The first grant round of the TeX Development Fund was completed in spring 2003. Since then, applications have been accepted on a rolling basis.
Descriptions of the selected projects follow, in reverse order of receipt. In all cases, please realize that we generally are not fully funding the work being described. The full descriptions are for background information, not to be taken as the grant contract.
We continue to be especially grateful for the level of support for the development fund from individuals in the TeX community: the fund has received donations from more than 200 individuals from around the world. We gratefully and publically acknowledge the donors here.
In both 2007 and 2008, TUG received a substantial grant of $US10,000, with another US$10,000 in matching funds, from an anonymous donor. Thanks to generous support from many individuals and companies, all those funds have now been matched. We will publicize future matching opportunities as usual.
The development fund committee members wrote a roadmap for TeX development, published in TUGboat. It emphasizes three of the projects below: the LuaTeX extension, the TeX Gyre fonts, and the TeXworks front-end. The document contains background and rationale information.
All published reports:
Most of the projects here are ongoing, and contributions to further the work are most welcome!
TeX development fund committee
Vafa Khalighi, Australia.
Amount: US$1000; date: 1 March 2018.
Porting the bidi package to LuaTeX. Description from the author:
The bidi package is a rich tool which is used around the world for typesetting theses, books, journal and conference papers, and any other academic work. As an example, this mathematics Ph.d. thesis from Sharif University is typeset with the xepersian package, which relies on the bidi package for bidirectional typesetting.
Another good showcase of the use of bidi package is the book From Alexandria, Through Baghdad published by Springer, about history of mathematics.
The bidi package modifies the LaTeX core for proper right to left typesetting, and also changes over 100 packages so that they can be used for right to left typesetting. The bidi package also implements unique TeX solutions for overcoming TeX--XeT limitations and bugs in right to left typesetting. One basic example of such limitations is that whatsits are reversed in right to left in the TeX--XeT model, so \special where it is used for color, graphics, etc., has issues with TeX--XeT. This is handled by the bidi package.
Luigi Scarso, Italy.
Amount: US$1000; date: 28 December 2015; completed 24 May 2017.
Bug fixes and maintenance for MetaPost path resolution, binary, decimal, and double number systems. Also, more efficient integration with mplib by excluding the libraries needed only in LuaTeX. May 2017 report.
Khaled Hosny, Egypt.
Amount: US$2000; date: 24 December 2014; completed 1 February 2016.
Building an OpenType math companion for Linux Libertine and Linux Biolinum fonts (ultimately named Libertinus), and fixing bugs in those fonts. Also coordinating with the authors of Linux Libertine (La)TeX support files to adapt the new and fixed fonts (ultimately they did not respond, so upstream Linux Libertine is unmaintained as far as we can tell). (repository, CTAN package).
Clerk Ma, China.
Amount: US$1000; date: 9 August 2013.
(All information here is from the applicants.) Most math books in China are produced by Founder Bookmaker. This system have used a set of Russian style math font for more than 30 years. These commercial fonts are designed with an unique encoding by Founder. And, these fonts cannot work in TeX or other programs.
We have a set of metal types which contain two Russian style fonts (serif and sans serif). By analyzing these metal types, we find Founder's fonts are derived from these fonts, and Founder only provided serif version (we will providing these math fonts in serif and sans serif). These metal types were imported from the U.S.S.R. in 1953.
We will trace the metal fonts to outlines (initially in EPS format). For more detailed adjusting, we will be using FontForge. Parts of our Chinese fonts are already processed in this workflow. For these Russian style fonts, we will also work in this way.
A short description of the Fandol fonts was published in TUGboat 34:3.
Clerk Ma, China.
Amount: US$2000; date: 26 June 2013.
Add a native editor and package manager GUI to TeX Live for Android. (See also Clerk's TUG 2013 presentation.)
Applicant: Marco Müller, Switzerland.
Amount: US$1000; date: 20 June 2013; completed 10 August 2014.
Enhance the Metaflop web application, which provides a graphical interface for adjusting Metafont parameters, with improvements to the underlying fonts, the preview mechanism, and the generation.
Applicant: Luigi Scarso, Italy.
Amount: US$2000; date: 31 May 2013; completed 11 March 2015.
Support shared libraries in LuaTeX using SWIG: overview, development site, proposal. Several libraries are already supported, e.g., mysql and graphicsmagick.
Applicant: Khaled Hosny, Egypt.
Amount: US$4000; date: 24 April 2012; completed 25 July 2013.
For updates to xetex, especially relating to OpenType math typesetting, and including updates as needed to luatex to keep the engines in sync. Other areas of work include finding fonts and syncing xdvipdfmx with dvipdfmx, as well as handling general bug reports. Report on the completed work.
Applicant: Uwe Lueck, Germany.
Amount: US$1000; date: 17 September 2011.
For updates to lineno and related packages.
Applicant: Idris Hamid, USA.
Amount: US$3000; date: 10 August 2010.
Focus on complex bi-directional issues in (Lua)TeX. Proposal.
Applicant: Taco Hoekwater, The Netherlands.
Amount: US$2000; date: 2 December 2009; completed 24 May 2011.
Implement better numerical handling in MetaPost, among other enhancements. Proposal. The initial article MetaPost 2 project goals by Hans Hagen and Taco Hoekwater was published in TUGboat 30:3.
Applicant: Taco Hoekwater, The Netherlands.
Amount: US$1500; completion date: 27 November 2008.
Implement SVG as a backend in MetaPost 1.200. This project was co-sponsored by a generous contribution to TUG made by Dave Crossland. Information is available in the MetaPost manual.
Applicant: Taco Hoekwater, The Netherlands.
Amount: US$1500; completion date: 28 November 2008.
Implement MetaPost as a library (MPlib) as well as an executable; done in MetaPost 1.100.
The article MetaPost as a reusable component was published in TUGboat 28:3, and MetaPost developments: MPlib project report and (by Hans Hagen) The MetaPost library and LuaTeX in TUGboat 29:3.
Applicant: Khaled Hosny, Egypt.
Amount: US$1000; completion date: 5 December 2011.
Bulaq Press, established in Cairo in 1820, has developed one of the most widely used Arabic typefaces that has been a standard in Arabic printing for more than 150 years. However, computer typesetting is yet to feature a fully conformant digitized version of that typeface; only a few proprietary fonts come close.
This project aims to digitize the Bulaq (Amiriya) Press typeface in the form of an OpenType font that implements all contextual features of the original typeface as well. Also, the project will work on extending it to cover other languages using the Arabic script. Finally, the project will consider writing any macro packages or support files needed to use the font in Arabic-capable TeX engines such as XeTeX and LuaTeX.
Applicant: David Crossland, Great Britain. Amount: US$500; date: 3 September 2008.
TUG is also contributing administrative support to this project. Report, 6 January 2009.
A summary article The Open Font Library was published in TUGboat 30:1.
Applicant: Jonathan Kew, Great Britain.
Amount: US$5000; date: 1 September 2008;
Amount: US$5000; date: 4 January 2008;
Amount: US$5000; date: 4 October 2007; stable release made 5 October 2009 for TeX Live 2009.
The TeXworks project is a new cross-platform front-end for TeX, in the spirit of TeXShop.
The target for the 1 Sep 2008 grant is a formal 0.1 release to supplement the snapshots already being made. It will include single-instance behavior for Windows and GNU/Linux/X11, at least one user interface localization, as a proof-of-concept and example for other translators, and package the whole thing in an appropriate installer package for Windows and Mac. (GNU/Linux distributions have their own package managers and prefer to work from sources; TeXworks is already available for Ubuntu.)
The article TeXworks: Lowering the barrier to entry was published in TUGboat 29:3.
Applicant: Basil Solomykov, Russia.
Amount: US$1000; date: 7 August 2008.
The Obyknovennaya Novaya (“Ordinary New Face”) typeface was widely used in the USSR for scientific and technical publications, as well as textbooks. The current implementation is in Metafont; the author aims to provide an outline version as well.
This was completed by 7 May 2014, with the release of the obnov package on CTAN, and an article to be published in the next TUGboat.
Applicant: Hans Hagen, Taco Hoekwater, The Netherlands.
Amount: US$5000; date: 19 September 2007.
Amount: US$2000; date: 29 March 2007.
The Oriental TeX project is a major effort to extend TeX to be suitable for Arabic scholarship. The implementation vehicle is LuaTeX. Colorado State University has funded the major portion of the development, under a grant proposed by Professor Idris Samawi Hamid in the Department of Philosophy. This allocation by TUG is a portion of the matching funds for the CSU grant. We gratefully thank Dr. Hamid and CSU for such strong support of this initiative in TeX development and infrastructure.
Applicant: Boguslaw Jackowski, Janusz Nowacki, Poland.
Amount: US$3500; date: 13 February 2008.
Amount: US$2500; date: 20 February 2007.
Amount: US$1000; date: 22 September 2006.
The TeX Gyre typefaces extends several notable free (libre) font families to other character sets in much the same way that Latin Modern extends Computer Modern. This is an ongoing, multi-year, project; the dates above are the completions of intermediate installments.
An introduction to the project was published in TUGboat 27:2. An update on GUST e-foundry projects was published in TUGboat 38:2.
Applicant: Alex A.J., India.
Amount: US$600; acceptance date: 9 November 2005.
The project aims to develop a package for Malayalam typesetting using the Omega system. Proposal.
This was completed by 1 February 2006. The report Typesetting Malayalam using Omega was published in TUGboat 27:2.
Applicant: Raph Levien, USA.
Amount: US$1000; acceptance date: 30 November 2005.
Completion and release of the Inconsolata font. Only one style is anticipated (no bold or italic is planned). Glyph coverage is to include Latin 1, 2, and 9, with a few other glyphs useful for TeX. Metrics and encoding will be tuned to make it as easy as possible to use Inconsolata as a drop-in replacement for cmtt.
Released in 2009 with (La)TeX support, discussed in the author's interview.
Applicant: Federico Garcia, USA.
Amount: US$1000; acceptance date: 11 April 2005.
Design of algorithms and code implementation for the first stage of TeXmuse project for musical typesetting.
The `first stage' consists of code that is able to typeset the basic musical text of Bach's 15 inventions. These pieces are for piano and only two voices: two staves and one voice per staff. Being from the Baroque, they feature interpretative notation (slurs, articulations, etc.) only in a very limited way. All of this makes these pieces a very good first stage in the development of TeXmuse.
This was completed by 31 August 2005, with TeXmuse uploaded to CTAN. The related article On musical typesetting: Sonata for TeX and Metafont, Op. 2 was published in TUGboat 24:2.
Applicant: Hrant Papazian, USA.
Amount: US$3000; acceptance date: 19 October 2004, eight-month term.
Design and implementation of a Baskerville typeface revival to high standards of typographic quality, historical sensitivity, and usability.
The typeface family will include two weights (regular and bold), each with a true italic. The fonts will cover the following three character ranges: Basic Latin, Latin-1 Supplement, Latin Extended-A.
(This project was later terminated at the request of the grantee, and money refunded.)
Applicants: Boguslaw Jackowski, Janusz Nowacki, Poland.
Amount: US$2000; acceptance date: 19 May 2004.
Continuing enhancement of the Latin Modern character set, including support for Vietnamese, Navajo, and other Latin-based scripts.
This was completed by 20 April 2004, with the Latin Modern 0.98.3 release. The related article Latin Modern: Enhancing Computer Modern with accents, accents, accents, was presented at TUG 2003 and published in TUGboat 24:1.
Applicant: Hàn Thế Thành, Vietnam.
Amount: US$1500; acceptance date: 26 March 2004.
1) New primitives to provide more control over the quality of
typesetting complex documents (feedback as well as manipulating the
result of breaking paragraphs into lines).
2) a primitive to ease the use of font expansion with pdftex, so one can use font expansion having expanded TFM's (which are complicated to generate for an average user).
This was completed by 14 October 2004, pdftex 1.20a includes this work. The related article Micro-typographic extensions of pdfTeX in practice was presented at Practical TeX 2004 and published in TUGboat 25:1.
Applicant: Gerben Wierda, Netherlands.
Amount: US$1500; acceptance date: 3 November 2003.
Make source release of new version of i-Installer, the engine used for installing and configuring GW's MacOSX TeX distribution.
This was completed by 28 February 2004, the new version is available, along with an article in TUGboat.
Applicant: John Plaice, Australia.
Since 1998, the Omega Project has been capable of generating MathML and XML directly from the typesetting engine. In this project, we propose to further develop and polish the XML- and MathML-generation capabilities of the Omega Project. The resulting code and macros will be distributed with all future Omega releases.
The Omega approach to generating markup languages from TeX input consists of two parts:
In this project, we propose to comprehensively cover the high-level LaTeX and AMS-LaTeX macros and define a matching DTD/Schema, and ensure that Omega can correctly translate a correct LaTeX document with mathematics into XML and MathML. High-level macros will be written, new macro primitives will be defined, and modifications will be made to the typesetting engine.
Applicant: John Plaice, Australia.
There are currently three large extensions to TeX:
Already e-TeX and pdfTeX have been combined into pdf-e-TeX, and more recently Giuseppe Bilotta has created e-Omega.
In this project, we propose to combine the key elements of e-TeX and pdfTeX into the Omega Project. In addition to combining several Pascal Web change files and integrating the associated C/C++ code, an important objective will be to harness the power of Omega's Translation Processes and context manipulation code to generate high-quality PDF files. The resulting code and macros will be distributed with all future Omega releases.
Applicant: David Kastrup, Germany.
The project described here is very large. Only a small part is funded through this grant: making it possible for the main work to be included on CTAN and integrated into the main LaTeX sources.
Thus, the main description here is provided for background information only and is not a description of the work that will be done for the grant.
The funded subset was completed around 16 July 2006, and is available from CTAN as the bigfoot package. An article was published in the EuroTeX 2006 proceedings.
The humanities have very arduous typesetting needs that are not accommodated by almost any available typesetting tools. The current workflow for this reason usually implies preparation of the various text apparatus in separate files, using arbitrary marks for establishing the correspondences between source text and critical apparatus, and then sending the thing to the publisher in some format such as Word and wait for intermediate typesetting results. The whole process is slow and error-prone. In particular the typesetting of multilingual sources (such as Biblical Hebrew, Greek, Latin, or even more recent languages) requires often skills for the proper typesetting that only the author of the texts has, making the whole production cycle very cumbersome.
The few available existing systems for such typesetting are so cumbersome with regard to text editing and processing as to make managing LaTeX sources appear laughably easy in comparison.
For that reason, giving LaTeX the necessary capabilities for easy typesetting and processing of such texts would make it an ubiquitous utility in the humanities. Since it would cause a sizable user base from outside the computer community to get acquainted with LaTeX, it would spawn a lot of interest in user-level introductory material and customizable editors, leading to a lot of possibilities to acquire further funds and jobs focused around making LaTeX better accessible and teaching its use.
Accommodation of the footnote apparatus
Critical editions usually contain multiple footnote apparatus. A typical set for an edition of a commentary would be
Of course, this is just a simple example: much more contrived apparatus can be seen, too. Still, there are some problems to be solved that are difficult to tackle: footnotes in apparatus may attached to footnotes earlier on the page. Since the text of a footnote appears lower than its anchor in the main text, proper ordering of footnotes might imply that the order of footnote anchors in the source text differs from that in the page layout: a footnote in a footnote anchored to the main text must appear below any footnote anchored to a later portion of the main text on the same side! Footnotes must be relocated to the next page if their reference inside of another footnote happens to fall on the next side, and the numbering must completely be changed, consequently. Only the last on-page footnote of every apparatus may be wrapped at all. But that means that wrapping a footnote from the main text might be allowed or not allowed depending on whether a footnote referring to another footnote happens to appear on this or the next page.
In short, a mess to be sorted out with multiple passes, and with a well-designed stable algorithm optimizing the amount of material fitting on the page given the necessary constraints.
In the first stage of the project, a separate footnote style will be designed that overrides only small parts of the standard LaTeXe output routine, probably building upon the ncctools package.
While LaTeX provides for margin notes and paragraphs, the mechanism is not versatile enough to cater for either margin notes in footnotes or multiple levels of margin notes (such as one column of verseslash line numbers, flanked by one column of cross references to other chapters/verses in the outer margin and page numbers for a different edition in the inner margin).
The possibilities for editions of course are limitless, nevertheless there are basic building blocks from which a page layout may be built up. The current LaTeX output routine does not accommodate such formats, neither would it be useful to support something like that as part of the base class. However, there are a lot of elements that can be systematically tackled and given interfaces, so that the average document designer would not have to program everything himself, but merely aggregate boxes, insertions and their processing in a reasonably easy way.