Re: Collection of examples suggestion

Hi Jeremy,

Just to avoid confusion: Community Groups cannot create any standards --
only W3C Working Groups (and I suppose the WHATWG) do that.

This group is primarily about improving the web platform from the
perspective of existing tools. So we're not after any particular (existing
or non-existing) standard for mathematical notation. If you like using TeX,
use tools that leverage that. Other people might (equally strongly) prefer
asciimath, MathML, or procedural notations. We have tools for all of these
(and more) and there's no need to abandon one for the others but there is
room to improve the web so existing tools can more easily achieve what
they're after.

Regards,
Peter.

On Mon, May 2, 2016 at 6:32 PM, Jeremy Kun <kun.jeremy@gmail.com> wrote:

> Hi Daniel,
>
> Feel free to use anything from the site. The mouse-hover text on the
> images show you the underlying source.
>
> About macros, it probably depends on what level the standard ends up at.
> Will be be Javascript? XML-like? A block like <math></math> in which one
> can write regular TeX? A new CSS-like domain-specific language written from
> scratch? (I like this last idea) Right now I'm just at the stage of
> thinking of features I'd like from the end result, more so than the means
> to get there.
>
> And we may decide that macros are better suited for an editor than a
> feature of a standard. But thinking about it further, I believe the root
> desire (in addition to not wishing to retype) is legibility and ease of use.
>
> In particular, I can't read a snippet of MathML to save my life. But if
> I'm writing an email to a colleague and I need to insert some math, I'll
> just type TeX and it works perfectly fine. For example: the claim is that
> f: X \to R has \lim_{x \to x_0} f(x) \neq \lim_{x \to x_1} f(x).
>
> So one feature of a standard I'd like is the ability to read the source
> without clawing my eyes out. Macros help this for particularly complicated
> documents, but again might be better suited for tools than the standard.
> Also, if we want semantic annotations to be part of the standard, it would
> make a lot of sense to have the option for document-wide annotations. Since
> when I'm writing a document it's probably always the case that if A is a
> matrix, it stays a matrix. I wouldn't want to have to write annotations for
> every formula involving A. The same underlying implementation could handle
> both of these features.
>
> On Mon, May 2, 2016 at 6:47 AM Daniel Marques <dani@wiris.com> wrote:
>
>> Hi Jeremy,
>>
>>
>>
>> Thanks for your contribution to the project. For sure we will consider
>> the way you are writing mathematics on the Web. Do we have permission to
>> take a few of your formulas for the purpose of the project?
>>
>>
>>
>> You say “Document-wide macros, the root problem being retyping and
>> re-modifying many times”. I know that LaTeX is programmable with macros but
>> I’m not sure to what extend people do that on Web pages. How do you do that?
>>
>>
>>
>> Dani
>>
>>
>>
>> *From:* Jeremy Kun [mailto:kun.jeremy@gmail.com]
>> *Sent:* viernes, 29 de abril de 2016 17:40
>> *To:* Daniel Marques; Peter Krautzberger
>>
>>
>> *Cc:* Moritz Schubotz; public-mathonwebpages@w3.org
>> *Subject:* Re: Collection of examples suggestion
>>
>>
>>
>> Another important example is probably Terry Tao's blog (
>> https://terrytao.wordpress.com/) and others who use the TeX2Wordpress
>> script you can find here:
>> https://lucatrevisan.wordpress.com/latex-to-wordpress/
>>
>>
>>
>> In other words, many math writers who use wordpress write entirely in TeX
>> and convert via this program to HTML.
>>
>>
>>
>> On Fri, Apr 29, 2016 at 8:37 AM Jeremy Kun <kun.jeremy@gmail.com> wrote:
>>
>> Hi all,
>>
>>
>>
>> I don't feel I have much to contribute to this group standard-wise, but I
>> have examples aplenty of how I display math on the web, and what kinds of
>> math I need to display on the web. I've been writing a math and computer
>> science blog for the last five years on Wordpress. Wordpress does something
>> quite weird: they accept a subset of TeX in their editor, and render each
>> to a static image. They do an acceptable but subpar job typesetting it, and
>> crucial features like equation aligning and matrix displays often break
>> with no traceable error. There's no support for commutative diagrams, and
>> when things break horribly I have to go and make my own images and upload
>> them by hand.
>>
>>
>>
>> Here's my main page: https://jeremykun.com
>>
>>
>>
>> And here's a post with lots of probability computations:
>> https://jeremykun.com/2013/10/28/optimism-in-the-face-of-uncertainty-the-ucb1-algorithm/
>>
>> Here's one that has commutative diagrams:
>> https://jeremykun.com/2013/05/24/universal-properties/
>>
>> Some big matrices used for algebraic topology:
>> https://jeremykun.com/2013/04/10/computing-homology/
>>
>>
>>
>> I tend to write 1-2 articles a month, around 4k - 5k words each, and over
>> the years I've built up a sizeable and complicated dependency graph across
>> the posts on my blog. One of the biggest complaints of my readers is that
>> they can't seem to navigate it all, and programmatically extracting
>> semantic information from the posts would allow me to write a program that
>> helps them.
>>
>>
>>
>> Some things that are very interesting and important to me as a writer:
>>
>>    1. Semantic/type annotations, which don't interfere with the actual
>>    math writing
>>    2. Document-wide macros, the root problem being retyping and
>>    re-modifying many times
>>    3. Maintaining the TeX math mode style of writing. MathML is just
>>    awfully verbose
>>    4. Full support for whatever subset of TeX math mode is used on
>>    mathoverflow.com
>>
>> As a side note, a close friend of mine (who is smart but not tech savvy)
>> is trying to launch a blog using Jekyll and Github and MathJax and all the
>> fancy tools. You can see how I probably would have done it if I were to
>> start a brand new blog today. Here's an example post:
>> http://samidavies.github.io/post/2016/04/01/Triangle-Removal_Lemma.html
>>
>>
>>
>> Cheers!
>>
>>
>>
>> On Fri, Apr 29, 2016 at 3:44 AM Daniel Marques <dani@wiris.com> wrote:
>>
>> Well… the idea is that the source code (html) of the page is as important
>> as the final rendering and mixing different ways to do math in a single
>> page would make difficult to understands the example.
>>
>>
>>
>> Regarding MathJax, I think that two different examples are valid. The
>> first one is an example with MathML on the page and the <script> tag. The
>> other one is the output (if I understand well that means the static HTML
>> and CSS). As another example, there is something like “accessibility
>> support such as volkers speech rule engine”. I do not know very much about
>> it but it is worth that somewhere it is explained because it might be of
>> interest for the CG.
>>
>>
>>
>> At this stage, I don’t think we are ready to provide a solution but
>> rather collect the existing know-how of the members. That’s the reason I
>> would prefer as many ways to do math as possible.
>>
>>
>>
>> My only concern is whether each example should follow a fixed template or
>> how they should be sort out.
>>
>>
>>
>> Dani
>>
>>
>>
>>
>>
>> *From:* Peter Krautzberger [mailto:peter.krautzberger@mathjax.org]
>> *Sent:* viernes, 29 de abril de 2016 11:16
>> *To:* Daniel Marques
>> *Cc:* Moritz Schubotz; public-mathonwebpages@w3.org
>>
>>
>> *Subject:* Re: Collection of examples suggestion
>>
>>
>>
>> Dani,
>>
>>
>>
>> Good idea! It will probably be relevant to have multiple representations
>> (different methods of creating HTML/CSS, SVG, PNGs etc). (Minor squibble:
>> MathJax does not fit into the list, I think, though its output does.)
>>
>>
>>
>> I would propose to start on the wiki but I can't seem to be able to find
>> it :-(
>>
>>
>>
>> I would like to propose to jump to GitHub right away, where we'd also
>> have a wiki.
>>
>>
>>
>> If people don't object, @Ivan would you have the time to set up a repo
>> for us?
>>
>>
>>
>> Best regards,
>>
>> Peter.
>>
>>
>>
>> On Thu, Apr 28, 2016 at 4:48 PM, Daniel Marques <dani@wiris.com> wrote:
>>
>> HI Moritz,
>>
>>
>>
>> I am not referring to an exhaustive list of the math that should be
>> covered but how math is put into Web pages. For example, using MathJax,
>> using SVG, using static HTML and CSS, using flexbox, etc. I would also like
>> to see examples with different ways to do accessibility (using alt or aria
>> attributes), etc.
>>
>>
>>
>> Dani
>>
>>
>>
>>
>>
>> *From:* Moritz Schubotz [mailto:schubotz@tu-berlin.de]
>> *Sent:* jueves, 28 de abril de 2016 16:19
>> *To:* Daniel Marques
>> *Cc:* public-mathonwebpages@w3.org
>> *Subject:* Re: Collection of examples suggestion
>>
>>
>>
>> Hi,
>>
>>
>> As promised on the (at least my) last MathJax meeting I'm working on that.
>>
>> Actually I'm working on that today.
>>
>> Currently I have the following lists of input all TeX sources
>> https://www.mediawiki.org/wiki/Extension:Math/CoverageTest
>>
>> https://github.com/wikimedia/texvcjs/blob/master/test/en-wiki-formulae.json
>> https://github.com/physikerwelt/utf8tex
>>
>> If you have more datasets I'm very interested.
>>
>> In addition I collected MathML input from input for inclusion in mediawiki
>> https://github.com/konrad/JATS-to-Mediawiki/issues/11
>>
>> Best
>>
>> Moritz
>>
>>
>>
>>
>>
>>
>>
>> On Thu, Apr 28, 2016 at 10:09 AM, Daniel Marques <dani@wiris.com> wrote:
>>
>> Hi Everybody,
>>
>>
>>
>> Great meeting and now we know each other better!
>>
>>
>>
>> My suggestion is to collect examples of how formulas are currently put
>> into a Web page. Each example would be a single formula written in an html
>> page together with any additional resource it depends (css, images, js). We
>> can store the examples in the github repo we talked in a previous email.
>> The examples should be as rich as possible (accessibility, styling, …)
>> depending on the feature is wanted to express.
>>
>>
>>
>> The idea behind the collection is that we understand better the different
>> solutions we are offering or we know from the members of the group. Later
>> on, when we try to write down the requirements, we will be able to point
>> into the examples.
>>
>>
>>
>> Regards,
>>
>>
>>
>> Dani
>>
>>
>>
>>
>>
>>
>> --
>>
>> Moritz Schubotz
>> TU Berlin, Fakultät IV
>> DIMA - Sekr. EN7
>> Raum EN742
>> Einsteinufer 17
>> D-10587 Berlin
>> Germany
>>
>> Tel.: +49 30 314 22784
>> Mobil.: +49 1578 047 1397
>> Fax:  +49 30 314 21601
>> E-Mail: schubotz@tu-berlin.de
>> Skype: Schubi87
>> ICQ: 200302764
>> Msn: Moritz@Schubotz.de
>>
>>
>>
>>