MathML:Cleanup: Difference between revisions
m (→DOM updates) |
No edit summary |
||
| Line 1: | Line 1: | ||
== [ | == [DONE] DOM updates == | ||
See {{bug|744783}}. | See {{bug|744783}}. | ||
| Line 5: | Line 5: | ||
[fredw]: The known issues are fixed. | [fredw]: The known issues are fixed. | ||
== GFX Glue Layer Removal == | == [DONE] GFX Glue Layer Removal == | ||
An important step for MathML is {{bug|663740}}, which will allow to start the work on OpenType MATH table, to support more [[MathML:Fonts|MathML fonts]]. More generally, see {{bug|430829}}, [[Gfx_glue_layer_removal]] and [http://groups.google.com/group/mozilla.dev.tech.gfx/browse_thread/thread/90d41db79b03f2fb/8921df6eaa49269c?#8921df6eaa49269c the thread on mozilla.dev.tech.gfx]. | An important step for MathML is {{bug|663740}}, which will allow to start the work on OpenType MATH table, to support more [[MathML:Fonts|MathML fonts]]. More generally, see {{bug|430829}}, [[Gfx_glue_layer_removal]] and [http://groups.google.com/group/mozilla.dev.tech.gfx/browse_thread/thread/90d41db79b03f2fb/8921df6eaa49269c?#8921df6eaa49269c the thread on mozilla.dev.tech.gfx]. | ||
== Attribute Mapping into Style == | [fredw]: the main refactoring {{bug|663740}} is done. | ||
== [DONE] Attribute Mapping into Style == | |||
We are using private attributes and pre-defined rules in mathml.css to map some MathML attributes into CSS. This approach shows its limits for length attributes ({{bug|330964}}) and exposes unwanted attributes to the DOM ({{bug|527201}}). We should rewrite this in a cleaner way ({{bug|69409}}). | We are using private attributes and pre-defined rules in mathml.css to map some MathML attributes into CSS. This approach shows its limits for length attributes ({{bug|330964}}) and exposes unwanted attributes to the DOM ({{bug|527201}}). We should rewrite this in a cleaner way ({{bug|69409}}). | ||
== Width Computation == | [fredw]: the private attributes have been removed. | ||
== [DONE] Width Computation == | |||
Our layout engine still has some problems with width computation. See for example {{bug|459363}} or {{bug|415413}}. Should we find a better algorithm? | Our layout engine still has some problems with width computation. See for example {{bug|459363}} or {{bug|415413}}. Should we find a better algorithm? | ||
[fredw]: the main preferred width bugs are fixed. | |||
== Box Model for MathML == | == Box Model for MathML == | ||
The <math> element is implemented as an inline-box, but this gives an incorrect height computation ({{bug|219873}}). We should also allow linebreaking inside <mrow>'s, as specified by the rules of MathML 3 ({{bug|508120}}). | The <math> element is implemented as an inline-box, but this gives an incorrect height computation ({{bug|219873}}). We should also allow linebreaking inside <mrow>'s, as specified by the rules of MathML 3 ({{bug|508120}}). | ||
Latest revision as of 10:00, 25 April 2014
[DONE] DOM updates
See bug 744783.
[fredw]: The known issues are fixed.
[DONE] GFX Glue Layer Removal
An important step for MathML is bug 663740, which will allow to start the work on OpenType MATH table, to support more MathML fonts. More generally, see bug 430829, Gfx_glue_layer_removal and the thread on mozilla.dev.tech.gfx.
[fredw]: the main refactoring bug 663740 is done.
[DONE] Attribute Mapping into Style
We are using private attributes and pre-defined rules in mathml.css to map some MathML attributes into CSS. This approach shows its limits for length attributes (bug 330964) and exposes unwanted attributes to the DOM (bug 527201). We should rewrite this in a cleaner way (bug 69409).
[fredw]: the private attributes have been removed.
[DONE] Width Computation
Our layout engine still has some problems with width computation. See for example bug 459363 or bug 415413. Should we find a better algorithm?
[fredw]: the main preferred width bugs are fixed.
Box Model for MathML
The <math> element is implemented as an inline-box, but this gives an incorrect height computation (bug 219873). We should also allow linebreaking inside <mrow>'s, as specified by the rules of MathML 3 (bug 508120).