Math on the Web ? Not Only ! I prefer to say Math with Web Technology.

 

Hi, 

I hope that all of you who came in Lyon enjoyed their stays in our town.


As I said before OpenJFX (JavaFX) MathML support is repaired. All
developper can used it from now with Java 8.192 and JavaFX 11/OpenJFX
11. They can reach Mac OS, Windows, Linux, ... so their can work with
Macs, PCs Raspberrys Pi, ... everywhere where JavaFX is supported and
may be mobiles with IOS and Android soon : 

I missed the meeting in Lyon because I was working on the second part of
my project, I'll wished to be ready for TPAC-lyon-2018 but I not succeed
:-(. 

I'm working hard because I hope to make another Pull Request to the
github javafxports/openjdk-jfx for the next release. 

- This work is tracked by Github Scientificware OpenJDK-JFX #issues26
[2]. 

- To validate my choices (described in my first mail), I build another
application called FXMessages. It's a mobile messenger application based
on SMS (that I also described precedently). I'm targeting IOS and
Android. The Android version [3] is near to be released (may be in one
week) and I'm going to test it with secondary school (for vocational
training) and also preparatory classe (two-year undergraduate intensive
course in mathematics and/or physics) students. 

___ 

The purpose of this application is editing only(#issue26). I already
worked on display and I hope my work will be available on Android as
soon as possible. But this part is Gluon's project [4] and I discussing
with them in MathML on JavaFX/Android or IOS #228 [5] and as you can
read in this report I've got a short term solution. This present
FXMessages version is based on CSS and the rendering is very bad. 

Guy. 

P.S. I'm not from Oracle and Gluon, I'm just a small contributor who
needs mathematics to work :-) 

----------------------------------------------------------------------------------------------------------------------------


Le 2018-06-22 10:20, Abossolo Foh Guy a écrit : 

> Hi Arno,
> 
> After our short discussion, I continued to work on WebKit support in JavaFX / OpenJFX.
> 
> I posted the Issue (JavaFX OpenJFS and MathML #71 [1]) on the new OpenJFX GitHub Contributor process and started to work as suggested by OpenJFX team.
> 
> I found and fixed the problem. I'm waiting now to get my pull request merged into OpenJFX. Some problems still remain but the support of MathML it's far better than in the previous version. I can now use it in my own applications. I hope it would interest others developpers.
> 
> On the picture, the reference goal TeX is on the left and the MalML rendering JavaFX/OpenJFX is on the right.
> 
> The small application for the test is not simply a web browser it's an editor (the HTLMEditor of JavaFX/OpenJFX). It can made some edits in formulas but we can not presently create formula. Many problems have to be fixed before we can use it. AsciiMath as authoring language seems to be a good entry point.
> 
> Of course one can use only the web browser JavaFX/OpenJFX component (WebView).
> 
> The test is from the Mozilla MathML Torture Test.
> 
> I hope to see it in the next OpenJFX release, I really need it.
> 
> Best regards.

----------------------------------------------------------------------------------------------------------------------------


> Hi Arno, 
> 
> Thanks for your answer, 
> 
>> ​That's an interesting use case. It might take a while to get there, given that SMS doesn't even support bold text, but a good use case to keep in mind. It might be easier to get support in other messaging platforms before SMS (SMS is tied to telecom standards that are slow to evolve).
> 
> By SMS, I mean, to use only its sequence of characters and to display it in a specific SMS Messenger. I agree with you that telecom standarts are slow to evolve that's why we need a specific messenger. 
> 
>> The status of JavaFX is a bit murky, but OpenJDK might be a way to contribute. I also note that the issue above has 0 votes. Furthermore, the issue seems to be with the MathML support in the host browsers, so really, until the host browsers (Chrome, Safari, IE) support MathML natively, the chances of seeing a fix for this bug are low.
> 
> I wrote "JavaFX", but in fact, I mean "OpenJFX" its open source version. OpenJFX includes a web browser "WebView" built over webkit that's why it supports MathML. 
> (https://docs.oracle.com/javase/9/docs/api/index.html?overview-summary.html) 
> But I don't understand why with the same webkit version, it works well on Safari and not on OpenJFX. As suggested by team OpenJFX leader, (http://mail.openjdk.java.net/pipermail/openjfx-dev/2017-December/021113.html) I tried to see where is the problem, but it's not pure Java and build OpenJFX is a bit difficult than build OpenJDK. 
> 
>> - is the height of the container that needs to be surround by the fence, within certain margins? 
>> - if so, use a single glyph representing the fence: U+007d "{" 
>> - if the height is above those bounds, create a "stacked" fenced, by putting a glyph at the top U+23a7 "⎧", a glyph at the bottom U+23a9 "⎩", a glyph in the middle U+23a8 "⎨" and for the rest of it, a "repeating" glyph, U+23aa "⎪" (and so on for all the different types of fences)
> Before choosing a way to proceed, I tried to understand how TeX works. There is several prebuilt sizes of each fence (3 ou 4). TeX first tries to use these sizes and if it needs hiegher, TeX builts the "stacked" fenced as you discribed. But what about italic and other orientations ! ?
> So,I wonder why nobody tried to sent a simple font file with the right sized glyph draw to the glyph engine (Harfbuzz or other) as TeX but for each sizes and not only for the first four.
> I think that drawing a bracket as a curve is simpliest and economics than the algorithm that puts glyphs over other glyphs.
> That's the way I choose to proceed, I directly draw "{", "(" and "[". The inconvenience, is that I have to create them by hand and then to draw them but without use the glyph engine (a bit complicate for me because I don't know how to built a font file). Fortunately Java can draw it easily. The advantage, I have got symetric, italic and horizontal glyphs too, with the same curve.
> Best regards. 
> Guy.

----------------------------------------------------------------------------------------------------------------------------


> Le 2018-01-25 20:32, Abossolo Foh Guy a écrit :
> 
> Hi !
> 
> I try to follow your discussions. I wanted to post a small commentary to try to progress. I hope not to disturb this exchange thread.
> 
> The mathematic notation on web and others support like mails ou sms are the user adoption. I'm a teacher and my dream is to exchange mathematic SMS with collegues or students.
> 
> On my authoring tool that I try to developp, I realised that I need three interfaces :
> - An user interface : "Write As You Speak or Heard", "3 times 12" in english, "3 fois 12" in french, ... . I think this is the only way to be fast adoped.
> - An internal interface to display or save formula : MathML presentation markup or Latex.
> - And finaly an other internal interface to exchange formula with other applications : MathML content markup.
> 
> Common Users don't have to deal with complex Markup and we need them to generalise mathematic on web : JavaFX support MathML but the display has been broken since 7 ans. The developpment team answer me that it not a priority bug !
> 
> https://bugs.openjdk.java.net/browse/JDK-8147476
> 
> I have a dream, I hope that one day, many angry user mails will change this :-)
> 
> Displaying streching characters is also a problem. I found that Java/Swing solution was interresting. In some configurations it's allow to mixed glyphs from fonts and user glyph drawing. In fact all glyphs are curves, we can use a specific application (like Harfbuzz) to draw it or draw "by hand" for complexe glyph. I don't understand why nobody achieved to extend Donald Knuth approch with TextFont. As you discribe it, most of rendering mathematic application use different font sizes first and many glyphs for higher characters (exactly as Knuth several years ago). But computers speed is faster than in the past and we should be able to draw directly stretching glyphs.
> 
> Guy.

 

Links:
------
[1] https://github.com/javafxports/openjdk-jfx/issues/71
[2] https://github.com/scientificware/openjdk-jfx/issues/26
[3] https://play.google.com/store/apps/details?id=com.fxmessages
[4] http://www.gluonhq.com
[5] https://github.com/javafxports/openjdk-jfx/issues/228