Link syntax (was: definition lists?)

[Waylan Limberg]

On Fri, Nov 21, 2008 at 10:43 AM, Thomas Nichols
<nichols7 at googlemail.com> wrote:

> Tangentially, I notice that this post by Waylan uses a link syntax which

> doesn't seem to be defined by

> <http://daringfireball.net/projects/markdown/syntax#link> but seems to

> work in some implementations at <http://babelmark.bobtfish.net/> - that is,

>

> from the git repo [1]


Actually, that syntax is a common shortcut used by many people in
various mailing lists I subscribe to. Many of the people on these
lists do not use (or perhaps even know) markdown syntax. I see it kind
of a cross between ``[git repo] [1]`` and a footnote. I never expect
these mailing list discussions to ever be feed through a markdown
parser and don't go all out with the formatting. I use just enough to
make it clear to the human reader in plain text without cluttering
things up. For links, I find ``[1]`` works best in such situations and
is the most commonly accepted approach in the circles I find myself
in.

I have concidered adding the ability to convert that syntax to links
in Python-Markdown, but have not as the label would be wrong IMO. In
the above example the result would be

    ... git repo <a href="http://...">1</a> ...

when it should be:

    ... <a href="http://...">git repo</a> ...

Although, now that I think about it, as the reference is removed by
the parser, if the single bracketed text (``[1]``) is left alone in
the text, it no longer has any meaning. That's probably worse that a
link with a meaningless label. Perhaps Markdown should create the link
when a reference for it exists. Although, I'd be more inclined to
leave the brackets around the text to give it more meaning in the
context:

    <a href="http:...">[1]</a>

or maybe

    [<a href="http://...">1</a>]

That said, its not really a big deal to me, as I only ever use that
syntax when I do *not* expect the text I'm typing (i.e.: mailing list
discussions) to be feed through a markdown parser. As long as that's
the case, none of the above arguments (for or against) are relevant as
I see it.

-- 
----
Waylan Limberg
waylan at gmail.com