Release early, release often!

Linus Torvalds

• 2.0.b6 - Fix support for base header level with Setext-style headers (thanks to Rob Walton); improve Windows support;

• 2.0.b5 - spaces at end of xslt filenames won't cause failure; use \url{} for "non-referenced" url's in LaTeX to allow linebreaks (though they still don't always break correctly --- this is a problem with hyperref not MMD); don't convert ^ to exponents in the clean-text-allow-latex.xslt file so that math code works properly; the S5 XSLT file at least partially works again now; update the TextMate bundle to work with Leopard; updated the envelope and letterhead files; include 6x9book-real-poetry XSLT that uses memoir's poetry features fairly well; rework the clean-text files to make them easier to update in the future and more modular; XHTML comments are now passed through as raw LaTeX; unescape encoding within comments;

• 2.0.b4 - empty labels for headers now produce valid XHTML (e.g. no id=""); fix bug in clean-text.xslt that caused a problem with closing double quotes; the .xslt extension is no longer required in metadata; added customizable letterhead XSLT; fix bug in table support that choked on extra spaces at end of lines; Major Change: switched to Text::ASCIIMathML for math support, meaning that everything is once again perl based (this enables math features on web sites using MultiMarkdown, for example); fix bug that occurred when 'Abstract' was not the first chapter;

• 2.0.b3 - move the clean-text routine from xhtml2latex.xslt into it's own file (to allow easier modification by users); create alternate version that does not protect certain characters in order to allow raw LaTeX code to be passed through; added latex-snippet.xslt stylesheet for inclusion in outside LaTeX template systems; added xhtml-poetry-support.xslt and xhtml-toc.xslt to demonstrate how to extend MMD functionality for XHTML output with new system; fix bug in SmartyPants that processed typography within <style> sections (thanks AmberV); fix handling of links by reference in headers and handling of attributes when links are referenced multiple times (thanks to Edward Nixon); fix bug in epigraphs (thanks etherean); improve id generation for footnotes - e.g. match behavior of PHP Markdown Extra (thanks to Özgür Gökmen); fix bug in id generation for ToC for XHTML documents; fix problem with \ldots command (thanks to etherean and James Howison); fix issue with   and tilde character; fix bug where footnote special characters were not unescaped (thanks to Chad Schmidt); clean up documentation a bit;

• 2.0.b2 - fix processing of footnotes so that ending in a blockquote doesn't break validity; fix bug in letter.xslt; overhaul XSLT system to allow for different XSLT files for different output formats (e.g. HTML, RTF, LaTeX);

• 2.0.b1 - fix bug in _StripLinkDefinitions that prevented detection of single character labels; change \textwidth to \linewidth in LaTeX export XSLT files (let me know if this causes problems); add Windoze compatibility to the perl scripts (thanks to Jason Bandlow for pointing out this problem, as well as for suggesting a fix);fix issues with glossary support and document the process; complete overhaul of the way namespaces are handled (stripnamespace.pl is no longer needed, XSLT files are rewritten, -nonet and -novalid should be optional for xsltproc); update the Drag and Drop applications to use the "Common" MMD Installation; update to Markdown 1.0.2b8 codebase; add support for natbib and \citep and \citet;

• 2.0.a2 - fix some minor problems with XSLTMathML; allow math to be enclosed in parentheses; change matching for bottomrule in tables; improve handling of tables with no header row (only a header column);

• 2.0.a1 - strip spaces from metadata keys for XHTML validity; make XHTML footnote output more compatible with Gruber's website and PHP Markdown Extra; update XSLT to address these changes (Note: this breaks compatibility with prior versions); add support for definition lists; fix bug when escaping WikiWords in code; add XHTML Header metadata, and update XSLT to ignore <style> tags; add support for the XSLT File metadata tag, which allows a single command to parse any MultiMarkdown file; add additional XSLT files; add the multimarkdown2XHTML.pl and related commands; article XSLT now uses the article option in memoir, rather than the article class; delete the report class (use memoir instead); fix a lot of "minor" bugs; add the "6x9book.xslt" option; allow custom cross-reference labels to headers; give preference to defined links over automatic cross-references; add "poetry" versions of several XSLT files (treat code blocks as formatted text, rather than code --- useful for formatting poetry)

• 2.0.a - New version numbering scheme; update to Markdown.pl 1.0.2b7 code; add support for [link reference] shortcut syntax (i.e. no trailing []) for MultiMarkdown crossrefs; add an extra newline in verbatims to add space before the next paragraph; synchronize numbering schemes of all related MultiMarkdown tools to make it easier to ensure compatibility; add revision numbers to source documents to help track incompatibilities; add LaTeX support for i.e. and e.g.; TextMate MultiMarkdown bundle available; update MultiMarkdownDragAndDrop tools to new codebase; now distributed as a zipfile.

• 1.0.1Multi19.4 - major update; fix issue where cross-references to images defined by alt text had to follow the image in the document; add support for MathML via ASCIIMathPHP; change name to id for footnotes; move _DoHeaders in front of _DoTables to allow cross-references inside tables; fix handling of citations without locator; a table with no header titles and no column alignment row is interpreted as a pull-quote - this is experimental and may be changed; the Bibliography Title metadata field is available for LaTeX to rename the bibliography section; multiple changes to XSLT files to improve compatibility; support for << math >> syntax using ASCIIMathPHP; change HeaderLevel to Base Header Level and process it in XSLT rather than in the OmniOutliner tool; support for Affiliation metadata element; add equation label to possible cross-reference list; compatible with epigraph feature for XSLT conversion to LaTeX; document table labeling feature and default to caption if no label present;

• 1.0.1Multi19.2 - require leading space before unescaping \WikiWord; fixed bug where attributes not included with images; add Bibliography Title metadata key; fix bug with invalid leading characters in header id attributes; allow '-' and '_' in metadata; fix handling of citations in footnotes; fix issue with quotes in link attributes.

• 1.0.1Multi19.1 - minor change to bibliography formatting to allow translation into a \BibTeX compatible format without the use of a .bib file;

• 1.0.1Multi19 - Major update; fix bugs discovered by testing with MarkdownTest 1.0; don't add leading blank line if no metadata exists; fix parsing of link definitions, including attribute parsing; various clean- ups to code and documentation; improve cross-reference handling of special characters; fix bug in handling of wiki links (/ is not automatically added any more); fix bug in title attributes of images; re-enable the inclusion of DOCTYPE in complete documents (this requires the use of the -nonet and -novalid options in xsltproc; fix bug in handling of **; fix bug where WikiWords in code blocks and spans were not unescaped; fix bug where digits were not allowed in metadata keys; fix numbering of footnotes so that they remain in proper order; add basic citation and bibliography features; major bug fixes and testing to precede the release of version 20 (2.0)

• 1.0.1Multi18 - further work to make WikiWord escaping work properly...

• 1.0.1Multi17 - add support for "char" alignment in table columns (NOTE: browsers do not currently support this); fix bug with \ in code spans when WikiWords are disabled; fix bug in bold/italic detection

• 1.0.1Multi16 - can now optionally have header in first cell of each row; fix bug in footnote counting (thanks to Mark Eli Kalderon for pointing this out);

• 1.0.1Multi15 - allow for multiple <tbody> span's within a table; ensure that the variable$g_empty_element_suffix is used everywhere; protect code spans from table parsing

• 1.0.1Multi14 - captions can now be before or after table; add syntax for column spanning within tables (body and header)

• 1.0.1Multi13 - added support for CSS metadata key; allow no alignment option on table cells; support for captions for tables

• 1.0.1Multi12 - added support for image/link attributes; fixed bug in table handling

• 1.0.1Multi11 - added support for table syntax

• 1.0.1Multi10 - allow emphasis at beginning of line

• 1.0.1Multi9 - fix bug in metadata parsing

• 1.0.1Multi8 - first draft of fix for "underscore within a word" problem that causes so many errors with URL's. Now a leading whitespace is required in front of the "opening" _ or * for it to be interpreted as emphasis or strong.

• 1.0.1Multi7 - add Wiki Links support

• 1.0.1Multi6 - correct bug in footnote id handling (Thanks to Jonathan Weber for pointing this out)

• 1.0.1Multi5 - allow disabling of metadata feature

• 1.0.1Multi4 - convert © entities to © (compatible with XSLT); generate cross-refs for images

• 1.0.1Multi3 - fix metadata parsing in the event a key was empty

• 1.0.1Multi2 - add support for footnotes. Major change - no longer use templates, but rather will focus on using XSLT to convert from XHTML output to other formats. I think this will be more flexible and less error prone.

• 1.0.1M - initial release