4 ms·

I don't believe that MathML should be integrated in a browser: MathML does not requires a lot of CPU, so you can get away with JS+canvas that implements the Mat

by Jyaif 6y ago

I don't believe that MathML should be integrated in a browser: MathML does not requires a lot of CPU, so you can get away with JS+canvas that implements the MathML spec. By having it in the browser you are making the browser more bloated (more disk space, more memory, more cache misses) and possibly less stable/secure, even-though less than 1% of the people are going to use this feature. Other issues include increasing the barrier of entry for new browsers, having to deal with differences in implementation of MathML when using it, and making the spec harder to improve.

abhinav22 6y ago
There’s so much bloat in Browsers already though. And millions of school children study math, so there’s a big market for online mathematics. I agree with you that the solution needs to be a good one, you don’t want a standard that is difficult to implement across browsers or is inefficient / bad. I’m no expert so I can’t comment on MathML, but I am just hopeful that some standardized math format can be agreed on; and web standards have the best chance for this —- hence its important for Chrome to implement what is a W3C standard. My opinion anyway.
jacques_chester 6y ago
This is an argument against any and all changes.
- Jyaif 6y ago
  No. I'm in favor of features that enable new uses cases, either by providing a new API (e.g. WebBluetooth) or by massively increasing performance (e.g. WebGL, WebAssembly). Features that can be built on top of those foundations should not be shipped in browsers for the reasons that I mentioned before.
malthejorgensen 6y ago
Totally agree – KaTeX is good enough. Browser implementors should spend their time on other things.
- runarberg 6y ago
  Browser implementors have been spending their time on other things. People have been asking for chrome support for almost a decade now. I believe MathML support in Firefox is entirely due to volunteers. And as you see from this post Chrome support is being implemented by Igalia, not the Chromium core developers.
felixfbecker 6y ago
This opinion is completely ignoring accessibility. A canvas is a black box to non-sighted users.

by Jyaif 6y ago

abhinav22 6y ago

There’s so much bloat in Browsers already though. And millions of school children study math, so there’s a big market for online mathematics. I agree with you that the solution needs to be a good one, you don’t want a standard that is difficult to implement across browsers or is inefficient / bad. I’m no expert so I can’t comment on MathML, but I am just hopeful that some standardized math format can be agreed on; and web standards have the best chance for this —- hence its important for Chrome to implement what is a W3C standard. My opinion anyway.

jacques_chester 6y ago

This is an argument against any and all changes.

Jyaif 6y ago
No. I'm in favor of features that enable new uses cases, either by providing a new API (e.g. WebBluetooth) or by massively increasing performance (e.g. WebGL, WebAssembly). Features that can be built on top of those foundations should not be shipped in browsers for the reasons that I mentioned before.

malthejorgensen 6y ago

Totally agree – KaTeX is good enough. Browser implementors should spend their time on other things.

runarberg 6y ago
Browser implementors have been spending their time on other things. People have been asking for chrome support for almost a decade now. I believe MathML support in Firefox is entirely due to volunteers. And as you see from this post Chrome support is being implemented by Igalia, not the Chromium core developers.

felixfbecker 6y ago

This opinion is completely ignoring accessibility. A canvas is a black box to non-sighted users.