Re: [mathjax-users] Require Fast MathJax

[David Farmer]

Dear Ramakrishna,

I hope Mr.Peter and Davide are not offended by you writing
that you hope they "listen and take notice".

I cannot speak officially as I am only a user, but I think your
plan cannot work, because it is impossible to say what should be
in the "mobile version."

I deduce that Katex is not adequate for you, because if it was
then you could just use it and you wouldn't need MathJax. So that
is an example of a minimal version that is too minimal.

So what should be there? Integrals? Is all of calculus part
of the minimal distribution, or none of calculus? Do you need
arbitrarily scaling parentheses and radicals, or is it okay
if the equations don't look nice?

Are they supposed to change their slogan to "Pretty good looking
math in some browsers"?

Maybe you want them to make it easy for users to pick and choose
which exact functionality they need. Assuming that is possible
(well, it is possible if you fork the code and do that yourself),
I really doubt that many people would make use of it.

I'm sure you could make your personal wish list of what should be
in the minimal version, but I doubt that would cover the majority
of people who want their own minimal version.

Regards,

David


On Fri, 23 Sep 2016, Ramakrishna Salagrama wrote:

> Hi,
>
> When speed is becoming one of the important factors in page ranking, and Mobile users are growing ever
> stronger, I think MathJax should also load within 300 to 500 ms instead of 3 to 4 seconds.  When pitted
> against Mathjax with Katex, Katex wins hands down in speed.  It loads very very fast.  Some say that its
> (Katex) functionality is limited (No accents etc.) but most of the websites use basic maths which does not
> require complex math setting.  
>
> One advantage with Mathjax is its set up is very easy but with improved speed, it should be de facto
> rendering code for every one.  I use mathjax to render on my site and it loads fast as I optimized my site in
> every direction and there is no single JS script loads before DOM loading except mathjax.
>  http://www.campusgate.co.in/2011/10/number-system-factors-and-coprimes.html
>
> So what I think is, why don't we have a lite or mobile version which has basic math display with less kb like
> KaTex, So that 90% of the websites may use this light version to display maths.  If any webpage require high
> level maths, they should be a setting that it should automatically fall back on using full version like math
> stackexchange and other scientific sites.
>
> I hope Mr.Peter and Davide listen and take notice of this.  
>
> Thanks and Regards
> Ramakrishna
>
> --
> You received this message because you are subscribed to the Google Groups "MathJax Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
> mathjax-user...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>
>

[Ramakrishna Salagrama]

Hi David, 

My intention is not to offend anyone.   I feel sorry If is sounds wrong. 

Regards

Ramakrishna

[Peter Krautzberger]

Hi Ramakrishna,

Thanks for sharing your opinion. 

I believe the team is well aware that it would be beneficial if MathJax were significantly faster. We are continuing to look into performance improvements, in particular as part of the MathJax v3.0 rewrite (which has been going through various planning phases this year and will begin early next year). Still, as a non-profit open-source project, our development resources are often limited.

As to your suggestion, I'm afraid it is not really possible to "strip down" MathJax's functionality to make it faster; the various parts of the layout process are too interconnected. There are also structural differences between how MathJax operates and how other math rendering libraries operate which make comparisons difficult. There are also some limitations outside our control, in particular the limitations of the web as a platform and browser implementations in particular; this might be a good example.

I think it's fair to say that MathJax has had a good track record of improving performance over the years but that we all want to it do better. Still, I would point out that there are some goals which rank higher than performance for us, most importantly layout quality but also specific types of features (such as Unicode support).

Finally I would say that, from a performance point of view, the biggest advancement in the last two years has been the introduction of mathjax-node, our APIs for using MathJax in NodeJS (i.e. on the server). For static content such as your site, I would always recommend this first.

Regards,

Peter.

To unsubscribe from this group and stop receiving emails from it, send an email to mathjax-users+unsubscribe@googlegroups.com.