Standard (flavored) Markdown

September 4, 2014 7:32 AM

In 2009, StackOverflow and Discourse developer Jeff Atwood called out prominent Apple commentator and Markdown developer John Gruber for failing to produce an unambiguous Markdown specification or maintain the Markdown reference implementation.

Three years later, Etherpad and Meteor developer David Greenspan approached Atwood about standardizing Markdown. Pandoc creator John McFarlane was brought in to help define the specification and test suite. Yesterday, they released Standard Markdown.

Gruber is not pleased.

posted by a snickering nuthatch (253 comments total)

* grabs popcorn *
posted by blue_beetle at 7:36 AM on September 4, 2014 [20 favorites]

Haven't looked at the standard but I have wanted something like this.
posted by grobstein at 7:38 AM on September 4, 2014

Sounds like Gruber's just irritated at what they called it, not that they forked it.
posted by cjorgensen at 7:43 AM on September 4, 2014 [7 favorites]

_grabs_popcorn_
posted by Iridic at 7:43 AM on September 4, 2014 [18 favorites]

Are the Yankees the best team in baseball?
Yes.

This standard is unacceptable.
posted by overeducated_alligator at 7:44 AM on September 4, 2014 [17 favorites]

I think the Gruber has reason to be upset. Name it something other then 'Standard Markdown', make it clear that it is not affiliated with, or endorsed by Gruber.

Also, the Markdown license says:

Neither the name “Markdown” nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

I realize he probably doesn't have a registered trademark on Markdown, and that this new project probably shares no code with, but at least respect the spirit of the license.
posted by papercrane at 7:44 AM on September 4, 2014 [12 favorites]

This reminds me how much I miss the good ol' nerd slapfights from back when Slashdot was a thing. BRB, fixing a bowl of grits.
posted by entropicamericana at 7:45 AM on September 4, 2014 [12 favorites]

C among not a few other software tools went a long way towards changing the worlds infrastructure without an exact specification.

Also the last question of the FAQ is specified incorrectly, should be:

Are the Yankees the best team money can buy?
posted by sammyo at 7:46 AM on September 4, 2014 [1 favorite]

It's totally fair to point out that Markdown has gotten a lot of incompatible flavors, and it's a great response to gather some of the bigger sites that use Markdown to agree on some common ground.

But there's a huge difference between "Let's fork this and work together on it," and "We will call this the standard." The answer to "Who can call something Standard Markdown" has two valid answers: "Gruber" and "Nobody at all." Markdown is not the child of a standards body, it is not owned by a standards body, and I find it really skeevy that there's no mention at all of Gruber in the narrative about how "Standard" Markdown came to be. Did they ask him? Did he decline because he thought it was unnecessary, or because he disagreed with the specific path they were taking? Did they not ask him at all, and just decide to make something up and call it the standard?

This project needs a new name, stat.
posted by Tomorrowful at 7:48 AM on September 4, 2014 [18 favorites]

Seems like the RSS nonsense all over again.
posted by bonehead at 7:48 AM on September 4, 2014 [10 favorites]

Atwood is incredibly hit-or-miss, but Gruber's just being an ass here. His non-stewardship of any specification and "ambiguity is a feature" (seriously, wat) stance on Markdown puts him in absolutely no position to make these kinds of judgements. You don't get to sit back and watch the community diverge into chaos and then get mad when someone tries to fix the mess you've made.

This is like your parents coming to visit for a weekend and organizing your kitchen with a labelmaker so that the knives, spoons, and forks are all in the same goddamn drawer. You can get pissy at them for having the gall to decide where your silverware goes, but it's clear they're only doing this because a) they like you enough to visit in the first place and b) they want to actually be able to do some cooking, which you apparently don't.
posted by 0xFCAF at 7:51 AM on September 4, 2014 [43 favorites]

This project needs a new name, stat.

Markdown
Farkdown
Fartdown
Fartgown
Fartgoon
Fartgoop
posted by mcstayinskool at 7:53 AM on September 4, 2014 [17 favorites]

Markdown is not the child of a standards body

You say it like it's a good thing. Standards committees are often excuses for slap-fighting, but the opposite of a bad standards committee that never generates a spec isn't "Do what you want". Markdown is a large enough part of the web ecosystem that it needs some kind of guiding hand to aid interoperability. I don't even know what the other option is and I have a love/ hate relationship with markdown because of the ambiguity and "Do what you think it best in this complicated corner-case because I didn't want to think about it." I'm with 0xFCAF's take.
posted by yerfatma at 8:00 AM on September 4, 2014 [1 favorite]

Meltdown? Or Nomnom - "Nomnom's Not Markdown"?

This has all the makings of a splendid religious war. It's been too long. And, as there's no right answer to how much specification is the correct amount, it should go on for a gratifyingly long time.
posted by Devonian at 8:01 AM on September 4, 2014 [1 favorite]

The choice is not standards or no standards, its good standards or bad standards
posted by shothotbot at 8:03 AM on September 4, 2014

Gruber went from clever and erudite blogger worth paying to read to anti-everything apple cheerleader in a few years.

It used to he had some notion of nuance. Now he just spends his days constructing arguments about how bad everything is unless it is from Apple. He literally cannot conceive of a world where an iPhone is not immediately perfect device in all ways.

So, you know, fuck him. If you want to be the one in charge markdown, maintain it. Otherwise, get out of the way. I'm sure there is a tiny news release from BlackBerry you have to shit all over.
posted by clvrmnky at 8:03 AM on September 4, 2014 [27 favorites]

0xFCAF Atwood is incredibly hit-or-miss, but Gruber's just being an ass here. His non-stewardship of any specification and "ambiguity is a feature" (seriously, wat) stance on Markdown puts him in absolutely no position to make these kinds of judgements. You don't get to sit back and watch the community diverge into chaos and then get mad when someone tries to fix the mess you've made.

What mess? Okay, there's a bunch of various Markdown-based "standards" used around the web, but is that really such a pain? Are they that different as to cause real problems? I doubt it. This isn't solving a problem any heavy Markdown user has to deal with all that much.

See also the XKCD standards comic I'm too lazy to link to on my phone.
posted by SansPoint at 8:05 AM on September 4, 2014

"ambiguity is a feature" (seriously, wat)

Someone on HN (or maybe Reddit?) gave the charitable* interpretation that Gruber wanted to create "markdown" the syntax, and now we have "markdown" the language. The syntax doesn't say how it should be used and gives guidelines for creating your own "markdown" language, but this gets confusing when you have dozens of languages.

Now we have people saying to forget the generic syntax, let's just have a language.


*perhaps very charitable
posted by Nonsteroidal Anti-Inflammatory Drug at 8:06 AM on September 4, 2014 [4 favorites]

Are they that different as to cause real problems?

Yes, absolutely. I have personally run into problems here in the course of my work. The fact that anyone even noticed and bothered to do this is also evidence that significant divergence has occurred. I recommend looking at the link in Atwood's post that shows fifteen (!) different outputs from 22 implementations. It really is a mess.
posted by 0xFCAF at 8:09 AM on September 4, 2014 [8 favorites]

What mess?

Consider this simple Markdown example:
...
Just for that, I count fifteen different rendered outputs from 22 different Markdown parsers.

Are they that different as to cause real problems?

Not as long as you never try to combine them. Ambiguity in a markup language is a hard problem. If we assume a couple of things:

• a lot of content on the net is written in Markdown
• we want to preserve that content

Then the differences present a large problem. It's fine to hand-roll your own solution because "It's my blog and I'll do what I want" but the web is supposed to be about sharing and interconnecting and aggregating so to be a good citizen and enable that kind of thing, it helps to have a standard.
posted by yerfatma at 8:10 AM on September 4, 2014 [2 favorites]

Gruber went from clever and erudite blogger worth paying to read to anti-everything apple cheerleader in a few years.

It used to he had some notion of nuance. Now he just spends his days constructing arguments about how bad everything is unless it is from Apple. He literally cannot conceive of a world where an iPhone is not immediately perfect device in all ways.

People have been saying this about him since the day he started writing.
posted by cjorgensen at 8:16 AM on September 4, 2014 [6 favorites]

Now he just spends his days constructing arguments about how bad everything is unless it is from Apple. He literally cannot conceive of a world where an iPhone is not immediately perfect device in all ways.

I agree that he is anti-everything apple cheerleader, but he does regularly criticize apple products and policies at least a little bit.
posted by empath at 8:18 AM on September 4, 2014 [5 favorites]

Gruber went from clever and erudite blogger worth paying to read to anti-everything apple cheerleader in a few years.

This ... is baffling. I understand he rubs some people the wrong way, but that really hasn't changed "in a few years" that I can tell. Maybe there's a link to something that expands on this point of view?

I don't mind the attempt to standardize Markdown - that's great - but I wish they hadn't named it "Standard Markdown".
posted by RedOrGreen at 8:24 AM on September 4, 2014 [5 favorites]

0xFCAF Yes, absolutely. I have personally run into problems here in the course of my work. The fact that anyone even noticed and bothered to do this is also evidence that significant divergence has occurred. I recommend looking at the link in Atwood's post that shows fifteen (!) different outputs from 22 implementations. It really is a mess.

The output is a mess, but that's on the parser implementers. I think it should be pretty clear that, if you change your list delimiter, your parser should start a new list. So the problem I'm seeing there is lazy/buggy parsers, not language ambiguity.
posted by SansPoint at 8:24 AM on September 4, 2014

Gruber went from clever and erudite blogger worth paying to read to anti-everything apple cheerleader in a few years.

That may be, but his development of Markdown, the spread of Markdown, and the creation of "Standard Markdown" have nothing at all to do with Apple in any respect. With respect, I really don't see what the fuck his pro-Apple stance has to do with this, unless it's just you saying you don't like him and you think he should "lose" as a result.
posted by Tomorrowful at 8:26 AM on September 4, 2014 [32 favorites]

The output is a mess, but that's on the parser implementers. I think it should be pretty clear that, if you change your list delimiter, your parser should start a new list. So the problem I'm seeing there is lazy/buggy parsers, not language ambiguity.

This is exactly what Standard Markdown is! Making a comprehensive list of test cases so that parser writers can do a better job.
posted by grahamparks at 8:28 AM on September 4, 2014 [7 favorites]

The output is a mess, but that's on the parser implementers

How can it be on the implementers if there isn't a standard to work from?
posted by yerfatma at 8:31 AM on September 4, 2014 [14 favorites]

TBH the one thing I want in standard Markdown is to support the TeX comment syntax (text after %), and I don't have any good reason for wanting it except it would simplify my Pandoc workflows.
posted by grobstein at 8:33 AM on September 4, 2014

> I think it should be pretty clear that ...

You just lost this argument.
posted by benito.strauss at 8:34 AM on September 4, 2014 [7 favorites]

If you run across an edge case that your parser puts out something that looks wrong, you go and change it. The syntax and explanations Gruber provided 10 years ago are enough. Just test.
posted by SansPoint at 8:34 AM on September 4, 2014

Back in the day I preferred textile but that's well and truly a dustbin-of-history format.
posted by Zed at 8:34 AM on September 4, 2014 [4 favorites]

This reminds me how much I miss the good ol' nerd slapfights from back when Slashdot was a thing. BRB, fixing a bowl of grits.

This IS a good ol' nerd slapfight from back when /. was a thing!

It was developed in 2004 by John Gruber, who wrote the first markdown-to-html converter in Perl
posted by mikelieman at 8:35 AM on September 4, 2014 [5 favorites]

The syntax and explanations Gruber provided 10 years ago are enough. Just test.

Almost a decade ago I worked on some W3C Standards (CSS3 and CSS print stuff). I thought I was a good spec writer at the time but I never realized what ambiguity was until I had to implement test cases for things that couldn't be rendered yet. If you think Gruber laid out enough information to cover every possible case, you're nuts.
posted by yerfatma at 8:37 AM on September 4, 2014 [12 favorites]

Several dozen implementors having different ideas of what an "edge case" is and what "looks wrong" is exactly how we got into this clusterarrangement. Very few of these problems are the sort where the renderers' authors would agree that they have bugs.
posted by 0xFCAF at 8:38 AM on September 4, 2014

The syntax and explanations Gruber provided 10 years ago are enough. Just test.

Did you read the spec? The first part lays out a lot of the ambiguities from Gruber's syntax.
posted by brentajones at 8:41 AM on September 4, 2014 [4 favorites]

The Markdown incompatibilities have driven me crazy for years; whenever I have to look up a bit of formatting, I have to try three or four different online references before I find an example that results in the correct output with whatever interpreter I'm using at the moment.

I have never even understood what problem Markdown is trying to solve. What's wrong with HTML? What percentage of people who write Markdown don't know HTML (or the minimal subset of HTML you need to know to be able to do anything you can do in Markdown)? It has got to be extremely low. Why do I have to spend my time figuring out how to get Markdown to spit out the HTML I want instead of just writing the HTML directly in the first place? I don't get the appeal.
posted by enn at 8:41 AM on September 4, 2014 [8 favorites]

I'm pretty much a non-coder aside from CSS/HTML/very-VERY-simple-PHP -- but I think even the worst possible interpretation of this Markdown turf war is still going to be a net benefit to the world.

The idea behind Markdown is human (read: non-coder) legibility for content source files. Even pure HTML is not democratic enough, as anyone who ever made a web site for a non-computer-literate person knows. Whatever gives longevity/ease-of-access to plain text files is a Good Thing for the web. There's too much crap locked inside proprietary systems.

Even if we end up with a ton of variations on Markdownish syntax, the human readability of the source files is the common denominator, so nothing is going to be "lost."
posted by overeducated_alligator at 8:41 AM on September 4, 2014 [2 favorites]

Markdown++ 2.0 Pro Enterprise 64-Bit Edition (Remastered) with Productivity Tools^TM for Windows®
posted by blue_beetle at 8:42 AM on September 4, 2014 [5 favorites]

Markdown is a great idea that never got the maintenance it deserved. The variety of small decisions implementers made while dealing with edge cases caused a lot of weird fragmentation issues. Maybe "Standard Markdown" wasn't the best thing to name it, but this is a terrific project and I hope it gets buy-in by the markdown-using community. I think, ideally, there should be some core standards and some specs for popular "extensions", like tables.

HTML is a fine tool for fancy web pages, but for most documents, I find Markdown much easier to read and to write quickly without thinking about the syntax too hard.
posted by demiurge at 8:44 AM on September 4, 2014 [1 favorite]

How about M@rkdown? It looks cool, and doesn't even violate the license.
posted by leotrotsky at 8:46 AM on September 4, 2014

I don't really care much about the fact that there's no spec right now, it just annoys me when a site lets me input Markdown and doesn't even bother to let me know what flavor they support. The in-house issue tracker at my place of work supports markdown as input and I had to figure out on my own what flavor/parser it uses.

So I guess that means a spec would be nice because I would at least know what I should expect when I use markdown input boxes on any old website.
posted by azarbayejani at 8:48 AM on September 4, 2014

I'm pretty much a non-coder aside from CSS/HTML/very-VERY-simple-PHP

Yes, but for an alligator that's actually really good.
posted by a snickering nuthatch at 8:48 AM on September 4, 2014 [18 favorites]

I have never even understood what problem Markdown is trying to solve.

I find light mark-up languages like Markdown more pleasant to write, but the big advantage is they're still pleasant to read as plaintext. You can write-once (and maintain one copy) and have source that's useful in its own right with automatically generatable HTML as a free side effect.

A lot of your comment refers to the problems that came from the lack of specification that this effort has finally tried to remedy... you might have ended up liking the idea better without the (justifiably) negative association with the pain cause by the differing implementations.
posted by Zed at 8:54 AM on September 4, 2014 [3 favorites]

I'm going to go against the grain here:

The original Markdown specification is an open-source project licensed under a BSD License. From a standards perspective, it's a project that is effectively incomplete, lacks a comprehensive specification, has a buggy reference implementation, and has not had any new work done on it for more than 10 years.

Under every interpretation of open-source ethics and etiquette I know of, this is a totally acceptable situation for another individual to effectively claim ownership of an obviously-abandoned project. The developers behind Standard Markdown are also among the small crowd that actually have enough clout to claim that they are better representatives of the "community consensus" than Gruber.

"Standard Markdown" is a fork of the original that very clearly states its goal and intent in the name. "Standard" Markdown does not claim to be the "Official" markdown, but does state the intent to become a fork with better specifications and documentation than the original.

That being said, I have Opinions and Skepticism about some of the goals of markdown's standardization. It's important for incompatibilities to be minimized or eliminated, but it's (IMO) not necessary for every Markdown parser/compiler to produce completely-identical output, as long as structure and layout are consistent.

Like TeX, Markdown compilers use their "best judgement" for how to render a block of content. The language is not intended to produce pixel-perfect reproducible documents, and should not be used as such. I worry that the new specification could hinder the development of renderers that produce non-HTML output from Markdown source.
posted by schmod at 8:55 AM on September 4, 2014 [24 favorites]

overeducated_alligator: I'm pretty much a non-coder aside from CSS/HTML/very-VERY-simple-PHP

Jpfed: Yes, but for an alligator that's actually really good.

Yeah, but he said he was over-educated.
posted by RedOrGreen at 9:04 AM on September 4, 2014 [1 favorite]

I find it really skeevy that there's no mention at all of Gruber in the narrative about how "Standard" Markdown came to be. Did they ask him?

I remember many tweets from Atwood at Gruber trying to get him on board with some standardizing years ago. I don't know if he reached out to him in other ways but I'd be surprised if he didn't. And if he didn't this effort has still be public and discussed.

This is pissy whining over the use of the word "standard" and it's stupid. Would it be magically okay if they'd called it Standardized Markdown?
posted by phearlez at 9:16 AM on September 4, 2014 [1 favorite]

Oh, and if you just follow back the link from the post here to the "kickoff posting" you see a very big "hey Gruber" part in it. Between that and the tweets I remember seeing I don't think anyone can say Gruber was deliberately left out. There's links farther back into the past about efforts to straighten out issues.
posted by phearlez at 9:26 AM on September 4, 2014 [1 favorite]

I don't think the whining is over the use of the word "standard", but the use of the word "Markdown", as it implies ownership over someone else's work.
posted by jenkinsEar at 9:26 AM on September 4, 2014 [2 favorites]

What's wrong with HTML?

It doesn't degrade as gracefully when a parser isn't available, as structural elements are implemented in tags. Markdown is basically just plain text with a few elements marked up, in a way that is designed to be more intuitively legible to humans.
posted by acb at 9:29 AM on September 4, 2014 [2 favorites]

I have never even understood what problem Markdown is trying to solve. What's wrong with HTML?

The core reason it exists is that HTML's syntax is wonky for creating formatted text and you can encode the same semantics in a less wonky way. Bad syntax does tend to stick around for the reason you mentioned, that if everyone already knows and is comfortable with the current syntax they don't have much reason to switch to something better (which is why most programmers still have to use semi-colons a lot). But if you were designing it all from scratch now, there wouldn't be very good reasons to make hand-edited human readable markup syntax look like HTML's format.
posted by burnmp3s at 9:30 AM on September 4, 2014

Jeez, and I thought the baseball tweets were unbearable.
posted by RobotVoodooPower at 9:33 AM on September 4, 2014

Markdown - because intuitive WYSIWG word processing, a solved problem circa 1984, is too hard for modern web standards.

I ain't talking full-fledged DTP suite, we're talking "Mac Write on a 128k device could do it, Metafilter can't, despite running on a firebreathing server and a bleeding edge 8-gigabyte notebook client."

It's not Metafilter's fault, it's that the tools required to do even workmanlike text entry and formatting are a screaming mindfuck to implement and a slog to use. I don't know what the way forward is, but a fork of Markdown probably ain't it.
posted by Slap*Happy at 9:36 AM on September 4, 2014 [1 favorite]

What's wrong with HTML?

Nothing is wrong with HTML, but given the choice between XML and JSON for small data exchange tasks, people will choose JSON more often than not. Stuff that is easier to craft and easier to read generally gets more traction, over time, unless the more complicated approach is more appropriate.

Surprised that reStructuredText hasn't gotten more attention, over the years. It seems like a pretty decent, flexible and improved alternative to Markdown, in a lot of respects.
posted by Mr. Six at 9:41 AM on September 4, 2014 [7 favorites]

I have never even understood what problem Markdown is trying to solve.

Ask the BBCode or reddit folks. For one thing, it's parsed, so it can be sanity checked for html exploits.

I'd appreciate a version of it here. The blockquote element, for example, has always been troublesome on metafilter, sort of supported, but hidden, so it's used much less frequently than it perhaps should be. Markdown offers a simpler way for users to access it.
posted by bonehead at 9:42 AM on September 4, 2014 [8 favorites]

I recently started working with markdown and realized on day one that there's more than one flavour of it. I'm all for a standardized version. And I'm sure that it itself will spawn forks and version incompatibilities.

I have mixed feelings about calling it standard markdown as both sides have good points. At least it's clear and concise. Maybe not 100% accurate and fair but years from now, will anyone still care? (Aside from Gruber ...)
posted by praiseb at 9:44 AM on September 4, 2014

I have never even understood what problem Markdown is trying to solve.

_This_ is easier and more natural to type than <i>this</i>, and is readable even if the parser isn't implemented.
posted by empath at 9:48 AM on September 4, 2014 [12 favorites]

Markdown - because intuitive WYSIWG word processing, a solved problem circa 1984, is too hard for modern web standards.

It's still nice to have a human-readable representation of the actual content. If I'm looking at a diff of my readme file on github, it's nice that I can see the changes in plain text and not have to rely on some special WYSIWG tool.
posted by burnmp3s at 9:48 AM on September 4, 2014 [1 favorite]

So.. Markdown got Kleenexed/Xeroxed
posted by DigDoug at 9:49 AM on September 4, 2014

It's also easier to just type simple markup than take your fingers off the keyboard, move to the mouse, select the text, then either look for a button or type some hotkey.
posted by empath at 9:50 AM on September 4, 2014

I also dislike wysiwyg text editing in general. It's fine for making an invitation to a backyard bbq, but it's really opaque to troubleshoot if you try do anything 'difficult' with it. See all the seething hatred people have for MS Word.
posted by empath at 9:52 AM on September 4, 2014 [8 favorites]

the use of the word "Markdown", as it implies ownership over someone else's work.

I'm not sure how you solve that problem when you're trying to create a fully-fleshed out version of something like this. Markdown the decade-old-unchanged-publication still exists on Gruber's site and people are free to keep writing parsers that follow it and make up shit when it's "pretty clear" what the proper solution to an ambiguity is.

Sure, you can pick a new name. That doesn't really do a lot to help all the places out there with in-flux issues actually using Markdown and dealing with stuff Gruber doesn't want to participate in. He's got time to crap on people doing the work but not to bless anything or participate - which is totally his right. And the "don't call it Markdown then" position isn't exactly new. But as is said here, the genie is out of the bottle when it comes to people using Markdown and making tools to do it.

So you choose - do you fill in the blanks of the original abandoned sketch to help the people using it and give it a name that makes it sufficiently obvious that this is an elaboration to help them? Or do you call it a whole new name and cope with folks not knowing it's there and push folks to change the name when they use your more rigorous standard?

I can see the "Gruber should get to control any/all use of the word Markdown" position but it seems like it's picking one person's ego over what's best in the interest of a very large community. A community using something supposedly released into the world for public use in the first place.
posted by phearlez at 9:52 AM on September 4, 2014 [10 favorites]

It's also easier to just type simple markup

Interestingly, this isn't true for soft keyboards. With half the web using touchscreens now, are these sorts of schemes past their day?
posted by bonehead at 9:54 AM on September 4, 2014 [1 favorite]

_This_ is easier and more natural to type than <u>this</u>, and is readable even if the parser isn't implemented.

Sure, but you're kind of cherry-picking there. I think it's a lot easier for someone unfamiliar with the syntax to figure out what <code>this</code> means than `this`.

Anyway, when I see Markdown it's usually in things like web comment forms, where essentially no one is going to try to read it as plaintext, ever. So why not just allow a (strictly-enforced, parsed) subset of HTML?
posted by enn at 9:54 AM on September 4, 2014

empath: _This_ is easier and more natural to type than <i>this</i>, and is readable even if the parser isn't implemented.

But this is one of the things I've never gotten about Markdown: Why doesn't _this_ become <u>this</u> and /this/ become <i>this</i>?
posted by jferg at 9:55 AM on September 4, 2014 [21 favorites]

Previous we had a problem because markdown wasn't a standard and was very ambiguous resulting in differing implementations and formatting. A group of developers got together and declared a "standard". So now we have two problems.
posted by humanfont at 9:57 AM on September 4, 2014 [2 favorites]

Good question, but that's more of a question about implementation than the need for a simplified parser in general. I also don't think discoverability is one of its reasons to exist. Its for people that post and comment a lot on blogs. Once you know what it does, it's just a hell of a lot faster and more convenient than using either html or wysiwig.
posted by empath at 9:58 AM on September 4, 2014

Ah, and the answer to why it's <i> rather than <u> is that markdown doesn't actually specify. It produces an <em> tag, which is up to the css as to how it gets displayed (italics is standard).
posted by empath at 10:01 AM on September 4, 2014 [4 favorites]

People have been saying this about him since the day he started writing.

Some people see the truth sooner than others.
posted by aspo at 10:03 AM on September 4, 2014

But this is one of the things I've never gotten about Markdown: Why doesn't _this_ become this and /this/ become this?

I suspect 'cause _this_ as used for over a decade when text was king in email and on Usenet was thought of as meaning italics and no one particularly missed not having a way to represent underlining.
posted by Zed at 10:03 AM on September 4, 2014 [2 favorites]

I'm all for this, because I use three or for different systems every day that have slightly different syntax and I'm too old to keep it all straight, so I have to read the docs every damn time. Now, whether all of those will adopt this remains to be seen. I'm sure StackExchange will, and I'm willing to bet GitHub might if the "controversy" doesn't mean the project is DOA. Hard to say with Pivotal.

Anyway, I'm all for it. Gruber is adequately credited—for my taste; he may have a legal argument, I dunno—and it's not like Atwood didn't announce they were doing this a long time ago. Now I have to go read the spec and see if there is anything reasonable to which I should object.
posted by jeffamaphone at 10:03 AM on September 4, 2014

I don't know if he reached out to him in other ways but I'd be surprised if he didn't

There was this Twitter exchange from two years ago that shows the animosity involved here: @codinghorror When you tell me to jump, should I ask “How high?”
posted by smackfu at 10:10 AM on September 4, 2014 [1 favorite]

Underlining is an artifact of typewriters not being able to do emphasis in text properly. I find it harder to parse than either increasing font weight or italicising. I'm not at all sorry to see it die.
posted by bonehead at 10:12 AM on September 4, 2014 [3 favorites]

Some people see the truth sooner than others.

I think a lot of people's apple animosity boils over into gruber animosity (or maybe vice versa). I read him all the time, because I've bought into the apple ecosystem and his blog tends to give me a good idea of the thinking that is going on in the company (since I assume he talks to a lot of people there), which I like to know, whether or not I agree with it or him.
posted by empath at 10:16 AM on September 4, 2014 [3 favorites]

I'm grateful to a group of people wading in and trying to make a standardized Markdown variant with a proper test suite, etc. It seems like a dick move to brand it "Standard Markdown" when the creator of Markdown isn't on board, and I would have advised a different name. But the community can benefit from a standard like this and if they did it right, the value of the technical work will outlast the nerdfight.

OTOH I was involved in the whole RSS wars, and the Atom spec, and it was such a mess. Similar problem; RSS was sort of owned by a completely inappropriate person. (For different reasons). And there were a bunch of variants and resistance to a growing standard etc etc. So a bunch of folks got together and made Atom. And it's a good spec, but in the meantime the whole feed syndication thing failed for other product reasons. And then Facebook and Twitter came and were only grudgingly sort of RSS/Atom compatible and all that standards work was mostly wasted.

Bottom line: what matters is what Reddit and GitHub implement as Markdown.
posted by Nelson at 10:17 AM on September 4, 2014 [7 favorites]

Yeah smackfu, that's the sort of thing I was remembering. And it's consistent with the sort of nonsense from the email list years earlier. Gruber wants to stomp on any changes or efforts that don't involve him and he wants to be pissy when people try to.

I see the moral position about "Markdown is Gruber's name" and respectfully disagree, but he sure as shit doesn't have any claim to being on some high road here.

Personally I like Gruber's writing and think the apple fanboy wahwah against him is tiresome. He's not shy about showing his biases and priorities. If you don't like it don't read it.
posted by phearlez at 10:18 AM on September 4, 2014 [1 favorite]

Markdown - because intuitive WYSIWG word processing, a solved problem circa 1984, is too hard for modern web standards.

There is no widely accepted text markup that is both readable and writable on all systems without special software, and doesn't look like hell in its raw form. (No, HTML doesn't count, it's barely human-readable and ugly as sin.)

I agree that this is stupid and unfortunate, but the situation exists and it's not really the fault of "web standards", it's been a problem since ... well, since forever.

Mostly, I blame the people who first decided that 7-bit ASCII should be enough for anybody and then built a bunch of networked operating systems around that assumption. If they'd bothered to consult, I don't know, anyone who does anything with print rather than just looking at a typewriter keyboard, they would have realized that expressing the English language requires a lot more than just [A-Za-z0-9] and a couple of crappy punctuation symbols.

As a result, everything from diacritics to italics/bold gets bolted-on by either the operating system or the application layer, and typically in naive, incompatible ways.

Because a lot of editors — both in-browser and elsewhere — still fall flat when trying to handle anything that's not in that low-ASCII set, we have kludges like Markdown.

Markdown is a beautiful flower compared to the horrorshow that is Punycode, as well. If we want to get enraged about gross hacks as a result of naive historical decisions, that's a better target.
posted by Kadin2048 at 10:23 AM on September 4, 2014 [3 favorites]

RSS was sort of owned by a completely inappropriate person.

wanna guess who just weighed in with his opinion? ;)
posted by xcasex at 10:25 AM on September 4, 2014 [8 favorites]

Most Microsoft Office programs will automatically italicize text if you do _this_ and bold text on *this*. I looked into it and I found at least one MS article about this autoformatting going back at least to Word 97 so it's been around for awhile before Markdown was created.
posted by Green With You at 10:27 AM on September 4, 2014 [2 favorites]

It seems like a dick move to brand it "Standard Markdown" when the creator of Markdown isn't on board, and I would have advised a different name. But the community can benefit from a standard like this and if they did it right, the value of the technical work will outlast the nerdfight.


I think an important thing to remember is that to launch a whole new project with a new name means one of two things happen with the sizable number of libraries already in place to work with Markdown: either they don't pick up this new more rigorous spec or they do.

If they don't pick it up it's either because they abandon it to now work with MarkedUp (say) and now we have this languishing set of libraries that are out there still being used by old projects or because MarkedUp gains no traction (and solves nobody's commonly-accepted problems).

If they do pick it up now they're offering MarkedUp-spec Markdown and that's what we see written all over the place. Until the world drops one word or the other and just calls it Markdown or MarkedUp. So either Gruber gets his ego-massage because his name is still used while other people do the work of keeping up a level of rigor or Gruber's chosen name ends up associated with places that haven't updated their shit in years and years.

I dunno. Maybe that's the more respectful thing even if, I think, it makes for a really ugly hash in the short term and sub-optimal results eventually. I can certainly understand the desire to hold onto a name; we decided to shut down our six-year-old news site recently and some people asked "why not just hand it off to a new brood?" But we didn't have anyone clamoring for that right and we didn't want to just arbitrarily cede control without some sense that things would go on in a way we'd approve of. But a toolset - hell, what is really just a loose spec since I don't think anyone is still using a Gruber-implemented chunk of code anymore - seems different in important ways.

But that's easy for me, lacking a dog in the fight, to say.
posted by phearlez at 10:28 AM on September 4, 2014

I was going to say, Markdown was whatever Editorially decided it was (best writing experience, ever, how I miss it so), but then I realized that it's essentially no more and I got sad again.
posted by fifteen schnitzengruben is my limit at 10:30 AM on September 4, 2014

It's a pretty common pattern. You release the initial draft of something and are super gung-ho about it. And then the fun stuff is done and actual work is what's left. And you don't ever feel like doing the work, so it just languishes. But the people who use your work get more and more invested in it, and some of them are willing to do the work, and they get very frustrated that you keep saying "I swear I'll do soon" but then years pass and it doesn't happen.

The good solution in this case is to fork it. The bad solution is to try to take over the project. Atwood and friends seem a bit too much in the latter camp.
posted by smackfu at 10:32 AM on September 4, 2014 [1 favorite]

I have never even understood what problem Markdown is trying to solve. What's wrong with HTML?

You’re comparing Markdown to the wrong thing. Markdown is an improvement to plan text files. It’s an improvement to README.TXT.
It reads by humans as well as plain text does, with the extra bonus that a simple parser can add some pizazz (like italics & bullet points), so that it reads even better by humans than plain text. That’s it.
The problems people have with Markdown stem from them wanting it to solve problems it wasn’t designed to solve. I imagine that, from Gruber’s perspective, there’s nothing that needs to be fixed.
posted by davel at 10:38 AM on September 4, 2014 [4 favorites]

Emphasis by underscore and asterisk have been routine, commonly understood usages in email and netnews/usenet for as long as I've been using those two (as well as carets and underscores for superscripts and subscripts, respectively). I wouldn't be terribly surprised to see examples from even earlier than the mid-80s either.
posted by bonehead at 10:42 AM on September 4, 2014 [1 favorite]

Yeah, it's nice for github especially, because people can upload text files with their code and you can cat it on a linux box and it reads fine without any parsing, but it looks better when you're browsing on the site.
posted by empath at 10:43 AM on September 4, 2014 [2 favorites]

People may be too young to have experienced plain text, so they may not get the point of Markdown.
posted by davel at 10:44 AM on September 4, 2014 [6 favorites]

I'm fine with calling this new thing "Markdown"; there's already like 20 competing incompatible things called some variant of "Markdown". But naming it "Standard Markdown" without the blessing of the one guy everyone would expect to be involved in the Standard is a bit cheeky. OTOH if that one guy is AWOL, well, what do you do? The community will speak on this. And given Reddit and GitHub are both involved in the new spec, I'm guessing this work will end up being accepted as a standard.

This list of what changed in Standard Markdown is very helpful. "As little as possible!", which was wise, but then they've flirted with danger by making hygeinic changes like changing the meaning of \ escapes. Probably OK, but risky that it causes a headache for someone somewhere.

I'm a little surprised they didn't take the chance to mandate a Unicode-compatible encoding. "This spec does not specify an encoding; it thinks of lines as composed of characters rather than bytes. A conforming parser may be limited to a certain encoding.". I get why they just punted, but I'd argue that any ASCII-only text processor in 2014 is broken by design and should not be allowed to be called "standards compliant".

The one aspect of Markdown syntax that trips me up every time is how its link syntax is backwards from HTML. Markdown: [text](http://example.com/). HTML: <a href="http://example.com/">text</a>. For some reason I find myself hand-converting Markdown to HTML and back frequently and having to swap the two is such a nuisance. But it'd be a huge mistake to try to change this in Markdown or add a new syntax now, so we live with it.
posted by Nelson at 10:46 AM on September 4, 2014 [3 favorites]

I think it's more likely that a lot of people consume stuff in a way that they don't ever have to cope with plain text they can't see rendered, regardless of their age.
posted by phearlez at 10:46 AM on September 4, 2014 [1 favorite]

What's wrong with HTML?

The biggest problem with HTML is that it collapses line breaks unless you know to use the <p> tag. Even MetaFilter "fixes" that.

The second problem with HTML is that you usually only want to allow people to format text, not write a web page. So you have to whitelist tags, and you end up with something that is a subset of HTML anyways.
posted by smackfu at 10:49 AM on September 4, 2014 [5 favorites]

There is a "standard markdown" that also happens to be better than markdown in every single way and more useful while also being compatible - it is called multi-markdown.
posted by Another Fine Product From The Nonsense Factory at 10:49 AM on September 4, 2014 [2 favorites]

I'm a little surprised they didn't take the chance to mandate a Unicode-compatible encoding.

Gruber probably spent the slightly-better-part of a week creating Markdown. It was a take it or leave it idea, not an RFC or ISO standard. People took it, and ran (too far, IMO) with it.
It didn’t need a Unicode mandate.
posted by davel at 11:02 AM on September 4, 2014

(Oh, The Standard Markup folks. Right, sorry.)
posted by davel at 11:03 AM on September 4, 2014

The project seems worthwhile, but they might want to consider another name. I suggest "StandDown".
posted by Joakim Ziegler at 11:10 AM on September 4, 2014 [3 favorites]

Passing through raw HTML is a recipe for disaster between user errors and security risks. Since you have to parse it to sanity check anyway, why not use something easier to type? Hence, markdown and all the other markup little languages.
posted by bonehead at 11:19 AM on September 4, 2014

For a better response from Gruber, listen to this. (Audio sucks, but Twitter sucks even more)
posted by smidgen at 11:19 AM on September 4, 2014 [1 favorite]

better than markdown in every single way and more useful while also being compatible - it is called multi-markdown.

For those who are unfamiliar: MultiMarkdown is a backwards-compatible extension to Markdown (actually it's a drop-in replacement in some blogging software, specifically Wordpress and Blosxom IIRC) which adds support for lists, tables, footnotes, citations, and exports to well-formed XHTML as well as a bunch of other formats. It's pretty cool, particularly for standalone static documents like READMEs and install docs.

There are two things I dislike about the syntax, though: first, *this* and _this_ are both italics, while **this** and __this__ are both bold. That has never made sense to me. Why is the difference between bold and italics the number of emphasis chars, rather than the type of characters? It would make more sense to have the underscore represent italics and the asterisk represent bold.

This is true in regular Markdown as well, so I blame Gruber.

My other pet peeve is that the slash character as an italics substitute isn't supported, and that's what I've always used and seen used on Usenet and other text-only channels (e.g. IRC) for italics. E.g. /this/ should be italics.

I hate underlining as much as the next person, but it seems to me that the logical syntax for Markdown-esque markup would have *bold*, /italics/, and _underline_. Humf.
posted by Kadin2048 at 11:20 AM on September 4, 2014 [8 favorites]

Gruber's response is more like "Smackdown" not "Markdown"

Sorry. I'll see myself out.
posted by zuhl at 11:42 AM on September 4, 2014

At a guess, I think it's because /this/ is used to mark off a regexp in a lot of programming languages..
posted by empath at 11:50 AM on September 4, 2014 [1 favorite]

For a better response from Gruber, listen to this.

Ugh, if anything that makes it even more maddening. There's no problem, nothing needs fixing, ambiguity is fine. Except you have a lot of people telling you there is one and they're having issues and can point to specifics. And he doesn't even know what version number he's assigned to it, despite talking up how much traffic that page gets. And some vague "maybe I should spend a week..." blah blah. Because suddenly you're going to do that after six years of not doing it?
posted by phearlez at 11:55 AM on September 4, 2014 [4 favorites]

I listened to the Gruber podcast from July 19, 2014. It's really frustrating at the start, disrespectful, particularly his sidekick (Marco Arment) chuckling mockingly at the very idea of a group of people working to improve Markdown.

The strongest arguments Gruber make are that it's a good thing that there's no standard Markdown, that app-specific variants like GitHub-flavored Markdown are a good thing. Also that no spec or standard is really necessary, it's successful as-is. They don't address the argument many Markdown implementors have made that the central syntax is ambiguous and confusing.

Gruber does go on to say he'd like to shut down a competing effort. Some discussion of trademark.

Real world example of Markdown brokenness: Reddit's version doesn't handle parentheses in URLs. Which are common in Wikipedia. So something like [this link](http://en.wikipedia.org/wiki/The_Standard_(Hong_Kong)) doesn't work. Standard Markdown explicitly allows one level of parentheses in the URL, but not arbitrary nesting.
posted by Nelson at 11:57 AM on September 4, 2014

How about:

GReTSSWeDDown (Gruber Refused To Standardize So We Did Down)
posted by symbioid at 12:01 PM on September 4, 2014

Or MarkFwd or Fwdown or MarkAhead, or CrybabyDown...
posted by symbioid at 12:03 PM on September 4, 2014 [1 favorite]

And I think every platform I post to that "supports" markdown on seems to do so in incompatible ways, or has a bad habit of supporting it here but not there, in this window but not that window.

But forget WYSYWIG. While Markdown isn't perfect, it's one of the better WYTIWYM (what you type is what you mean) systems. It's definitely superior to LaTeX, which drags you back into the morass of worrying about quarter inches, centimeters, and picas any time you need to do anything graphical. Editing HTML is fine as long as you're just on the sentence level, but operations like splitting or transforming block units is painful. Lists especially are annoying, and a bad or missing closing tag can break your whole document.

And with a tool like pandoc, you can go to HTML, LaTeX, or ODT (and from there to RTF or MSWord).
posted by CBrachyrhynchos at 12:04 PM on September 4, 2014 [1 favorite]

Should have named it GrowUp.
posted by ijoshua at 12:15 PM on September 4, 2014 [4 favorites]

The strongest arguments Gruber make are that it's a good thing that there's no standard Markdown, that app-specific variants like GitHub-flavored Markdown are a good thing.

Does he actually make an argument to that effect? Or does he just say it's the case? Because prima facie it isn't a good thing and I'm not sure why he might think it is.
posted by kenko at 12:21 PM on September 4, 2014

You should probably listen to the podcast yourself.

Sorry for the confusing language though, to be more explicit I should have written "one argument Gruber makes for why Markdown doesn't need a standard is that app-specific variants are a good thing". I think he's absolutely right about the value of application-specific flavors. For example, on GitHub it's great that "Issue #3" is automatically linked to the bug report page for issue #3. I just think you could have a strong standard for core Markdown, then have extensions. "GitHub flavored Standard Markdown".
posted by Nelson at 12:25 PM on September 4, 2014 [1 favorite]

That reminds me a lot of Dave Winer declaring that if you didn't like RSS as it was, you could just add an extension, disregarding that the biggest problems were with the vaguely-defined core elements.

(I'm another veteran of the RSS/Atom wars - We're at the "RSS 1.0" stage of proceedings, aren't we?)
posted by grahamparks at 12:43 PM on September 4, 2014 [1 favorite]

It's really frustrating at the start, disrespectful, particularly his sidekick (Marco Arment) chuckling mockingly at the very idea of a group of people working to improve Markdown.

Well, there's obviously five years of hurt feelings going around here, so at this point, the last person Gruber would work with is Atwood. People aren't robots.
posted by smackfu at 12:51 PM on September 4, 2014

Sure, it's great that github can link #3 to issue number three, something that wouldn't make sense for other applications of Markdown. But that's a far cry from saying that the (many) ambiguities and underspecifications in Gruber's spec and in Markdown.pl should all be left to different application developers to resolve in their own way, which seems like the only relevant direction of argument.
posted by kenko at 1:04 PM on September 4, 2014

Is there some initial ugly encounter between them that I am not aware of, beyond Atwood's taking the position that Gruber wasn't doing well by Markdown? Nobody likes being criticized but it's not like Atwood was alone in frustration with the state of Markdown - as he links to - and Atwood DID play a part in making something huge and awesome that makes notable use of Markdown.
posted by phearlez at 1:04 PM on September 4, 2014 [2 favorites]

Is Atwood ever not terrible? Between the Shanley crap, "only special unicorns can be programmers" (citing a study the author himself repudiated) and this, he just never seems not-gross.
posted by dame at 1:05 PM on September 4, 2014

Real world example of Markdown brokenness: Reddit's version doesn't handle parentheses in URLs. Which are common in Wikipedia. So something like [this link](http://en.wikipedia.org/wiki/The_Standard_(Hong_Kong)) doesn't work. Standard Markdown explicitly allows one level of parentheses in the URL, but not arbitrary nesting.

You actually can make it work by escaping the parentheses like so:

[this link](http://en.wikipedia.org/wiki/The_Standard_\(Hong_Kong\))

...but yeah, that's kinda kludgey.
posted by jason_steakums at 1:06 PM on September 4, 2014

> If I'm looking at a diff of my readme file on github, it's nice that I can see the changes in plain text

The trouble with line-based diffs of markup is that all contributors have to use the same line length, and have text wrap triggered in exactly the same place, otherwise you'll get a false positive difference on identical text. Old SGML editors used to do diffs based on normalized content, so would only show the elements that had really changed, rather than the ones that were merely reformatted. Heck, those ancient editors even tried to do something WYSIWYG-ish, something that I'm mystified that all of the vogue text markup dialects eschew.

For me, Markdown is a johnny-come-lately version of txt2html, and a feature-starved version of AsciiDoc. Maybe I'll create a text markup format for TEI just for the lulz ...
posted by scruss at 1:13 PM on September 4, 2014 [1 favorite]

Is Atwood ever not terrible?

I think Atwood writes some fine stuff. But then, I thought the Shanley-inspired thing was good on its own and not an unreasonable riff on other work. If your issue is with the initial lack of acknowledgment Atwood explained his reason for not initially linking and capitulated and I can't say that I find a lot of reason to disagree with his observations.
posted by phearlez at 1:15 PM on September 4, 2014 [1 favorite]

For me, Markdown is a johnny-come-lately version of txt2html, and a feature-starved version of AsciiDoc.

Yep, there are plenty of "turn plaintext into HTML" formats. And most of them are pretty much the same for the features that 90% of the users use, and the rest is just a matter of taste. That's what makes it odd that these guys are so invested in taking Markdown as it is but just cleaning it up, when they could pretty easily start over and end up at the same place, without all this drama.
posted by smackfu at 1:21 PM on September 4, 2014 [3 favorites]

His reasons are tone policing and his suggestions suck. Which if you followed the incident, you know.
posted by dame at 1:29 PM on September 4, 2014

For a better response from Gruber, listen to this. (Audio sucks, but Twitter sucks even more)

I took the liberty of transcribing this segment into a Gist.
posted by Nothlit at 1:35 PM on September 4, 2014 [1 favorite]

Nothlit: I forked your Gist to make the type "Markdown". No, really, it helped.
posted by Nelson at 1:39 PM on September 4, 2014 [5 favorites]

So first you ask a question when you really just mean "Atwood sucks" and now you tell me what I "know." I disagree that he sucks. I disagree with your assessment of that incident.
posted by phearlez at 1:39 PM on September 4, 2014 [2 favorites]

Is Atwood ever not terrible?

Yes, the 95% of the time he writes about programming. When he writes about broader social issues, he falls into the typical myopic programmer trap a lot.
posted by yerfatma at 1:45 PM on September 4, 2014 [1 favorite]

Well your answer of "no" is "I interpret that incident far differently than you do," not "oh he did some objectionable things but here are some better ones." So, it was a bad answer to a rhetorical question. Yay?

But honestly, a Markdown thread probably shouldn't be a referendum on Atwood, so I will do my MeFite part and end this derail now.
posted by dame at 1:49 PM on September 4, 2014 [1 favorite]

I took the liberty of transcribing this segment into a Gist.

Now I've catched up with Atwoods writings and some of it comes off as more than a bit glib and some of it is funny in the way he falls into the typical isolated programmer trap, but his writings on technical issues is quite good, as far as technical writing goes.

But that podcast segment has some issues, sure there's the entitlement -- it's his code and his child after all -- but most of it is actually in what i'd perceive as very smug and elitist, both by Arment and Gruber.
posted by xcasex at 1:50 PM on September 4, 2014

Well, that's kind of their schtick.
posted by entropicamericana at 2:01 PM on September 4, 2014

For a better response from Gruber, listen to this. (Audio sucks, but Twitter sucks even more)

It's worth noting that the context of this in the podcast was immediately following a discussion on RSS vs Atom, and how RSS was a practical standard, while Atom was a standard-by-committee that supporting everything but did nothing well.
posted by smackfu at 2:07 PM on September 4, 2014

Worth noting in these discussions: Marco Arment has been trying to tone down the rhetoric that he uses online, and acknowledged that it's been a problem in the (very recent) past.

Gruber could take a few notes from this... His contributions to this debacle have been combative and non-constructive. I agree that we probably shouldn't be taking etiquette lessons from Jeff Atwood, but I genuinely don't see a problem with anything that he's done so far.
posted by schmod at 2:09 PM on September 4, 2014 [1 favorite]

The cheeky attitude evinced in the naming process that will now power an infinite popcorn machine could be cast as an homage to Gruber's rhetorical style.
posted by 99_ at 2:10 PM on September 4, 2014

Well your answer of "no" is "I interpret that incident far differently than you do,"

Actually I prefaced with something that I thought was a very nice writing on parenthood, and I think you can find any number of other good and thoughtful things he's written. I just also disagree on that specific issue being a clear he-is-awful.

I don't think a referendum on Atwood is appropriate here but certainly he's a driving force behind this endeavor, so if you're inclined to think he's a straight-up jackhole that seems pertinent. Personally I don't, and I think that when he swings and whiffs his heart is still in the right place or at least is looking at a noble destination.

Gruber and Arment, in the discussed podcast, on the other hand, come across to me as "I don't have a problem so everything is fine and too bad if you have problems which I deny there are any." And that's a notable improvement on his demeanor about any Markdown stuff for years, including any number of polite efforts from Atwood to get him involved. Hell, if he'd just devoted 1/10th the time of that podcast on writing back a flat refusal to Atwood - the one paragraph Atwood claimed he wanted - saying no, I am not okay with this and don't use the name... well then I guess he wouldn't get to make hay out of it now?

Obviously the case is made by many that he owes them no answer, but it seems really counter-intuitive to me to invest the time in creating an open project and release it into the world with the hope that it would make things better and then turn around and resist any input from those people about the ways it's 2% off from perfect.
posted by phearlez at 2:17 PM on September 4, 2014 [2 favorites]

From Hacker News, Atwood wrote this in response to "I would like to think that they at least approached him."

We did -- no response to the current spec after a week of waiting for a reply.

A much earlier lengthy email response from Gruber around 11/2012 evaluated each of the changes Stack Overflow and GitHub made to Markdown one by one (there were some he agreed with) and essentially said "ambiguity is a feature" at the end.

We disagree about that.

posted by smackfu at 2:33 PM on September 4, 2014 [1 favorite]

I'm disappointed by how many people don't care to understand Gruber's reasoning in this. It's easy to mock people you don't agree with, and I find this happening disproportionately to Apple-affiliated folks. ("His sidekick" Marco Arment? Seriously?)

Yes, ambiguity is a feature in this case. Markdown was designed from the top down by a writer, for writers. This gives it the rare characteristic of being intuitive to understand and use, but not necessarily easy (or even unambiguously possible) to implement. The lack of a standard helps keep people honest by making them follow the spirit of the law, not the letter, as you see happening with things like HTML and CSS. In fact, I'd argue that's the entire philosophy behind Markdown: usable and understandable by ordinary humans, all the way down to the spec. (Ever try reading the C spec? It feels like you need a PhD to get anything out of it.)

I'm with Gruber that Markdown is successful because of its ambiguous (read: easy-to-understand) design, not in spite of it. The proof is clearly in the pudding. My feeling is that the people running into edge cases are trying to use Markdown as some sort of replacement for HTML, not a massively simplified syntax for writing articles and pages.

What I see here is a group of developers who arrogantly think that they know better than Gruber what to do with Markdown. They could have called it anything else, but they chose to make the issue political. It's juvenile, especially coming from people who we're supposed to look up to in the tech community.

According to Gruber, Markdown is the hardest thing he's ever worked on. He didn't spend a week on it, the idea was not obvious, and its success was not accidental. If I was him, I'd absolutely be seeing red too. Have some respect for the man.
posted by archagon at 2:41 PM on September 4, 2014 [9 favorites]

But it is not intuitive to understand and use. Every time I'm posting somewhere and want anything more complicated than bold or italics, I have to look up the syntax, and fairly often I'll hit on something that isn't well specified and be reduced experimenting to determine the actual behavior of the implementation in front of me.
posted by eruonna at 2:58 PM on September 4, 2014 [2 favorites]

Really? What sorts of things are you trying to do? Barring occasional linking mishaps, I've almost never had a problem with Markdown in all my writing. (But again, that's mostly because it's so simple: paragraphs, headings, links, images, lists, quotes, code, and a bit of text formatting. You'd have to try pretty hard to screw that up.)
posted by archagon at 3:00 PM on September 4, 2014

the idea was not obvious

What was "the idea"?

Maybe I'm missing something here (I don't mean that in a sarcastic sense, I mean it genuinely, and if I am missing something then I'd appreciate an explanation), but Markdown has always seemed to me like yet another markup language. What's "the idea" that differentiates it from, say, Wikitext? Which had been around for, what, a decade or so by the time Markdown came around? I mean, obviously they're different syntaxes, but that doesn't really seem to be any sort of grand "idea" to me.
posted by Flunkie at 3:03 PM on September 4, 2014 [5 favorites]

According to Gruber, Markdown is the hardest thing he's ever worked on. He didn't spend a week on it, the idea was not obvious, and its success was not accidental. If I was him, I'd absolutely be seeing red too. Have some respect for the man.

If he's not going to maintain it, getting pissy when other people do so is a bit silly.
posted by sebastienbailard at 3:19 PM on September 4, 2014 [4 favorites]

The transcript is definitely appreciated. I don't think it helps my perception of Gruber's side of the argument though.

Gruber: [...] There's no need for a Markdown 2.0. And there's no need for a standard or a spec, but people get really worked up about it.

Yeah, I think I see where the disagreement stems from, right there.

I mean, the fact that people are "really worked up about" having a standard and a specification would seem to invalidate his first point that there's no need for one. If people want a standard, then clearly there's a need for one; I don't think people go around making up standards just because, which seems to be what he's implying.

Never have I met a programmer who really enjoys writing up standards documents. I'm sure that person exists, somewhere, and maybe it's even Atwood (?), but I don't think he'd have gotten any traction if there wasn't demand among implementors for a decent reference implementation.

Maybe the way people are using Markdown isn't the way Gruber intends Markdown be used, or he doesn't quite appreciate the ways in which they want to use it and which his laissez faire attitude towards compatibility doesn't allow for. But it seems pretty shortsighted to just say that "there's no need for a standard" when there are people clamoring for a standard.

I can't really take the whinging over the name seriously when he just hand-waves away the crux of the issue like that.
posted by Kadin2048 at 3:20 PM on September 4, 2014 [12 favorites]

I'm disappointed by how many people don't care to understand the reasoning for Standard Markdown. It's easy to mock people you don't agree with, and I find this happening disproportionately to people who are slightly critical of anything or any folks affiliated with Apple.
posted by juiceCake at 3:22 PM on September 4, 2014 [2 favorites]

Seems like the RSS nonsense all over again.

It is completely unfair to compare Gruber to Dave Winer. Gruber has said exactly seven words via twitter about this.

What coders apparently fail to realize is that Markdown is intended as a human-readable markup. The primary goal is to make it easy to write Markdown by mere mortals who aren't code geeks. The way it does this is with an extremely restricted feature set. It has ten features. I can remember how to use ten features. And that is the whole point. It turns piles of inscrutable HTML code for blocks and tables into simple markup tags.

So far there has been exactly one feature in an extension to Markdown that I approve of and use regularly: ~~Strikeout~~.
posted by charlie don't surf at 3:23 PM on September 4, 2014 [1 favorite]

Yes, ambiguity is a feature in this case

Did you look at the examples of underspecification? There's no obvious correct-according-to-spirit interpretation here; it's just a morass.
posted by kenko at 3:24 PM on September 4, 2014 [3 favorites]

> If he's not going to maintain it, getting pissy when other people do so is a bit silly.

It's his project. He considers it (mostly) complete. Evidence points to it (mostly) working great for (most) people. Nobody has the right to just swoop in and patronizingly say "great work, but we'll take it from here".

> I'm disappointed by how many people don't care to understand the reasoning for Standard Markdown. It's easy to mock people you don't agree with, and I find this happening disproportionately to people who are slightly critical of anything or any folks affiliated with Apple.

Thanks for contributing to the discussion.
posted by archagon at 3:24 PM on September 4, 2014

If people want a standard, then clearly there's a need for one; I don't think people go around making up standards just because, which seems to be what he's implying.

Except there isn't a clear need for one.

Standards.
posted by cjorgensen at 3:25 PM on September 4, 2014 [1 favorite]

Also, per Gruber, "If they pick a name as unique from 'RSS' as 'Atom', there will be no drama at all. I’ll wish them good luck." All this drama can so simply be avoided, but no, let's all thump our chests at each other over some bullshit.
posted by archagon at 3:26 PM on September 4, 2014 [1 favorite]

> Really? What sorts of things are you trying to do?

I recall having a lot of issues with lists, but I don't have specific examples.
posted by eruonna at 3:31 PM on September 4, 2014

cjorgensen: "Except there isn't a clear need for one."

That's... not at all analogous to what's going on here. "Standard Markdown" is exactly the same as the original Markdown, except that some ambiguities have been clarified.

No existing Markdown parsers will become any more broken than they already are as a result of "Standard Markdown."

And, yes. There is very clearly a need for a standard, given that Standard Markdown was created by representatives of Github, StackExchange, and Reddit, which are (by far) the three largest groups of Markdown users.
posted by schmod at 3:36 PM on September 4, 2014 [7 favorites]

I liked Gruber's sly commentary on the TwitPic shutdown this afternoon.
posted by Ian A.T. at 3:39 PM on September 4, 2014 [1 favorite]

My feeling is that this is a pet project of these individuals, not their respective companies. I doubt companies as big as Github, SE, and Reddit would adopt a Markdown implementation with possible legal/trademark issues, especially when an alternative like MultiMarkdown is clean.
posted by archagon at 3:39 PM on September 4, 2014

He considers it (mostly) complete.

His consideration does not make it actually mostly complete. His own perl implementation is buggy. His spec is extremely loose, such that the same short text produced 15 different outputs when fed into 22 renderers. It may work ok for many people but that doesn't mean that there's no room for improvement.

And you know what? If he thinks it's mostly complete and therefore has stopped working on it, that's a great reason for the people who want to fix it up to swoop in and work on it more, because he won't.
posted by kenko at 3:41 PM on September 4, 2014 [3 favorites]

Aaron Boodman, creator of Greasemonkey, has an interesting perspective. "I realized that [Greasemonkey] had become bigger than my original intent, and that it was going to continue in that direction with or without me."

("His sidekick" Marco Arment? Seriously?)

I chose those words carefully, and generally I have a lot of respect for Marco. In this podcast he's playing the role of sidekick, as Ed McMahon or Andy Richter as ever has been. I imagine the rest of the podcast has him offering more of interest. But for this segment, I chose an ill-tempered characterization because the kind of neckbeard mockery he was doing makes me angry.

given that Standard Markdown was created by representatives of Github, StackExchange, and Reddit, which are (by far) the three largest groups of Markdown users.

End of the day that's the only thing that's going to matter. As long as I can post a link to a Wikipedia URL with the same markup on every site, I'm happy.
posted by Nelson at 3:42 PM on September 4, 2014 [4 favorites]

Gruber just stepped up the game

I’d still object to that but not as vehemently. “Standard Markdown” is flat-out theft, and a lie. (2 lies, actually.)

and is pouncing forth with personal attacks at Atwood.
posted by xcasex at 3:42 PM on September 4, 2014

Atwood started with the stupid personal attacks, if you look at his tweet history.

Plus, there's his passive-aggressive "we're going ahead with or without you!" attitude which I imagine is hard to stomach when the (arguable) theft of your project name is on the line.
posted by archagon at 3:44 PM on September 4, 2014 [1 favorite]

archagon: "He didn't spend a week on it, the idea was not obvious, and its success was not accidental. If I was him, I'd absolutely be seeing red too. "

If you don't want people copying or modifying your work, the BSD license is probably a bad choice.

(That being said, "Standard Markdown" is in a bit of a gray area when it comes to the BSD license's endorsement clause.)
posted by schmod at 3:44 PM on September 4, 2014 [3 favorites]

I wonder if MultiMarkdown got "specific prior written permission" or not?
posted by jason_steakums at 3:50 PM on September 4, 2014

the same short text produced 15 different outputs when fed into 22 renderers

To be fair, a subset of those differences are really not significant: whitespace, the id attribute on the h1. Surprising how many of them get the numbered list wrong though.
posted by We had a deal, Kyle at 3:54 PM on September 4, 2014

So in summary, MultiMarkdown is ok, but Standard Markdown is horrid, because.. reasons. (not withstanding that both projects share the same author)
Standard Markdown even goes as far as to view Grubers Markdown as _canonical_.

then we have the whole bit about trademark enforcement, if the trademark isnt enforced, you can't just pop up out of irrelevancy a decade later and claim it.

as far as Gruber himself is concerned Markdown under his stewardship is _complete_ it's done.

So my interjection is quite simply, why not just be a good shepherd and let it evolve on its own, if you have no interest in it anymore?

and for chrissakes, it's too much of a storm in a teacup full of schmemantics.
posted by xcasex at 3:54 PM on September 4, 2014

Call it Very Strongly Suggested Markdown
posted by jason_steakums at 3:57 PM on September 4, 2014 [2 favorites]

I think it's pretty clear that MultiMarkdown is a fork (for "Multi" doesn't mean anything by itself), whereas Standard Markdown sounds like an official revision of Markdown, which it's not.
posted by archagon at 4:02 PM on September 4, 2014

I think it's pretty clear that MultiMarkdown is a fork (for "Multi" doesn't mean anything by itself), whereas Standard Markdown sounds like an official revision of Markdown, which it's not.

you can't cherrypick trademark enforcement, either you do it every time, or you don't. and if you don't, you lose the trademark.

but i'm in the camp that sees good things in this, so our thoughts concerning this debacle will differ.. wildly.
posted by xcasex at 4:04 PM on September 4, 2014

That is not accurate re: trademark enforcement
posted by phearlez at 4:10 PM on September 4, 2014 [1 favorite]

The standardizers just need a name to mark their work as their own, say, MarkedOwn.
posted by Zed at 4:19 PM on September 4, 2014 [7 favorites]

That is not accurate re: trademark enforcement

you're right! :D
and it'd seem i'm in the wrong, the correct wording on my part would be "can result in a loss of protection" not loss of trademark.
posted by xcasex at 4:20 PM on September 4, 2014 [1 favorite]

Shmarkdown
posted by jason_steakums at 4:26 PM on September 4, 2014 [3 favorites]

and thus Common Markdown was born.
posted by xcasex at 4:36 PM on September 4, 2014 [4 favorites]

I kind of love the bit in the Common Markdown post where they originally emailed Gruber, got no response for two weeks, and then decided to proceed - and then all this ugly business ensued - and they basically did the same thing after emailing him last night with a list of proposed names and not getting a response by today.
posted by jason_steakums at 4:45 PM on September 4, 2014 [5 favorites]

I kind of love the bit in the Common Markdown post where they originally emailed Gruber, got no response for two weeks, and then decided to proceed - and then all this ugly business ensued - and they basically did the same thing after emailing him last night with a list of proposed names and not getting a response by today.

Yeah, I don't like how Gruber is handling this at all. he's coming off worse and worse.
posted by xcasex at 4:46 PM on September 4, 2014 [1 favorite]

Oh given how nice Gruber has been about MultiMarkdown I'm sure it'll be just hunky dory.
posted by phearlez at 4:50 PM on September 4, 2014

I assume the name "Common Markdown" was at least partially chosen to suggest Common Lisp, another name born out of a prickly community of minority computer nerds.
posted by Nelson at 4:50 PM on September 4, 2014 [3 favorites]

I'm also amused that Jeff Atwood keeps going ahead with "no response must mean agreement, right?" despite all evidence to the contrary.
posted by smackfu at 5:20 PM on September 4, 2014 [4 favorites]

I mean, these are not coworkers, they are effectively strangers. No one should expect an email response if they ask a stranger to do something.
posted by smackfu at 5:23 PM on September 4, 2014

I think there's a lot of between the lines in that post as Atwood tries to be nice, and the lack of response action amounts up to "we are not going to let you grind this effort to a halt by only ever voicing objections."
posted by phearlez at 5:26 PM on September 4, 2014 [8 favorites]

Or for that matter by not even voicing objections.
posted by kenko at 5:42 PM on September 4, 2014

Like Zed I'm from the Textile camp, and while it's good to see Markdown get standardized (seriously, are people really advocating against standardizing this?) it makes me a little sad that this is likely to wind up pushing Textile further out to pasture, because I feel like it gets a lot of things right that Markdown misses:

*bold* and _italics_ are intuitive

Ordered and unordered lists are concise and unambiguous:

# ordered list
## sublist

* unordered list
** sublist

Images are simple: !image.png!

Also, considering that Markdown is virtually always targeted at HTML output, the lack of support for things like tables and CSS seems a bit delusional. There are variants that remedy this, but Textile makes it so much easier to produce something more than lowest-common-denominator output.
posted by bjrubble at 5:51 PM on September 4, 2014 [4 favorites]

Well that apology was actually more reasonable than I expected.

We still have the problem of Yet Another Standard (xkcd link above) but at least they're backing off from the most offensive aspect, so good for them.
posted by RedOrGreen at 5:53 PM on September 4, 2014

In the vein of "now there are", if you look at markdown.pl, you get...

# Regex to match balanced [brackets]. See Friedl's
# "Mastering Regular Expressions", 2nd Ed., pp. 328-331.
my $g_nested_brackets;
$g_nested_brackets = qr{
	(?> 					# Atomic matching
	   [^\[\]]+				# Anything other than brackets
	 | 
	   \[
		 (??{ $g_nested_brackets })	# Recursive set of nested brackets
	   \]
	)*
}x;

I think we had "two problems" in 2004 already.
posted by Monday, stony Monday at 6:44 PM on September 4, 2014 [3 favorites]

Oh, yeah. Textile was pretty great, and had the advantage of being first-to-market, and is still actively maintained.

That said, I think that Markdown is slightly more intuitive overall -- a few Textile-isms were really hard to remember (URLs in particular), and Markdown looks a bit nicer when expressed as plaintext.

I also won't forgive Atlassian for the bastardized version of Textile that they use in JIRA.... Their vendor-proprietary extensions are *really* hard to remember,
posted by schmod at 6:48 PM on September 4, 2014 [1 favorite]

The problem with Gruber is that he doesn't dialogue. He sneers. I don't know how you talk to someone like that.
posted by kbanas at 6:48 PM on September 4, 2014 [9 favorites]

You ask them to parse http://ddos-link.com/[test.......................................] with their nice regex? (Given its heavy reliance on the NFA features of Perl's regexes, Markdown.pl is probably also vulnerable to similar attacks).

Unfortunately, John Gruber doesn't seem to be familiar with langsec.
posted by Monday, stony Monday at 6:58 PM on September 4, 2014 [3 favorites]

I'm also amused that Jeff Atwood keeps going ahead with "no response must mean agreement, right?" despite all evidence to the contrary.

I dunno, I read that post and kind of got the idea that Gruber was an asshole who would never agree to anything at all, and that some concessions were being made for the sake of unwarranted politeness but seriously fuck that guy. YMMV.
posted by leopard at 7:16 PM on September 4, 2014 [3 favorites]

Atwood wrote that post so you'd think Gruber was a jerk. Now he may be right, but his rhetorical skill should not be underestimated.

But none of this crap matters. Does Common Markdown add value? Is it a better specified, clearer piece of tech? Will people use it? That's what matters, not whether Johnny and Jeffy are squabbling.
posted by Nelson at 8:27 PM on September 4, 2014 [2 favorites]

Is Atwood ever not terrible? Between the Shanley crap, "only special unicorns can be programmers" (citing a study the author himself repudiated) and this, he just never seems not-gross.

...

Yes, the 95% of the time he writes about programming.

When I first saw the Coding Horror blog title and logo, I thought it was related to or endorsed by Steve McConnell, who used that graphic and phrase in Code Complete. At the time, Atwood was not actually a very strong technically, and it seemed like the attention his blog got was totally unwarranted and there was this really weird dissonance with its seeming association with McConnell*. My initial impression of Atwood was therefore not very good. He's done other variously douchey things, but over time he has at least improved as a programmer.

*There isn't any actual relationship; Atwood just wanted to use the title and logo and McConnell gave him permission.

One of the motivations for looking into this and posting it was thinking "huh, this is something I actually agree with Atwood on". I can't discount the possibility that I am taking his side because of my dislike of Gruber, but I think that's somewhat unlikely because Markdown is not very closely linked with Gruber in my mind. The reason for that, in turn, is that Gruber has been completely uninvolved with anything Markdown for years. That makes me less inclined to take his pissing and moaning seriously.

The idea that [lacking a standard is an advantage] is, at first glance, repellent, but it should be engaged with.

One can imagine the ambiguities in the standard as presenting possible choices to each implementer. For some ambiguities, there will be choices that are (obviously or subtly) better than others. Let us call these ambiguities "improvable". For some ambiguities, there will be a number of possible choices that are not compatible with one another, but none of the choices are better than the others (for everybody). Let's call these ambiguities "arbitrary".

We can imagine different implementations of Markdown copying the choices of other implementations, hopefully selecting the best known choices for improvable ambiguities. Over time, the population of Markdown implementations could, in the absence of a standard, tend towards the best possible choices for improvable ambiguities. The improvement in improvable ambiguities represents a positive utility (in a collective sense) that increases over time.

However, the change in implementations produces incompatibilities between implementations. This represents a negative utility that decreases over time (as implementations stabilize towards the best possible choices for improvable ambiguities).

Furthermore, the arbitrary ambiguities provide a constant source of incompatibilities- constant, at least, until a standard is adopted, at which point they go down.

At some point, the benefit you get from not standardizing (you might find an even better set of choices in your improvable ambiguities) will not exceed the penalty you're paying from the arbitrary ambiguities. The question isn't really "should we standardize?"; it's "when should we standardize?"

When the major players start complaining about the lack of a standard, as they have, I'm inclined to believe that there is not much more fertile ground to be tilled re improvable ambiguities and it's time to coordinate to get rid of the arbitrary ambiguities.
posted by a snickering nuthatch at 8:44 PM on September 4, 2014 [3 favorites]

As someone who only now understands why she's always had to google markdown syntax (and my blog is in octopress for feck's sake), I like the idea of some kind of common dialect.

As for Atwood somehow being awful for moving forward on a rename so quick, I think that's unfair. If Gruber is actively upset about it and has said he's fine with the project if they lose the name "Standard Markdown" and has asked that the domain name be pulled AND the project owners have agreed to do all that, I think it's reasonable that communication on working out the name change be prompt.
posted by R343L at 8:57 PM on September 4, 2014 [2 favorites]

I've never looked very closely at Markdown but every time I come across it the first thing I wonder is why not wiki markup or some subset of it instead of re-inventing the wheel? It had already been around for ten years when Gruber came out with Markdown.pl.

And is the equivalent of

=== Third level header ===

seriously

h3. Third level header

in Markdown, as claimed by the Wikipedia article? Some aspects of it actually seem like a step backwards as far as intuitiveness, but if it's now standardized maybe there will be improvements.
posted by XMLicious at 9:04 PM on September 4, 2014

Does it even matter what it's called? If Github, Reddit, and StackExchange start enforcing a higher standard, a lot of implementations will follow, and who cares about the ones who don't?

Seems like even Gruber fans will run into an ambiguity, have to make a decision how to resolve it, and what, refuse to go with the new spec out of spite? And if they don't run into any ambiguities, then... it's a non-conflict?

I kind of feel like Gruber's being an insufferable dick, but also that changing the name wouldn't be that hard either. With all the added publicity now, it's not like ForkDown is going to languish from nobody knowing about it.
posted by ctmf at 9:48 PM on September 4, 2014

Atwood's group has agreed to change the name from "Standard Markdown" to "Common Markdown" which seems much less an act of out and out appropriation and more like an act of community-minded populism.

I think the name change is a great move. Respectful. Accurate. On point. KUDOS.

I don't think Gruber is wrong in his opinion. The widespread adoption of Markdown is likely partly due to the fact that it's dead simple to understand and that simplicity heavily outweighs its few (3 according to the "Common Markdown" project) specification flaws.

I also don't think Gruber is being arrogant; he's being defensive about territory he's staked out and I think he's entitled to feel as he does. (He is a bit dismissive of Atwood but I think Atwood invites Gruber's dismissiveness, as I explain below.) That doesn't mean I think a "better" implementation should not be crafted without him.

Dissatisfied users of Markdown should have access to a version that clarifies unacceptable ambiguities and, hopefully, Atwood's group can deliver this to users so dissatisfied.

Atwood is a bit presumptuous and condescending when he calls Gruber out and Atwood's acting as if he can project manage Gruber into working for his version of Markdown, whatever it will be called, really deserves to be ignored. Seriously.

Another way to think about Atwood's presumption is imagining the effect of Atwood cleaning up the PERL Gruber wrote oh-so-long ago and emailing Gruber privately as a contribution to Markdown.

Funny right?

The absurdity of such scenario indirectly reveals Atwood's ambition to forge a project of his own and, make no mistake, ambition and impatience are written everywhere in Atwood's calls against Gruber's inscrutable inertia, as it were.

At least "Common Markdown" isn't a "pull the rug out from under" move as "Standard Markdown" (the chutzpah!) and I hope Atwood's group's project satisfies those users and developers who really need something better than Markdown.

I'd love to see such a thing come to be.
posted by mistersquid at 10:22 PM on September 4, 2014 [2 favorites]

Gruber is doubling down now >_>
posted by xcasex at 10:32 PM on September 4, 2014

I kind of feel like Gruber's being an insufferable dick, but also that changing the name wouldn't be that hard either. With all the added publicity now, it's not like ForkDown is going to languish from nobody knowing about it.

So if App.net changed its name to "Standard Twitter" you'd be ok with that?

They invented Trademark laws because people would sometimes be a dick and steal the name. Sadly, the absence of trademark seems to be license for "I can be a dick and steal this name".
posted by Talez at 10:49 PM on September 4, 2014

The widespread adoption of Markdown is likely partly due to the fact that it's dead simple to understand and that simplicity heavily outweighs its few (3 according to the "Common Markdown" project) specification flaws.

I count thirteen, and you know what? Cleaning those up won't make it more complicated. It will, in fact, make it simpler, by codifying how various things interact, which are currently not codified.
posted by kenko at 11:03 PM on September 4, 2014 [2 favorites]

Look at number six on that list, for instance. Which is it? Not having an answer to that question is a mark of complexity (born of a failure to think things through), not simplicity.
posted by kenko at 11:04 PM on September 4, 2014

Sadly, the absence of trademark seems to be license for "I can be a dick and steal this name".

Markdown is trademarked? That changes this whole conversation. Gruber is entirely justified in his actions then.
posted by 99_ at 11:15 PM on September 4, 2014

I went in to edit when I realized that I misread the comment, but you know, rather than try any other tack, I'm only going to say, I don't even know. I wish we could find a way to use the energy these disputes generate for anything else.
posted by 99_ at 11:27 PM on September 4, 2014 [1 favorite]

Markdown is trademarked? That changes this whole conversation. Gruber is entirely justified in his actions then.

It's not. What I'm saying is just because Gruber hasn't trademarked it while calling something "x Markdown" is a legitimate action, it doesn't make it any less of a dick move.
posted by Talez at 11:29 PM on September 4, 2014

If you think Gruber is unreasonable and exclusively snarky, you haven't listened to him in conversation. He's often very reasonable and garners a lot of respect from his peers. I imagine an e-mail along the lines of "hey, we're a small group of people who want to make a fork of Markdown with good documentation and we want to give it a name that doesn't confuse people about what it is, how's this for a few suggestions?"*, or even just a less grand-standing project, would have (eventually) gotten a positive response from him.

Also, given Atwood's previous conversations with Gruber about standardizing Markdown two years ago (which didn't end well), I find it very difficult to imagine that he thought this wouldn't happen. Atwood's a smart guy, and it's pretty obvious to anyone who knows Gruber at all what kind of reaction "Standard Markdown" would elicit. The cynical part of me think it's some sort of weird PR move.

Anyway, it looks like Gruber is OK with "Common Markdown", so I guess we can all go home.

* In his recent tweets, he said he got one e-mail from this group, and that he was on vacation when he got it. Plus, who reads e-mail anymore? Tweet at him or something.
posted by archagon at 11:34 PM on September 4, 2014

Underlying this whole discussion is an apparent conflict between an individualist ethos so common among classic programmers versus a community-minded perspective. What matters more - one person's "ownership" of a name and concept, or a community having a more functional shared tool/language?
posted by parudox at 12:47 AM on September 5, 2014 [2 favorites]

What matters more - one person's "ownership" of a name and concept, or a community having a more functional shared tool/language?

the latter.

Anyway, it looks like Gruber is OK with "Common Markdown", so I guess we can all go home.

that's how you read these tweets?

By my reckoning that's what us common folk call "vehemetly opposed".

Gruber knew this was coming, he said as much on the podcast episode, that he saw the discussion being held on the mailing-list.
If a person knows this is an ongoing effort, knows that the group has emailed him (and by his own words yet again, answers emails) why wait to answer?
to me it just seems like he (Gruber) wanted to compound a bad image for the group behind the effort by dragging it out.

But what it comes down to is one person who's doubling down, like last time & the time before that, and a person who's willing to go further and make a community effort.

I just cannot see a downside to a community effort, for something that today is in wide use.
posted by xcasex at 1:00 AM on September 5, 2014

We should just be writing everything in LaTeX anyway...
posted by PenDevil at 1:19 AM on September 5, 2014 [2 favorites]

> that's how you read these tweets?

I'm pretty sure all those tweets were posted before the apology, in response to the original title. And @codinghorror said they finally got a dialogue going over e-mail.
posted by archagon at 1:28 AM on September 5, 2014

Also, don't confuse "community" with "a tiny, vocal pack of programmers".
posted by archagon at 1:33 AM on September 5, 2014 [2 favorites]

those tweets were posted _after_ the apology.
posted by xcasex at 1:41 AM on September 5, 2014 [1 favorite]

Doesn't mean Gruber saw it. They were clearly continuing a previous conversation, not responding to new information.
posted by archagon at 1:42 AM on September 5, 2014

I was hoping they’d go with "Dickheads Who Steal the Project Names of Other People’s Work as a Shameless Publicity Ploy Markdown”.

-- Amy Jane Gruber, moments after the apology was posted.
posted by xcasex at 1:49 AM on September 5, 2014

"Bottom line: what matters is what Reddit and GitHub implement..."

Most depressing thought of the day.
posted by axon at 3:40 AM on September 5, 2014 [5 favorites]

Markup-lite is terrible. It should be no surprise to anyone, because markup-heavy, no matter how well and how thoroughly it repairs the deficiencies of markup-lite, will also be terrible. We've learned this lesson dozens of times in the last ... oh god, 35 years. Likely even longer. ohgodohgodohgod I can't even
posted by wobh at 3:55 AM on September 5, 2014 [2 favorites]

Amen to that, wobh. Little markup languages are great for their very narrow applications, but as soon as someone tries to create an extension, blammo! instant complexity. You get complete, or you get consistent in this world. Not both.

I get that plain text is easy to write if you think that a web textbox or vi is the wasp's nuts of text editing. But once you want to apply structure beyond the most very basic typewriter conventions, you start having to apply markup, and it will get ugly. Hit yourself over the head with (or slightly less pleasantly, read) a copy of Goldfarb until you attain enlightenment. There are no easy solutions to a hard problem. Implicit markup will always be misunderstood by different people, because your conventions ain't always my conventions.

Markdown isn't solving any problems that can't be marked up in 1995-vintage HTML 2.0. If your document requires more structure than that (such as sections, or heaven forfend, tables) you are way beyond Markdown's domain. All Markdown's really doing is skipping HTML's insistence of starting a paragraph block element with <p>. The other conventions in Markdown can be just as succinctly represented as HTML, especially if you consider using the old SGML </> convention for “close the last inline element I used”. No web browser supports that, but it's a handy trick to use in an SGML-aware editor.
posted by scruss at 5:27 AM on September 5, 2014 [2 favorites]

A very educational post and thread. With drama even. Almost makes me want to get a Twitter account. Almost.

I'm on the side of Making Things Better. I believe that creating and putting something out into the OS community is massively noble, but also requires an acceptance that your child may go on to have a life of its own and interact and be influenced by other people.

Changing 'Standard' to 'Common' is a pretty reasonable concession to the originator's feelings. Although their goal is to create a canonical implementation, 'Common Markdown' is still just another Markdown implementation at this point (with grand intentions) and any further move by Gruber to shut this down would truly be petty. But humans are petty...
posted by Artful Codger at 5:36 AM on September 5, 2014

Bottom line: what matters is what Reddit and GitHub implement as Markdown.

To me, the big websites that use Markdown aren't especially relevant as far as standardization, despite being big names. The average user on Reddit or Stack Overflow or GitHub is using maybe 10% of the Markdown spec in their comments (stuff that just looks like email formatting from 1990), and likely do not even know they are using Markdown. And all three of these sites are using customizations today that aren't included in the spec. Standardizing may please their developers but the users won't care.

OTOH, the more relevant case for a standardized Markdown is when people start using it as a generic document format. Then it becomes very important that what you write in Editor X is rendered properly in Blog Engine Y. And that if you migrate to Blog Engine Z, it doesn't change all your formatting.

Maybe they should have down something like MarkdownDoc, to explicitly address the second case, and also support metadata like author and title.
posted by smackfu at 5:36 AM on September 5, 2014

> I'm with Gruber that Markdown is successful because of its ambiguous (read: easy-to-understand) design

Has anyone made a cogent defense of the "ambiguities are a feature" stance? It seems totally insane to me, but maybe it's just because I haven't seen it explained (just stated).
posted by Horselover Fat at 5:58 AM on September 5, 2014

Oh, and I forgot to say: "ambiguous" and "easy-to-understand" are in no way synonymous. Arguably there's a negative correlation.
posted by Horselover Fat at 6:02 AM on September 5, 2014 [2 favorites]

I think there is a parallel to some extent in the idea of "Worse Is Better".

The MIT guy did not see any code that handled this case and asked the New Jersey guy how the problem was handled. The New Jersey guy said that the Unix folks were aware of the problem, but the solution was for the system routine to always finish, but sometimes an error code would be returned that signaled that the system routine had failed to complete its action. A correct user program, then, had to check the error code to determine whether to simply try the system routine again. The MIT guy did not like this solution because it was not the right thing.

The New Jersey guy said that the Unix solution was right because the design philosophy of Unix was simplicity and that the right thing was too complex. Besides, programmers could easily insert this extra test and loop. The MIT guy pointed out that the implementation was simple but the interface to the functionality was complex. The New Jersey guy said that the right tradeoff has been selected in Unix-namely, implementation simplicity was more important than interface simplicity.

The MIT guy then muttered that sometimes it takes a tough man to make a tender chicken, but the New Jersey guy didn't understand (I'm not sure I do either).

posted by jenkinsEar at 6:34 AM on September 5, 2014

Appropriate that a personality like JWZ would be quoted in an issue like this, jenkinsEar. I'm not sure if he studied at Gruber's school of conflict management or the other way around but they're unquestionably brothers from another mother in that way.

Changing 'Standard' to 'Common' is a pretty reasonable concession to the originator's feelings. Although their goal is to create a canonical implementation, 'Common Markdown' is still just another Markdown implementation at this point (with grand intentions) and any further move by Gruber to shut this down would truly be petty. But humans are petty...

Nobody can prove a hypothetical, of course. But... this project includes Github, Stack Exchange, Reddit, Discourse and who knows how many other sites that, combined, probably are host to 90%+ of the Markdown written in a given year. I'd bet $100 that if they'd done what Gruber claims he thinks they should have in podcast and picked a whole new name, he'd be pissed off and snarking that they were now using this thing that was in all appearance exactly his creation with no proper nod to him.
posted by phearlez at 6:47 AM on September 5, 2014 [5 favorites]

Markdown isn't solving any problems that can't be marked up in 1995-vintage HTML 2.0. If your document requires more structure than that (such as sections, or heaven forfend, tables) you are way beyond Markdown's domain.

Certainly. However markdown's domain easily accounts for a large majority of texts produced. It's not ideal for technical or academic writing and it doesn't have a framework for dealing with long-form writing. Although in many cases, those gaps can be better dealt with at the second-draft stage using text pulled from a lighter system into SGML, XML, or LaTeX.

Many of us are writing in forms and genres where the gold standard for submission is only incrementally derivative of the typewritten page after all. Because all of that is going to get stripped down to bare text and recreated by a page designer to publication standards or shoved into a CSS framework.
posted by CBrachyrhynchos at 6:56 AM on September 5, 2014 [1 favorite]

Markdown isn't solving any problems that can't be marked up in 1995-vintage HTML 2.0. If your document requires more structure than that (such as sections, or heaven forfend, tables) you are way beyond Markdown's domain. All Markdown's really doing is skipping HTML's insistence of starting a paragraph block element with . The other conventions in Markdown can be just as succinctly represented as HTML, especially if you consider using the old SGML convention for “close the last inline element I used”. No web browser supports that, but it's a handy trick to use in an SGML-aware editor.

The problem that markdown solves is writing texts that are readable either as plaintext or fully rendered. Unrendered HTML is ugly.
posted by empath at 7:40 AM on September 5, 2014 [1 favorite]

And here's the other itch that Markdown scratches. My priorities when I'm working on my writing are sentence structure, character, dialogue, tension, conflict, balancing mystery with explanation, pacing and rhythm if I'm attempting poetry, allusion, metaphor, reference, rhetoric, voice, flow, and setting. Does the language say what needs to be said, what I want for it to say, and what the audience of the piece need to hear?

Messing around with structural and semantic markup, 99% of which can be generated by a script, gets in the way of that flow. And that flow is a precious constraint when you're working on meaning.
posted by CBrachyrhynchos at 8:01 AM on September 5, 2014

The drama is mostly over, but I wanted to point this out ... Gruber has stated repeatedly that markdown has grown steadily on its own, without any revision on his part, as evidence that no change is needed for the language.

However, look at google trends.


Reddit's commenting form has a direct link to the daringfireball.net markdown page, as does Github.

It seems more likely to me that the Markdown page is popular because a huge site like Reddit has made it popular, in spite of its shortcomings. If Reddit's developers are saying that they need more from the spec, that's a pretty good reason to listen to their needs. If Reddit and Github had decided to switch to their own Markdown-inspired language, I would not be surprised if inbound traffic to Gruber's site dropped a lot.

That might have been the best solution for everyone involved, really.
posted by rustcrumb at 9:06 AM on September 5, 2014

I'm 90% sure that if reddit and github and stack overflow changed the name to Standard Foo, everyone would call it Foo and forget that Gruber had anything to do with it.
posted by empath at 9:30 AM on September 5, 2014 [4 favorites]

empath, exactly. And then, as phearlez said, I'm sure Gruber would be mad that he was not credited enough.
posted by rustcrumb at 9:32 AM on September 5, 2014

Markdown isn't solving any problems that can't be marked up in 1995-vintage HTML 2.0.

The main problem it solves (for me, at least) is "HTML is fragile, messy, and awkward to type" -- which was as true in 1995 as it is today.

If your document requires more structure than that (such as sections, or heaven forfend, tables) you are way beyond Markdown's domain.

Why is that?

# Section 1

## Subsection 1.1

| Name | Email |
| Joe Blow | [e-mail] |
| Jane Foo | [e-mail] |

Tables, especially, are so much easier to deal with in lightweight markup. If someday my muscle memory for typing <table><thead><tr><th> disappears from disuse, I will do a little jig.

There are no easy solutions to a hard problem.

The whole point of systems like Markdown is that the vast majority of markup isn't "hard" at all, so a simple system is easily sufficient. Moreover, in many use cases (such as input from untrusted users) you really want to eliminate the "hard" parts anyway.

Implicit markup will always be misunderstood by different people, because your conventions ain't always my conventions.

Hence the efforts to (gasp!) standardize these conventions.
posted by bjrubble at 9:42 AM on September 5, 2014 [1 favorite]

And then people entering the community five years later would, after a while, realize "Hey, Foo is just a version of Markdown. I'd have called it something like "Markdown 2.0" — makes it easier to see the connection. Why the heck did they change the name to something so different. " And the olds would groan and provide links to the tweets.
posted by benito.strauss at 9:44 AM on September 5, 2014

If you don't think there needs to be a standard... don't use it.

If you don't care about the standard because you are just going to write whatever... don't use the standard.

The only people who really need to care about the standard are engineers worried about things like portability, WHICH MATTERS. I mean if all you do is post to metafilter you don't care that much about how html edge cases work, but if noone cares about then you are stuck in the 90s web.
posted by aspo at 10:00 AM on September 5, 2014 [2 favorites]

Tables, especially, are so much easier to deal with in lightweight markup.

Incidentally, tables are in neither Gruber's spec nor the Common Markdown spec.
posted by a snickering nuthatch at 10:09 AM on September 5, 2014 [2 favorites]

We should send Gruber some good vibrations and name it MarkyMarkdown.
posted by ryoshu at 10:31 AM on September 5, 2014 [5 favorites]

> # Section 1
   ## Subsection 1.1

Those generate H1 and H2 elements. You'd have to presume that each top-level heading starts a new section, which it may not do. Compare with Docbook's section or TEI's divisions. That, and the table thing not being in any Markdown I've used.
posted by scruss at 10:35 AM on September 5, 2014

I'm 90% sure that if reddit and github and stack overflow changed the name to Standard Foo, everyone would call it Foo and forget that Gruber had anything to do with it.

empath, exactly. And then, as phearlez said, I'm sure Gruber would be mad that he was not credited enough.

What if the new name was 'Gruber'?
posted by ctmf at 11:02 AM on September 5, 2014

What if the new name was Butt-Head Blogger?
posted by aspo at 11:07 AM on September 5, 2014 [1 favorite]

tables are in neither Gruber's spec nor the Common Markdown spec.

org-mode for the win!
posted by Zed at 11:26 AM on September 5, 2014 [2 favorites]

Here's a pretty good (and sober) article that generally corresponds with my thoughts on why (name stuff aside), Common Markdown is not the end of it all and why, if it actually gets any traction (which I sort of doubt, but we'll see), it could be actively problematic: Why Common Markdown isn't the Solution.
posted by Hartster at 12:35 PM on September 5, 2014 [3 favorites]

# Section 1
## Subsection 1.1

Those generate H1 and H2 elements. You'd have to presume that each top-level heading starts a new section, which it may not do. Compare with Docbook's section or TEI's divisions.

If by "sections" you mean "a full and well-formed document hierarchy" then yes, Markdown doesn't do that -- but neither does HTML. (At least not using HTML's own semantics; of course you can always use HTML as generic SGML to define your own structure, but in that case you arguably shouldn't be using HTML at all.)

That, and the table thing not being in any Markdown I've used.

I think most of the Markdown extensions support tables. MultiMarkdown and GitHub do, at least. But the question isn't so much whether tables are supported, as whether tables should be supported -- personally, I frequently use simple tables for which a radically simplified text formatting is perfectly sufficient, takes 1/10th the time to generate, and is easy to parse visually, so in my book this is exactly the kind of thing that Markdown should be targeting.
posted by bjrubble at 1:32 PM on September 5, 2014 [1 favorite]

I frequently use simple tables for which a radically simplified text formatting is perfectly sufficient

These days, even when I'm writing a program to generate tabular data I just generate org-mode source and read it in org-mode. (This is for my own use... probably end up using CSV if I were sharing it.)

All this talk is making me feel like writing up my own lightweight markup spec. Naturally, mine would do everything right!
posted by Zed at 1:46 PM on September 5, 2014

All this talk is making me feel like writing up my own lightweight markup spec. Naturally, mine would do everything right!

You could call it "Correct Markdown" -- I'm sure nobody would mind.
posted by bjrubble at 1:53 PM on September 5, 2014 [3 favorites]

These days, even when I'm writing a program to generate tabular data I just generate org-mode source and read it in org-mode. (This is for my own use... probably end up using CSV if I were sharing it.)

That's another bonus to drafting in a text-based mode. While you can script SGML or HTML without SAX or DOM, doing so is rather fragile. Even with a parser/writer system, you can get problems. I just spent the morning wrestling with a corrupted DOCX file that was hiding paragraphs based on cursor position. I might need to just strip everything back to text and reformat 50-odd pages.
posted by CBrachyrhynchos at 1:56 PM on September 5, 2014

Hey Metafilter, we like Andy Baio, right? The twitter conversation that ended with this tweet does a pretty good job of showing why I feel like Atwood is in the wrong in trying to co-opt the name: https://twitter.com/codinghorror/status/507814630303870976

Atwood just doesn't seem to understand Gruber's objection *at all*.
posted by macrael at 2:11 PM on September 5, 2014

You could call it "Correct Markdown" -- I'm sure nobody would mind.

Since it'd sort of be a 2.0, I was thinking MarkTwain. Imagine the searchability!
posted by Zed at 2:21 PM on September 5, 2014 [2 favorites]

It seems weird to release something on with a free license like BSD and then restrict the name from forks of it. I don't think anything works that way in the linux world.
posted by empath at 2:22 PM on September 5, 2014 [1 favorite]

Atwood just doesn't seem to understand Gruber's objection *at all*.

To be fair to Atwood, Gruber hasn't been great about articulating his objection directly. It's clear now, but Atwood is probably still stuck thinking about it in the context before it was clear. A mental block, if you will. Or he doesn't care, I guess.
posted by HiddenInput at 2:23 PM on September 5, 2014 [1 favorite]

It seems weird to release something on with a free license like BSD and then restrict the name from forks of it. I don't think anything works that way in the linux world.

I was wondering about this. Is there a precedent for this? It seems counter to the spirit of the BSD license to try to withhold the name from forked projects.


The way to win this would've been for Atwood et al to choose a different name. In a couple years nobody would remember Markdown and Gruber's contribution would be a footnote.

It sucks that the name is so good and apt but I bet they could've found something reasonable.
posted by wemayfreeze at 3:24 PM on September 5, 2014

That "house" metaphor really isn't working the way Atwood thinks it is.
posted by We had a deal, Kyle at 3:46 PM on September 5, 2014

I think adverse possession would be a better house metaphor.
posted by 99_ at 4:13 PM on September 5, 2014

And like that, our long national nightmare is over: "The Standard / Common Markdown project, in accordance with the wishes of @gruber, has been renamed to http://commonmark.org/"

https://twitter.com/codinghorror/status/508027568839479297
posted by macrael at 4:18 PM on September 5, 2014 [1 favorite]

Wait 'til you see my new book: "Dive in to CommonMark"!

Part of me suspects in a few years Gruber will regret his obstinancy. It'd be a shame if he's remembered as "the guy who made that thing before CommonMark, then abandoned it and fought the folks who did a bunch of great work on it." I think the CommonMark folks have bent over backwards to give Gruber credit and respect, so hopefully people won't forget. Or maybe he's right and the world really would prefer benign neglect.
posted by Nelson at 7:01 PM on September 5, 2014 [1 favorite]

CommonMark?? Dammit now I have to change the T-shirt design AGAIN! Geez.
posted by Artful Codger at 8:06 PM on September 5, 2014 [1 favorite]

hmmm. In the oldest markdown license page the Internet Archive has, Gruber said he was releasing it under the GPL, with a link to the GPL v2. The GPL v2 doesn't mention project names. A few months later, that became GPL v2 or, at your option, any later version. It was that way through at least Dec. 7, 2004. From Feb. 5, 2005 on we see the same as the current version which actually doesn't license it under a BSD license, but under what the page calls a "BSD-style" license. Its first two clauses are the same as those of the three-clause BSD license, but where the BSD three-clause has:

Neither the name of the nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

Gruber had:

Neither the name “Markdown” nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

Which papercrane called out above, but I was still distracted by thinking "BSD license."
posted by Zed at 11:22 PM on September 5, 2014

I have never even understood what problem Markdown is trying to solve. What's wrong with HTML? What percentage of people who write Markdown don't know HTML (or the minimal subset of HTML you need to know to be able to do anything you can do in Markdown)? It has got to be extremely low. Why do I have to spend my time figuring out how to get Markdown to spit out the HTML I want instead of just writing the HTML directly in the first place? I don't get the appeal.

Oh God oh God.

I feel like I came in late on this debate but I have so many feelings. Like, I have written extended articles on somewhat related topics.

The problem is that HTML is (relatively speaking) a very verbose markup language. It uses named elements with additional marker syntax to start and end every run of formatting. Its vocabulary is both larger than most simple copyediting tasks require, and missing syntax for very common elements like footnotes. It's also nigh impossible to read in 'raw markup' form.

Markdown renders to HTML, but is actually much closer in syntax to simple text markup standards like Setext and BBCode. They were designed to be used for 'pure runs of text', not document layout, they only support a small subset of HTML's vocabulary, and most implementations add extra bits that would take more complex markup. The result is that Markup-formatted text is perfectly readable in a plaintext email, a tweet, rendered to HTML, or printed to paper without any extra parsing. It's also hard to fuck up: you can mess up the formatting of your text in Markdown, but you can't easily break the layout of the entire page the text appears on. That's depressingly easy with HTML.

For these reasons, it's popular for user-generated content on some sites -- stuff like comments and user profiles.

The problem of course is that many people need or want a little bit more than markdown offers. ("I want captioned images," or "I need classes on my blockquotes" or "I realllllly really need random bits of CSS.") Gruber solved that by saying you could jam in HTML if you wanted to, as well. Many other parsers added additional syntax for stuff like IDs and classes. But steadily, that begins to compromise the reasons that make Markdown compelling.

In conclusion, if all you really need is simple formatting of text, and it will be passing through many systems and ad-hoc paths like IM and email and be reviewed in those forums, Markdown is really nice. But if you need more than that, or Markdown's syntax isn't rich enough to describe what you are 'saying' in your document, it's going to be super painful and frustrating.
posted by verb at 12:42 AM on September 6, 2014 [4 favorites]

Although I expect few here actually need this, here is a corrected link to verb's article The Battle for the Body Field for your convenience. I liked it.
posted by wobh at 2:07 AM on September 6, 2014

Can people stop saying something is 'perfectly readable' and use more relative terms? I think the confusion with Markdown and its value (over HTML) comes for this assertion. For very simple expressions of emphasis and other rudimentary typesetting in a browser it is a demonstrably an improvement over HTML for most people who write a lot of code, provided you learn and understand the dialect. Gruber's comments about ambiguity being a feature sort of acknowledged that everyone could develop their own set of rules as to what make the most 'intuitive' sense to them when writing that would ease their own workflow. So one person might prefer /italics/ to >italics< but neither is intrinsically better.

Because the rules came from from established communities does not mean it is intuitive or self-evident any more than arguing serial commas are just obviously correct (or not).
posted by 99_ at 8:05 AM on September 6, 2014

Whether it's /italics/ or >italics< I daresay a reader can work out that there's emphasis going on. Without contextual clues in the text, blockquotes might be hard to work out if you've never seen the '>' convention (as people increasingly haven't in our increasingly HTML-email world) and links and images are ugly, but the rest of it... headlines, lists, emphasis... I'm hard-pressed to imagine a reader being unable to read and understand a text in Markdown for their presence.

I agree, of course, that it's not at all intuitive precisely how one's supposed to produce Markdown -- learning that takes effort; it's harder to know what HTML it's supposed to represent (or, rather, impossible, hence the desire for a more exact spec). But even though I'm hostile to nearly all uses of the word 'intuitive' in regard to anything to do with computers, I'd consider it a fair statement that most of Markdown's syntax is intuitive to a reader.
posted by Zed at 9:51 AM on September 6, 2014

They should have called it GruberText.
posted by humanfont at 11:54 AM on September 6, 2014 [3 favorites]

Can people stop saying something is 'perfectly readable' and use more relative terms? I think the confusion with Markdown and its value (over HTML) comes for this assertion.

It's self-evident that it's easier to read than html, whether you know the conventions or not.
posted by empath at 1:59 PM on September 6, 2014

Justing using the example from their text:

# Hello there

This is a paragraph.

- one
- two
- three
- four

1. pirate
2. ninja
3. zombie

vs

<h1 id="hellothere">Hello there</h1>

<p>This is a paragraph.</p>

<ul>
<li>one</li>
<li>two</li>
<li>three</li>
<li><p>four</p></li>
<li><p>pirate</p></li>
<li>ninja</li>
<li>zombie</li>
</ul>

posted by empath at 2:02 PM on September 6, 2014

"Easier to read than HTML" != "perfectly readable"

Markdown is still a markup. My point was people who have a working knowledge of HTML, but perhaps not an advanced one (those tasked with creating text content for the web but not anything more advanced) may not see the value because it's not a direct correlation to WYSWIG editing. You are still adding characters to plain text that will be removed in the rendering process. You still need to have a frame of reference as to what can be done or not done (in terms of formatting) to do it. By saying 'perfectly readable' it seems to be implying that resultant text is styled (which it is not) and that it has no extraneous characters (also not).
posted by 99_ at 2:44 PM on September 6, 2014

empath, your HTML is a bit chubby (and broken; you skipped the <ul> to <ol> transition there). This is more like the thing:

<h1>Hello there</h1>
<p>This is a paragraph.
<ul>
<li>one
<li>two
<li>three
<li>four
</ul>
<ol>
<li>pirate
<li>ninja
<li>zombie
</ol>
posted by scruss at 3:47 PM on September 6, 2014 [2 favorites]

Thanks for contributing to the discussion.

You're welcome, and likewise.
posted by juiceCake at 11:50 PM on September 7, 2014

My point was people who have a working knowledge of HTML, but perhaps not an advanced one (those tasked with creating text content for the web but not anything more advanced) may not see the value because it's not a direct correlation to WYSWIG editing. You are still adding characters to plain text that will be removed in the rendering process. You still need to have a frame of reference as to what can be done or not done (in terms of formatting) to do it.

And yet somehow millions of people use it everyday, despite those crippling flaws.
posted by empath at 5:39 AM on September 8, 2014

just wondering if markdown is supported here in the comments.

nope, guess not. ok.

-bowerbird
posted by bbirdiman at 12:26 PM on September 23, 2014 [1 favorite]

>just wondering if markdown is supported here in the comments.

If you use it, *most* people will understand it.
posted by charlie don't surf at 3:55 PM on September 24, 2014 [1 favorite]

It's harkening back to Usenet formatting, which evolved naturally and easily - *italic*, **bold**, _underline_, -strikethrough-, ^superscript^ vsubscriptv =double strikethrough, or italic, or bold, or underlined, heck, let's just call it "em"= - after a while, your brain formats it for you.

I'd rather a modern graphical interface format it actually in a presentable way for most devices.

Dude. Just today, I used a screensharing service a customer had paid for. It took less time for me to see his entire desktop, scaled sensibly in my browser window, animated at full speed, than it took for Metafilter to load this thread. It took less time to load than Google's home page, for crying out loud, and I'm on a rickety old corporate-issue MacBook from the Bush II Era. Pure javascript over cable modems.

I'm kind of at the point where I'm beginning to believe that HTML is considered harmful. Slow, inconsistent, opaque, resistant to modern UX techniques to the point where it should hang out with the glowering trenchcoat geezers that all have dog-eared Oreilly manuals dating back to the '90s that are still referred to. With modern OCR, even searchability of plaintext is no longer a concern.

Screw HTML. Screw CSS. Direct javascript rendering is where we need to go.
posted by Slap*Happy at 7:30 PM on September 24, 2014

With modern OCR, even searchability of plaintext is no longer a concern.

That's a good one. Are you booking gigs? Because I'm going to a librarian/archivist conference...
posted by overeducated_alligator at 6:02 AM on September 25, 2014

Then you'll want to be familiar with the state of the art.
posted by Slap*Happy at 7:39 AM on September 25, 2014 [1 favorite]

the glowering trenchcoat geezers that all have dog-eared Oreilly manuals dating back to the '90s that are still referred to

No one told me I was supposed to have a trenchcoat! <glowers>
posted by Zed at 12:47 PM on September 26, 2014