Could you elaborate a bit on which markdown features you support (or even better: write it into your module haddocks)? Thinks like - autolink detection - ```language blocks? Also, you build on performance-oriented libraries - it would be cool if you could make a small benchmark comparing with the standard C/Python/Ruby parser implementations; AFAIK there is a standard Markdown test suite that this could run against. Concerning the project proposal: I especially find the last feature useful for programming documentation, and would love to have them in a potential haddock succesor. I was also pleasantly surprised that pandoc seems to handle all of this (even with code syntax highlighting). On 05/04/13 02:10, Michael Snoyman wrote:

In case it can be useful in any way for this project, my markdown package[1] is certainly available for scavenging, though we'd likely want to refactor it to not use conduit (I can't imagine conduit being a good dependency for Haddock).

[1] http://hackage.haskell.org/package/markdown

On Thu, Apr 4, 2013 at 11:22 PM, Kim-Ee Yeoh wrote:

On Fri, Apr 5, 2013 at 3:04 AM, Simon Heath wrote:

...

I humbly suggest reStructuredText rather than Markdown, which is what is used by the Python community for documentation. Since it's specifically made for documentation it may be nicer. But, I don't want to spark a format argument.

Could you say something about /why/ you make the suggestion? I, for one, would be happy to google and read links, but what's missing from that experience would be input from a fellow haskeller. In context. In real-time. On topic.

-- Kim-Ee

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

I'd actually like to say that Python has great "documentation traditions" not just thanks to adoption of reStructuredText format, but a bit more. The greatest thing is an adoption of Sphinx [1] documentation engine. In Sphinx you write in extended (by special extensions) reStructuredText format, which is able to easily link to function definitions (with :func:`foo.bar.baz` or :class:`foo.bar.Baz`), other documents (with :doc:`other/doc`) and (most importantly) has extension called "autodoc" which goes and generates documentation for classes and functions automatically (by gathering it from docstrings and other things, similar to haddoc, if I'm not mistaken). My main point is that thanks to adoption of "mixed documentation" techinique people usually tend to keep documentation with has these properties: - it's up to date - there's (most of the time) no need to keep separate API and non-API documentation - API documentation can be easily be extended with long introduction without having to write it in source code (you can write it in document and then include autodoc) I think Haskell's documentation would greatly benefit from (possible?) adoption of something like Sphinx, but I don't think it supports Haskell currently. [1]: http://sphinx.pocoo.org

+++ Simon Heath [Apr 04 13 13:04 ]:

I humbly suggest reStructuredText rather than Markdown, which is what is used by the Python community for documentation. Since it's specifically made for documentation it may be nicer. But, I don't want to spark a format argument.

There is also the Pandoc program, which is a universal-ish markup- language-converter, conveniently written in Haskell. Might be a place to start for this, regardless of the language chosen: http://www.johnmacfarlane.net/pandoc/

Simon

(Pandoc author here.) It probably wouldn't make sense for a key infrastructure component like Haddock to depend on a behemoth like pandoc. But I could help out with a markdown-superset parser if needed. I have an experimental thing here that could be used as a basis (it's 7x faster than pandoc and uses 1/5 the memory, BSD licensed): https://github.com/jgm/Markdown Another idea: If someone contributed a Haddock markup writer to pandoc, then documentation could be written in markdown (or RST or whatever) and converted automatically to standard Haddock markup. David Lazar has recently contributed a Haddock markup reader, but there is no writer. Note: Creating a writer would be a bit tricky, because Haddock markup isn't expressive enough for many of the constructions pandoc allows -- for example, if I'm not mistaken, you can't have multiple paragraphs inside list items. Decisions would have to be made about how to deal with such cases. There are also a few Haddock constructions that don't correspond to anything in pandoc. John

On 4 Apr 2013, at 22:53, Aleksey Khudyakov wrote:

If we are going to change haddock syntax we should add ability to add math formulae to documentation. It's not currently possible and it makes documenting numeric code properly difficult.

How about support for .lhs files? - both those with bird-tracks (which I don't use anymore) and \begin{code}...\end{code} (which I do use). My .lhs files are also LaTeX sources - I guess some way to nicely integrate haddock markup/down/whatever with LaTeX stuff would be needed +1 in any case

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

-------------------------------------------------------------------- Andrew Butterfield Tel: +353-1-896-2517 Fax: +353-1-677-2204 Lero@TCD, Head of Foundations & Methods Research Group Director of Teaching and Learning - Undergraduate, School of Computer Science and Statistics, Room G.39, O'Reilly Institute, Trinity College, University of Dublin http://www.scss.tcd.ie/Andrew.Butterfield/ --------------------------------------------------------------------

I'm not proposing the LaTeX is used for hyperlinking the reference - hence my comment about nicely integrating Perhaps a \begin{haddock} ... \end{haddock} environment* ? * This would only affect those using LaTeX/lhs - everyone else could haddock** as usual ** haddock = whatever markdow/up/sideways scheme you guys come up with... On 5 Apr 2013, at 16:22, Aleksey Khudyakov wrote:

On 5 April 2013 12:20, Andrew Butterfield wrote:

...

On 4 Apr 2013, at 22:53, Aleksey Khudyakov wrote:

...

If we are going to change haddock syntax we should add ability to add math formulae to documentation. It's not currently possible and it makes documenting numeric code properly difficult.

How about support for .lhs files? - both those with bird-tracks (which I don't use anymore) and \begin{code}...\end{code} (which I do use).

My .lhs files are also LaTeX sources - I guess some way to nicely integrate haddock markup/down/whatever with LaTeX stuff would be needed

I'm not sure that it would help. If we to use haddock markup it need to support math typesetting. And LaTeX IMHO isn't right tool for creating hyperlinked API reference

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

-------------------------------------------------------------------- Andrew Butterfield Tel: +353-1-896-2517 Fax: +353-1-677-2204 Lero@TCD, Head of Foundations & Methods Research Group Director of Teaching and Learning - Undergraduate, School of Computer Science and Statistics, Room G.39, O'Reilly Institute, Trinity College, University of Dublin http://www.scss.tcd.ie/Andrew.Butterfield/ --------------------------------------------------------------------

On Fri, Apr 5, 2013 at 10:12 AM, Petr Pudlák wrote:

While I like the idea of allowing any markup language (let's say supported by Pandoc) and freedom it gives to developers, it also has also drawbacks: It makes contributing more difficult, if a project uses some wierd, non-standard markup language.

I feel pretty strongly about this; I don't think we gain much by allowing arbitrary formats. I think the same argument for coding conventions applies here -- consistency is much more important for readability than flexibility. --Rogan

Concerning math expressions, what about using Markdown with MathJAX, like math.stackexchange.com does?

Best regards, Petr Pudlak

2013/4/5 Andrew Butterfield

...

I'm not proposing the LaTeX is used for hyperlinking the reference - hence my comment about nicely integrating

Perhaps a \begin{haddock} ... \end{haddock} environment* ?

* This would only affect those using LaTeX/lhs - everyone else could haddock** as usual

** haddock = whatever markdow/up/sideways scheme you guys come up with...

On 5 Apr 2013, at 16:22, Aleksey Khudyakov wrote:

...

On 5 April 2013 12:20, Andrew Butterfield wrote:

...

On 4 Apr 2013, at 22:53, Aleksey Khudyakov wrote:

...

If we are going to change haddock syntax we should add ability to add math formulae to documentation. It's not currently possible and it

makes

...

...

documenting numeric code properly difficult.

How about support for .lhs files? - both those with bird-tracks (which I don't use anymore) and \begin{code}...\end{code} (which I do use).

My .lhs files are also LaTeX sources - I guess some way to nicely integrate haddock markup/down/whatever with LaTeX stuff would be needed

I'm not sure that it would help. If we to use haddock markup it need to support math typesetting. And LaTeX IMHO isn't right tool for creating hyperlinked API reference

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

-------------------------------------------------------------------- Andrew Butterfield Tel: +353-1-896-2517 Fax: +353-1-677-2204 Lero@TCD, Head of Foundations & Methods Research Group Director of Teaching and Learning - Undergraduate, School of Computer Science and Statistics, Room G.39, O'Reilly Institute, Trinity College, University of Dublin http://www.scss.tcd.ie/Andrew.Butterfield/ --------------------------------------------------------------------

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

Johan Tibell writes:

On Thu, Apr 4, 2013 at 9:49 AM, Johan Tibell wrote:

...

I suggest that we implement an alternative haddock syntax that's a superset of Markdown. It's a superset in the sense that we still want to support linkifying Haskell identifiers, etc. Modules that want to use the new syntax (which will probably be incompatible with the current syntax) can set:

{-# HADDOCK Markdown #-}

Let me briefly argue for why I suggested Markdown instead of the many other markup languages out there.

Markdown has won. Look at all the big programming sites out there, from GitHub to StackOverflow, they all use a superset of Markdown. It did so mostly (in my opinion) because it codified the formatting style people were already using in emails and because it was pragmatic enough to include HTML as an escape hatch.

For what it's worth, I think Markdown is a fine choice for very much the same reason. RST has some nice properties (especially for documenting Python), but Markdown is much more common. Moreover, I've always found RST's linkification syntax a bit awkward. Cheers, - Ben

Would it be too much to ask that a notation be used which has a formal syntax and a formal semantics?

We will document our superset, sure. That's what others did as well. The point is using Markdown as the shared base.

On 5 April 2013 13:24, Richard A. O'Keefe wrote:

On 5/04/2013, at 2:00 PM, Johan Tibell wrote:

...

...

Would it be too much to ask that a notation be used which has a formal syntax and a formal semantics?

We will document our superset, sure. That's what others did as well. The point is using Markdown as the shared base.

Nononono. Sure, the others "documented" their supersets. But they did *NOT* provide what I am asking for: - a FORMAL SYNTAX and - a FORMAL SEMANTICS. I tried to use one of these systems, and found myself unable to determine which combinations of features were legal and what legal combinations of features *meant*. I corresponded with some people who had built markdown parsers, and the answer was the same each time: they had reversed engineered some other parser (typically a Perl one) and bashed on it until they were bug-compatible.

If I want to get a particular effect in LaTeX or even in HTML+CSS, I can usually figure it out *without* having to run any program. If I want to get a particular effect in Markdown, I flounder around and end up doing without.

I am sick of "documentation" that vaguely hints at things, and I am especially sick of Markdown so-called documentation.

To say it one more time: I was unable to use the official Markdown documentation, http://daringfireball.net/projects/markdown/syntax, to guide the construction of a parser.

For example, <br> is a valid URL enclosed in <. . .>, so is it a link, as the "Automatic Links" section would suggest, or is it embedded HTML, as the "Inline HTML" section would suggest? Can you tell *from the documentation*?

For another example, is *foo**bar**ugh* supposed to map to <em>foo<strong>bar</strong>ugh</em> or to <em>foo</em><em>bar</em><em>ugh</em>? Again, I'm not asking "what does this or that *program* do", I'm asking "can you tell from the documentation what they *ought* to do?"

If there is an unambiguous specification of Markdown somewhere (specification; not program), I would be glad to see it.

I don't think so; this was one of the big issues recently when people were trying to get Gruber to actually _do_ something with Markdown as there were all these corner cases.

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

-- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com http://IvanMiljenovic.wordpress.com

On 5 April 2013 15:49, Kim-Ee Yeoh wrote:

On Fri, Apr 5, 2013 at 10:44 AM, Ivan Lazar Miljenovic wrote:

...

I don't think so; this was one of the big issues recently when people were trying to get Gruber to actually _do_ something with Markdown as there were all these corner cases.

In that case, surely this is an opportunity to convene a committee (a la H98) to craft a formal spec?

There have been calls for such a committee: http://www.codinghorror.com/blog/2012/10/the-future-of-markdown.html

-- Kim-Ee

-- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com http://IvanMiljenovic.wordpress.com

Can't we just add some features to haddock? There are a lot of ways to improve haddock a lot, and no one is doing them, so my impression is that haddock doesn't really have active maintainers. Adding a whole new backend seems risky, unless it results in new maintainers joining. For my personal bikeshed contribution, I would like to see haddock move in the way of fewer markup characters and rules, not more. Since haddock is not "statically checked", the only way to find out if I put in an error is to run haddock and then visually inspect the output, unless of course it was a syntax error, in which case the error message is often not very good. I can easily haddock individual files since I have a custom build system, but I imagine cabal users would have to haddock the entire project every time. I regularly see haddock errors in released packages so I'm not the only one. There are lots of ways to improve haddock a lot. For example, better parse error messages. Make ""s smarter so they don't try to link things that are obviously not modules. Or complain if it's not a module. Or better, get rid of them entirely and use single quotes for that. And make single quotes work for non-imported symbols. Incremental support for cabal. Perhaps even deprecate @ and use ' for that too. One thing I think HTML got right is that there are only two characters that need to be quoted. Of course that's at the cost of all the markup being wordy, but the more you move in the markup-style DWIM the more little rules you have to remember. On Thu, Apr 4, 2013 at 6:00 PM, Johan Tibell wrote:

...

Would it be too much to ask that a notation be used which has a formal syntax and a formal semantics?

We will document our superset, sure. That's what others did as well. The point is using Markdown as the shared base.

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

On 9 April 2013 05:08, MigMit wrote:

Отправлено с iPad

08.04.2013, в 21:44, Evan Laforge написал(а):

...

Can't we just add some features to haddock?

No, we can't. At the very least we should FIX haddock before adding features.

How specifically does haddock need to be fixed?

...

There are a lot of ways to improve haddock a lot, and no one is doing them, so my impression is that haddock doesn't really have active maintainers. Adding a whole new backend seems risky, unless it results in new maintainers joining.

For my personal bikeshed contribution, I would like to see haddock move in the way of fewer markup characters and rules, not more. Since haddock is not "statically checked", the only way to find out if I put in an error is to run haddock and then visually inspect the output, unless of course it was a syntax error, in which case the error message is often not very good. I can easily haddock individual files since I have a custom build system, but I imagine cabal users would have to haddock the entire project every time. I regularly see haddock errors in released packages so I'm not the only one.

There are lots of ways to improve haddock a lot. For example, better parse error messages. Make ""s smarter so they don't try to link things that are obviously not modules. Or complain if it's not a module. Or better, get rid of them entirely and use single quotes for that. And make single quotes work for non-imported symbols. Incremental support for cabal. Perhaps even deprecate @ and use ' for that too.

One thing I think HTML got right is that there are only two characters that need to be quoted. Of course that's at the cost of all the markup being wordy, but the more you move in the markup-style DWIM the more little rules you have to remember.

On Thu, Apr 4, 2013 at 6:00 PM, Johan Tibell wrote:

...

...

Would it be too much to ask that a notation be used which has a formal syntax and a formal semantics?

We will document our superset, sure. That's what others did as well. The point is using Markdown as the shared base.

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

-- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com http://IvanMiljenovic.wordpress.com

2013/4/8 Evan Laforge

Can't we just add some features to haddock? There are a lot of ways to improve haddock a lot, and no one is doing them, so my impression is that haddock doesn't really have active maintainers. Adding a whole new backend seems risky, unless it results in new maintainers joining.

As one of the official maintainers (the other one is Simon Hengel), I agree that Haddock can barely be called actively maintained and it has been like that for a long time now. It's sad, but we lack of time, which is the usual excuse of course. A more resourceful maintainer would be great, so if someone reading this is motivated and wants to help out, we'd love to be contacted by you. I don't think a SoC project for Markdown support would be risky at all (with a sufficiently good student), however. And it wouldn't require a new backend, the backend would still be the existing HTML backend in Haddock. David

On 6 April 2013 01:57, John Wiegley wrote:

...

...

...

...

...

Johan Tibell writes:

...

I suggest that we implement an alternative haddock syntax that's a superset of Markdown.

Definite +1 from me too.

+1 Conrad.

I forgot the mention the craziness with the *significant trailing whitespace*. On Fri, Apr 5, 2013 at 9:49 PM, dag.odenhall@gmail.com < dag.odenhall@gmail.com> wrote:

Personally I think Markdown sucks, although perhaps less than Haddock markup.

Still:

* No document meta data * No code block meta data like language for syntax highlighting * No tables * No footnotes * HTML fallback is insecure * Confusing syntax (is it []() or ()[] for links?) * Syntax that gets in the way (maybe I don't want *stars* emphasized) * Above point leads to non-standard dialects like "GitHub Markdown" (no, GitHub doesn't use markdown) * Not extensible, leading to even more non-standard hacks and work-arounds (GitHub Markdown, Pandoc Markdown, other Markdown libraries have their own incompatible extensions) * Not well suited for web input (e.g. four-space indentation for code blocks), although not that important for Haddock

An important thing to note here is that no, Markdown has *not* won because no one is actually using *Markdown*. They're using their own, custom and incompatible dialects.

Only two of the above points apply to reStructuredText (web input and syntax getting in the way), and those particular points don't apply to Creole. Therefore I tend to advocate Creole for web applications and reStructuredText for documents.

On Thu, Apr 4, 2013 at 6:49 PM, Johan Tibell wrote:

...

Hi all,

Haddock's current markup language leaves something to be desired once you want to write more serious documentation (e.g. several paragraphs of introductory text at the top of the module doc). Several features are lacking (bold text, links that render as text instead of URLs, inline HTML).

I suggest that we implement an alternative haddock syntax that's a superset of Markdown. It's a superset in the sense that we still want to support linkifying Haskell identifiers, etc. Modules that want to use the new syntax (which will probably be incompatible with the current syntax) can set:

{-# HADDOCK Markdown #-}

on top of the source file.

Ticket: http://trac.haskell.org/haddock/ticket/244

-- Johan

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

+1 for concistency. Also, consider interop with non-haskell environments. For example showing the documentation of a function in emacs, eclipse, on github, and from a javascript library. All of these can be engineered around, and tooling can be provided. But let me give an example: the other week I was looking for a command-line tool to extract javadoc to display as contextual information in emacs. There is no such tool. Javadoc is "java only". For me, if I could not hack it up in an hour, it was too much work. The solution was rather to craft a specific google search, use "I'm feeling lucky", and extract the subsection containing the documentation for the function. Often the most useful format for documentation is contextual help in an IDE/editor, so don't forget that use-case. Alexander On Sat, Apr 6, 2013 at 1:04 AM, John MacFarlane wrote:

I like markdown and use it all the time. While I acknowledge the problems that have been pointed out, markdown has the advantage of being easily readable "as it is" in the source document, and not looking like markup.

But I do want to point out one problem with markdown as a format for documentation in Haskell files. Consider:

---------------------------------------------------- module MyModule {- # Introduction

This is my module -} where import System.Environment

main = getArgs >>= print ----------------------------------------------------

Now try to compile with -cpp, and you'll get an error because of the '#' in column 1. '#' in column 1 is common in markdown (and even indispensible for level 3+ headers).

One could work around this by disallowing level 3+ headers, by allowing the headers to be indented, or by introducing new setext-like syntax for level 3+ headers, but it is a problem for the idea of using a markdown SUPERset.

John

+++ dag.odenhall@gmail.com [Apr 05 13 21:59 ]:

...

I forgot the mention the craziness with the *significant trailing whitespace*.

On Fri, Apr 5, 2013 at 9:49 PM, [1]dag.odenhall@gmail.com <[2]dag.odenhall@gmail.com> wrote:

Personally I think Markdown sucks, although perhaps less than Haddock markup. Still: * No document meta data * No code block meta data like language for syntax highlighting * No tables * No footnotes * HTML fallback is insecure * Confusing syntax (is it []() or ()[] for links?) * Syntax that gets in the way (maybe I don't want *stars* emphasized) * Above point leads to non-standard dialects like "GitHub Markdown" (no, GitHub doesn't use markdown) * Not extensible, leading to even more non-standard hacks and work-arounds (GitHub Markdown, Pandoc Markdown, other Markdown libraries have their own incompatible extensions) * Not well suited for web input (e.g. four-space indentation for code blocks), although not that important for Haddock An important thing to note here is that no, Markdown has *not* won because no one is actually using *Markdown*. They're using their own, custom and incompatible dialects. Only two of the above points apply to reStructuredText (web input and syntax getting in the way), and those particular points don't apply to Creole. Therefore I tend to advocate Creole for web applications and reStructuredText for documents. On Thu, Apr 4, 2013 at 6:49 PM, Johan Tibell <[3]johan.tibell@gmail.com> wrote:

Hi all, Haddock's current markup language leaves something to be desired once you want to write more serious documentation (e.g. several paragraphs of introductory text at the top of the module doc). Several features are lacking (bold text, links that render as text instead of URLs, inline HTML). I suggest that we implement an alternative haddock syntax that's a superset of Markdown. It's a superset in the sense that we still want to support linkifying Haskell identifiers, etc. Modules that want to use the new syntax (which will probably be incompatible with the current syntax) can set: {-# HADDOCK Markdown #-} on top of the source file. Ticket: [4]http://trac.haskell.org/haddock/ticket/244 -- Johan _______________________________________________ Haskell-Cafe mailing list [5]Haskell-Cafe@haskell.org [6]http://www.haskell.org/mailman/listinfo/haskell-cafe

References

1. mailto:dag.odenhall@gmail.com 2. mailto:dag.odenhall@gmail.com 3. mailto:johan.tibell@gmail.com 4. http://trac.haskell.org/haddock/ticket/244 5. mailto:Haskell-Cafe@haskell.org 6. http://www.haskell.org/mailman/listinfo/haskell-cafe

...

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

On 8 April 2013 07:12, Roman Cheplyaka wrote:

Looks like a bug in cpphs to me (CC'ing Malcolm). It should respect comments. E.g. GNU cpp strips C comments.

Not quite: http://hackage.haskell.org/trac/ghc/ticket/4836

Roman

* John MacFarlane [2013-04-05 16:04:32-0700]

...

I like markdown and use it all the time. While I acknowledge the problems that have been pointed out, markdown has the advantage of being easily readable "as it is" in the source document, and not looking like markup.

But I do want to point out one problem with markdown as a format for documentation in Haskell files. Consider:

---------------------------------------------------- module MyModule {- # Introduction

This is my module -} where import System.Environment

main = getArgs >>= print ----------------------------------------------------

Now try to compile with -cpp, and you'll get an error because of the '#' in column 1. '#' in column 1 is common in markdown (and even indispensible for level 3+ headers).

One could work around this by disallowing level 3+ headers, by allowing the headers to be indented, or by introducing new setext-like syntax for level 3+ headers, but it is a problem for the idea of using a markdown SUPERset.

John

+++ dag.odenhall@gmail.com [Apr 05 13 21:59 ]:

...

I forgot the mention the craziness with the *significant trailing whitespace*.

On Fri, Apr 5, 2013 at 9:49 PM, [1]dag.odenhall@gmail.com <[2]dag.odenhall@gmail.com> wrote:

Personally I think Markdown sucks, although perhaps less than Haddock markup. Still: * No document meta data * No code block meta data like language for syntax highlighting * No tables * No footnotes * HTML fallback is insecure * Confusing syntax (is it []() or ()[] for links?) * Syntax that gets in the way (maybe I don't want *stars* emphasized) * Above point leads to non-standard dialects like "GitHub Markdown" (no, GitHub doesn't use markdown) * Not extensible, leading to even more non-standard hacks and work-arounds (GitHub Markdown, Pandoc Markdown, other Markdown libraries have their own incompatible extensions) * Not well suited for web input (e.g. four-space indentation for code blocks), although not that important for Haddock An important thing to note here is that no, Markdown has *not* won because no one is actually using *Markdown*. They're using their own, custom and incompatible dialects. Only two of the above points apply to reStructuredText (web input and syntax getting in the way), and those particular points don't apply to Creole. Therefore I tend to advocate Creole for web applications and reStructuredText for documents. On Thu, Apr 4, 2013 at 6:49 PM, Johan Tibell <[3]johan.tibell@gmail.com> wrote:

Hi all, Haddock's current markup language leaves something to be desired once you want to write more serious documentation (e.g. several paragraphs of introductory text at the top of the module doc). Several features are lacking (bold text, links that render as text instead of URLs, inline HTML). I suggest that we implement an alternative haddock syntax that's a superset of Markdown. It's a superset in the sense that we still want to support linkifying Haskell identifiers, etc. Modules that want to use the new syntax (which will probably be incompatible with the current syntax) can set: {-# HADDOCK Markdown #-} on top of the source file. Ticket: [4]http://trac.haskell.org/haddock/ticket/244 -- Johan _______________________________________________ Haskell-Cafe mailing list [5]Haskell-Cafe@haskell.org [6]http://www.haskell.org/mailman/listinfo/haskell-cafe

References

1. mailto:dag.odenhall@gmail.com 2. mailto:dag.odenhall@gmail.com 3. mailto:johan.tibell@gmail.com 4. http://trac.haskell.org/haddock/ticket/244 5. mailto:Haskell-Cafe@haskell.org 6. http://www.haskell.org/mailman/listinfo/haskell-cafe

...

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

-- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com http://IvanMiljenovic.wordpress.com

And cpphs strips C comments too. :-) But seriously, John's use-case is the exact opposite of what you suggest. John wants to keep the # inside the comment block. You suggest to remove the comment-block altogether? When I checked the example with cpphs, it turns out that the # line is retained, generating a warning but not an error. I expect Gnu cpp is the software that throws an error. In my opinion, it is perfectly valid to have intentional preprocessor directives inside Haskell comments. Regards, Malcolm On 7 Apr 2013, at 22:12, Roman Cheplyaka wrote:

Looks like a bug in cpphs to me (CC'ing Malcolm). It should respect comments. E.g. GNU cpp strips C comments.

Roman

* John MacFarlane [2013-04-05 16:04:32-0700]

...

I like markdown and use it all the time. While I acknowledge the problems that have been pointed out, markdown has the advantage of being easily readable "as it is" in the source document, and not looking like markup.

But I do want to point out one problem with markdown as a format for documentation in Haskell files. Consider:

---------------------------------------------------- module MyModule {- # Introduction

This is my module -} where import System.Environment

main = getArgs >>= print ----------------------------------------------------

Now try to compile with -cpp, and you'll get an error because of the '#' in column 1. '#' in column 1 is common in markdown (and even indispensible for level 3+ headers).

One could work around this by disallowing level 3+ headers, by allowing the headers to be indented, or by introducing new setext-like syntax for level 3+ headers, but it is a problem for the idea of using a markdown SUPERset.

John

+++ dag.odenhall@gmail.com [Apr 05 13 21:59 ]:

...

I forgot the mention the craziness with the *significant trailing whitespace*.

On Fri, Apr 5, 2013 at 9:49 PM, [1]dag.odenhall@gmail.com <[2]dag.odenhall@gmail.com> wrote:

Personally I think Markdown sucks, although perhaps less than Haddock markup. Still: * No document meta data * No code block meta data like language for syntax highlighting * No tables * No footnotes * HTML fallback is insecure * Confusing syntax (is it []() or ()[] for links?) * Syntax that gets in the way (maybe I don't want *stars* emphasized) * Above point leads to non-standard dialects like "GitHub Markdown" (no, GitHub doesn't use markdown) * Not extensible, leading to even more non-standard hacks and work-arounds (GitHub Markdown, Pandoc Markdown, other Markdown libraries have their own incompatible extensions) * Not well suited for web input (e.g. four-space indentation for code blocks), although not that important for Haddock An important thing to note here is that no, Markdown has *not* won because no one is actually using *Markdown*. They're using their own, custom and incompatible dialects. Only two of the above points apply to reStructuredText (web input and syntax getting in the way), and those particular points don't apply to Creole. Therefore I tend to advocate Creole for web applications and reStructuredText for documents. On Thu, Apr 4, 2013 at 6:49 PM, Johan Tibell <[3]johan.tibell@gmail.com> wrote:

Hi all, Haddock's current markup language leaves something to be desired once you want to write more serious documentation (e.g. several paragraphs of introductory text at the top of the module doc). Several features are lacking (bold text, links that render as text instead of URLs, inline HTML). I suggest that we implement an alternative haddock syntax that's a superset of Markdown. It's a superset in the sense that we still want to support linkifying Haskell identifiers, etc. Modules that want to use the new syntax (which will probably be incompatible with the current syntax) can set: {-# HADDOCK Markdown #-} on top of the source file. Ticket: [4]http://trac.haskell.org/haddock/ticket/244 -- Johan _______________________________________________ Haskell-Cafe mailing list [5]Haskell-Cafe@haskell.org [6]http://www.haskell.org/mailman/listinfo/haskell-cafe

References

1. mailto:dag.odenhall@gmail.com 2. mailto:dag.odenhall@gmail.com 3. mailto:johan.tibell@gmail.com 4. http://trac.haskell.org/haddock/ticket/244 5. mailto:Haskell-Cafe@haskell.org 6. http://www.haskell.org/mailman/listinfo/haskell-cafe

...

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

On 8 Apr 2013, at 14:52, Roman Cheplyaka wrote:

...

In my opinion, it is perfectly valid to have intentional preprocessor directives inside Haskell comments.

Could you give an example where this is useful? ... macro expansions inside the comments are rather exotic.

{- | Some module documentation. #define WEBSITE http://some.really.rather.long/and/tedious/URL/that_I_dont_want_to_type_too_... You can find more information about Foo at WEBSITE/Foo and Bar at WEBSITE/Bar -} As you say, the #define could equally live outside the comment, but I don't see why we should have an arbitrary restriction that it _must_ live outside the comment. As you also say, "the liberty to write whatever one wants inside a comment feels important", and if that includes the intentional use of CPP, why not? Regards, Malcolm