Markdown development

[Waylan Limberg]

On Wed, Mar 17, 2010 at 8:37 PM, Arno Hautala <arno at alum.wpi.edu> wrote:

>

>

> Me?  I'm happy with PHPMDExtra.  And really that's just for the

> footnotes and definition lists.  Plenty of other forks support both.

> The headache is that so many support one convention but not another.

> It also emphasizes the point that this is Markdown, not HTML.  It

> can't and shouldn't support everything.  Take "Markdown in Python".

> Does a Markdown implementation really need to be able to output an

> image gallery [2], or RSS [3]?  Even the table syntax used by MDE

> gives me bad vibes.  If your table is simple enough to be defined in

> MDE, there's probably a better way to convey that data.  If it's

> complicated enough to stretch what MDE can do, you should probably

> just code the table.  Markdown is not always the right tool for the

> job.

>


For the record, we are dropping the Imagelinks extension completely from
Pyrton-Markdown. It never actually worked without some hard-to-find
no-longer-maintained third party code. It will be gone with the next
release. Both it and the RSS extension were proof of concepts to show what
our extension API "could" do - not that they should necessarily be something
we should do. They ended up being distributed with the core project ever
since and forgotten about. We recently deleted the Imagelinks extension when
someone complained it didn't work -- which reminded us it existed.  We
should probably do the same with the RSS extension. Especially now that we
have other, useful extensions (i.e.: extra [1]).

In fact, I've recently expounded [2] on why I think my code highlighting
extension [3] is no longer necessary, but that is probably one of our most
used extensions, so we'll continue to maintain it for the time being.

[1]: http://www.freewisdom.org/projects/python-markdown/Extra
[2]: http://achinghead.com/archive/88/syntax-highlighting-web/
[3]: http://www.freewisdom.org/projects/python-markdown/CodeHilite

So, I have to agree with Arno here. Adding more doesn't necessarily make
Markdown better. We've already tried it (as have many implementations) and
it doesn't always work. For example I wish I never would have released my
wikilinks extension. It has resulted in more feature creep type of requests
that anything else in the project. Personally, I don't even like wikilinks
for any purpose. But I digress.

The one thing we are proud of about Python-Markdown is the ease with which
anyone can write extensions of their own [4]. Although not quite
as versatile, PHP Markdown has started to follow suit (although mostly
undocumented - search the list archives for clues). Personally, I think this
is where implementations should look to move next - not expanding the
syntax, but providing an API so that it's users can in their own code to
meet the specific needs of their specific projects.

[4]: http://www.freewisdom.org/projects/python-markdown/Writing_Extensions

-- 
----
\X/ /-\ `/ |_ /-\ |\|
Waylan Limberg
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20100317/e91fb6f2/attachment.htm>