr 23rd 2010 edited

@Reid: I think I only missed the surjectivity condition.

• CommentAuthorrwbarton
• CommentTimeApr 23rd 2010 edited

@Harry: In the body of the question, yes, but the original title omits all the conditions.