sherwood said:> If agreement is reached, then the group > looks at variants 6,7,8,9 and inquires > if they would like to join in this effort.well, that kind of leveraged consensus would have been the way to go about this process a year ago. but not today.> Has anyone collected a would-be canonical list > of either the ambiguous cases in original MD, > or the variants between the versions of MD?well, i think it has become rather obvious that macfarlane is the most knowledgeable person -- by far -- when it comes to these matters... and, just as above, a year or two back, this would have been the direction for progress to be made... but again, not today. *** allan odgaard said:> I see a bigger chance of filling the niche between > Markdown and reStructuredText that John was alluding to, > getting traction, and then see that as a new format > added to their wiki (it already has several), and > understood when READMEs have the proper extension. > Then we can also finally mandate a file extension: > I suggest calling the new format ?Markright? > with ?.mr? as file extension :)well, i would _love_ to hear more from john about that "sweet spot" that he said he sees... utterly fascinating... but as for the general outline that mr. odgaard plots? well, one more time: "in the past, yes, but not today". here's the deal... in the dayes of olde, all of the markdown variants were roughly equal in power, which is why their standoff was so intractable, and ended up persisting for so very long. then 6 things happened. pandoc assessed 'em all, and decided on multimarkdown. i don't think y'all here appreciated the importance of that. macfarlane is perhaps _the_ smartest guy in light-markup and he researched the question carefully, so his decision is one that we can safely assume to be the wisest one around. even more importantly, john rewrote the parser. as i said, i'm not smart enough to fully understand all of the jargon, but my tenuous grasp on what he did is that he got it right -- a top-down one-pass model; fast, powerful, and flexible. perhaps most important, the parser is specific and correct. it's a simple model that is concrete, easy to grok, and solid. everything is totally disambiguated, so there's no room for any of those troubling inconsistencies to creep in and hide. third, john's model stimulated work done by ali rantakari.> http://hasseg.org/peg-markdown-highlight/this work was for a _syntax_highlighter_, so it wouldn't be too important, in and of itself, except for a couple factors: (1) it used pandoc's parser, and (2) it was _very_ accurate, fixing the problems that plagued _every_ existing_attempt_ at markdown syntax-highlighting, a tribute to the parser. ali wrote this up for his computer science master's studies. ali's work caught the eye of fletcher penney, inspiring him to make a text-editor for multimarkdown, out any day now. i'm not sure how it fits in, because i'm pretty sure that it predated the work by rantakari, but fletcher also changed his parsing model, changing over to the pandoc method. i believe john and fletcher actually worked together on it. so fletcher's parser is fast, powerful, flexible, and accurate. i believe fletcher will confirm the importance of the change. once you start doing it right, you realize the wrong way was well, it was wrong, there's just no other way you can say it. finally, brett terpstra also came out with "marked", which displays on-the-fly markdown output for any text-editor. "marked" can be configured to use any markdown variant, or even another light-markup system (such as textile), but the _default_ is multimarkdown. and we're all familiar with the power of the default, and how very few users change it. thus, for those "marked" users, markdown is multimarkdown. so what does all of this mean? it means that the various markdown variants are no longer balanced in terms of power. multimarkdown has an edge. it first gained that edge on the basis of pandoc's decision. but the edge is going to shoot up immensely _very_soon_, as soon as "multimarkdown composer" hits the app store... none of you has shown that you understand the immensity of having an on-the-fly display while a person is writing... (i knew it'd be important, so i programmed it for myself, but it was only _using_ it that i learned it was _immense_. so my saying that you don't understand it isn't an insult.) and -- even aside from "composer" -- multimarkdown has "marked" doing promotional work for it via any text-editor. what this means is that fletcher has absolutely no incentive to change what he is doing to accommodate a "consensus"... now i'm not saying that he's going to be some _asshole_ who is a power-freak and wants to lord it up over people. it's the opposite, i'd say; he seems like a really decent guy. but there's _no_ reason for him to change multimarkdown, especially since he's recently jacked up his investment in it, and most especially because it's going to pay off big-time... so for him to change stuff _now_ would be extremely risky. when you just put the lighted match to the firecracker fuse, it's not the time to open it up and rearrange the gunpowder. it's time to run away, safely away, and then turn around and _enjoy_the_explosion_. -bowerbird -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20111020/292c35cb/attachment-0001.html>
To clarify: peg-multimarkdown was derived from peg-markdown, not pandoc. Ali Rantakari's work on his highlighter's PEG was also derived from peg-markdown, not pandoc. Composer's highlighter is based on Ali's (which performed highlighting for markdown, not MMD). Composer will highlight MMD *or* MD, depending on preferences. I modified the PEG to support MMD syntax, and made some changes to the highlighter code itself to make it more flexible and to enable context-sensitve editing, and not just highlighting. I am very thankful for the work by both John and Ali, as I would never have been able to complete my projects without their work. As for whether MMD is emerging as the most popular markdown derivative or not, I have no idea. I have multiple search feeds that keep me informed about what people are saying about MMD, but I could be missing a lot of traffic talking about how good other implementations are. F- On Oct 20, 2011, at 1:37 PM, Bowerbird at aol.com wrote: <snip>> third, john's model stimulated work done by ali rantakari. > > http://hasseg.org/peg-markdown-highlight/ > this work was for a _syntax_highlighter_, so it wouldn't be > too important, in and of itself, except for a couple factors: > (1) it used pandoc's parser, and (2) it was _very_ accurate, > fixing the problems that plagued _every_ existing_attempt_ > at markdown syntax-highlighting, a tribute to the parser. > ali wrote this up for his computer science master's studies. > > ali's work caught the eye of fletcher penney, inspiring him > to make a text-editor for multimarkdown, out any day now. > > i'm not sure how it fits in, because i'm pretty sure that it > predated the work by rantakari, but fletcher also changed > his parsing model, changing over to the pandoc method. > i believe john and fletcher actually worked together on it. > so fletcher's parser is fast, powerful, flexible, and accurate. > i believe fletcher will confirm the importance of the change. > once you start doing it right, you realize the wrong way was > well, it was wrong, there's just no other way you can say it.<snip>> -bowerbird > _______________________________________________ > Markdown-Discuss mailing list > Markdown-Discuss at six.pairlist.net > http://six.pairlist.net/mailman/listinfo/markdown-discuss-- Fletcher T. Penney fletcher at fletcherpenney.net -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20111020/3faf8671/attachment.html> -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 4899 bytes Desc: not available Url : <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20111020/3faf8671/attachment.bin>
On Oct 20, 2011, at 1:37 PM, Bowerbird at aol.com wrote:> none of you has shown that you understand the immensity > of having an on-the-fly display while a person is writing... > > (i knew it'd be important, so i programmed it for myself, > but it was only _using_ it that i learned it was _immense_. > so my saying that you don't understand it isn't an insult.)I hadn't thought of it as "immense", but I often edit Markdown documents with Jonathan Rentzsch's MarkdownLive (actually a slightly tweaked fork thereof). As completely basic as it is, I strongly prefer it over a plain text editor, even one that syntax-highlights. MarsEdit's live preview is pretty huge too. I know a lot more great Markdown options have become available, just haven't gotten around to exploring them. --Andy
> then 6 things happened. > > pandoc assessed 'em all, and decided on multimarkdown.Huh? No. Pandoc's markdown extensions aren't the same as mmd's. I have explained in earlier posts why I don't like mmd's table and metadata syntax. Pandoc goes back to 2005 and was not influenced by mmd. John -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20111020/76ddbf50/attachment.html>
andy lee said:> I hadn't thought of it as "immense", > but I often edit Markdown documents > with Jonathan Rentzsch's MarkdownLive > (actually a slightly tweaked fork thereof).ok, maybe just maybe the markdown "grapevine" isn't quite as utterly pathetic as i thought it was... markdownlive should've been seen as a huge advance. mostly, though, as far as i could see, it was just ignored.> As completely basic as it iscalling it "completely basic" is actually very high praise. crude and primitive is what it is, if we are being honest. but it demonstrated the point. and it is _open-source_... and done in xcode, if i remember correctly, meaning that any of you variant people who are brave enough to try to take on "composer" while you still have a fighting chance would have a good head-start in programming your app... really. i'm serious. think about it. before it's too late.> As completely basic as it is, I strongly prefer it over > a plain text editor, even one that syntax-highlights.in other words, this one innovation makes up for a crude and primitive app, even over other apps that are full-featured, and even highlight syntax. (syntax highlighting is a good thing, to be sure, but pales in comparison to a formatted display.) put that "one innovation" into a decent text-editor, and use it for two years, like i have, and you too will describe the effect as being "immense", just like i do. it will make markdown viable for the "average" user. that will greatly enhance the power of the users who are not currently able to generate .html well, or at all. and it will ease the grunt work of the users who do... and both those developments will end up being huge.> MarsEdit's live preview is pretty huge too.oh crap. i always write these posts off the top of my head, and i completely forgot about marsedit. sorry about that... i already had 6 developments, which was probably enough, quantity-wise, but i _definitely_ shoulda mentioned jalkut, since i believe he did the work of removing a dependency, which in turn gave multimarkdown _much_ greater viability. so that was a significant omission on my part; sorry, daniel. anyway, big ups to andy lee for fleshing out the full history! -bowerbird -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20111020/bafe0c83/attachment-0001.html>
john said:> Pandoc's markdown extensions aren't the same as mmd's.that's not what i meant, john. (apologies for confusion.) i meant that pandoc provides converters for markdown (as-defined-by-john-gruber) and for multimarkdown... and that's it. so _none_ of the other markdown variants. at least that's my understanding. (if i'm wrong, tell me.) and to my mind, that's a very important distinction there. (but if you don't feel it's significant, please do inform us.) -bowerbird -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20111020/f2b62b13/attachment.htm>
john said:> No, it doesn't provide a converter for multimarkdown, > though someone on the pandoc-discuss list > mentioned that he was working on one.well, maybe that was what confused me, but that "someone" could not have been "the smartest person in light-markup", because that's _you_, john, so i was mistaken when i thought you'd given multimarkdown the implicit distinction of being the only markdown variant honored with a pandoc converter. though now that i've thought about it, you told me you would include a converter for z.m.l., providing that i wrote the thing, so i guess there is no implied macfarlane validation anyway... no matter how you swing it, pandoc is a phenomenal machine. *** david said:> That's your mistake. Pandoc provides converters for > (one interpretation of) the "core" markdown syntax, > and for its own extended markdown syntax, which differs > from MultiMarkdown's extended markdown syntax > in significant ways.yes, john informed me of that. :+) but thanks anyway; when i get my facts wrong, i actively welcome when a correction is made, and prefer duplicate corrections to none at all. doesn't make much difference, though, because i'll just substitute the marsedit choice of multimarkdown for the "missing" macfarlane endorsement. the important thing, anyway, is that fletcher changed to a better parsing model, which gave clear sailing over regular-expression systems... *** john said:> Pandoc has binary installers for Windows and OSX (intel):that must be kinda new. and it is a welcome development! (my eyes rolled when it took like 3 hours to install haskell.) -bowerbird -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://six.pairlist.net/pipermail/markdown-discuss/attachments/20111020/8c8c929d/attachment-0001.html>