Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Monday October 06 2014, @09:50AM   Printer-friendly
from the markdown-dustup dept.

Stack Overflow co-founder Jeff Atwood is leading a community-driven update to the syntax supported by Markdown, the plain text-to-HTML conversion utility developed in 2004 by John Gruber with help from the late Aaron Swartz. Markdown allows site users to compose simple constructs such as paragraphs, italics, boldface, unordered lists, and ordered lists, in their form posts, without inserting balky HTML tags; indeed, often without being aware that any conversion is being performed at all.

Gruber's perl script, Markdown.pl, was released 10 years ago under the BSD license, and has since been incorporated - and modified - by sites all over the web. Atwood became frustrated over the varying HTML output that different sites (using different forks or rewrites of the original Markdown script) produced from the same input text. Working with developers from GitHub, Reddit, and other sites, Atwood came up with a specification and test suite for a unified Markdown syntax; originally called 'Standard Markdown', Atwood changed the name to 'CommonMark' after Gruber objected.

Not everybody is thrilled by the new spec, though. The disagreements seem to have less to do with specifics of the unified syntax, than with the idea that a common syntax is being forced on web developers who were happy with what they had. The change in project name was not enough to mollify Gruber, who apparently suspects that his work is being hijacked by a consortium.

Internet scripting guru Dave Winer blogged in support of Gruber's position:

In [a Reddit] thread, a programmer laments that there is no formal grammar for Markdown. This was totally predictable. Once the format is in play, the extent to which theorists will want to debate it is almost boundless. By the time it's over, Markdown will be Lisp. Maybe that should be a corollary for Godwin's Law, except about format wars on the Internet.

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 0) by Anonymous Coward on Monday October 06 2014, @10:00AM

    by Anonymous Coward on Monday October 06 2014, @10:00AM (#102349)

    Those HTML tags are a bitch. But the real problem is of course, why not a middle road. Contrary to the choice between ordered and unordered lists, we should have a somewhat ordered list. Maybe even one that's sometimes ordered.

    • (Score: 2) by wonkey_monkey on Monday October 06 2014, @10:10AM

      by wonkey_monkey (279) on Monday October 06 2014, @10:10AM (#102351) Homepage

      Those HTML tags are a bitch.

      Really? I never have any trouble*.

      *dammit. I used <quote> to make myself look a fool, but SN understands them now! *shakes fist*

      --
      systemd is Roko's Basilisk
      • (Score: 2) by AndyTheAbsurd on Monday October 06 2014, @01:31PM

        by AndyTheAbsurd (3958) on Monday October 06 2014, @01:31PM (#102401) Journal

        <quote>*dammit. I used <quote> to make myself look a fool, but SN understands them now! *shakes fist*</quote>

        You have to understand <a href="http://en.wikipedia.org/wiki/HTML_character_entities">HTML character entities</a> if you want to make sure to make yourself look like a fool!

        --
        Please note my username before responding. You may have been trolled.
    • (Score: 2) by c0lo on Monday October 06 2014, @11:03AM

      by c0lo (156) Subscriber Badge on Monday October 06 2014, @11:03AM (#102358) Journal

      Those HTML tags are a bitch.

      As it happens, I like bitches, life included.
      So, what's wrong with HTML tags? Why should I care about CommonMarkDown?

      --
      https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
      • (Score: 3, Informative) by tonyPick on Monday October 06 2014, @11:15AM

        by tonyPick (1237) on Monday October 06 2014, @11:15AM (#102364) Homepage Journal

        Markdown is quite a lot simpler, and handles some of the more verbose oddities (the lt escape for < for example). This means it's a fair bit more readable and faster to put out things like lists, quoted blocks, etc.

        You know wikiformatting as used in wiki, trac, etc? That's pretty much a markdown-like syntax. And as per TFA there's a lot of ambiguities and inconsistencies in the way some sequences get interpreted, which means the syntax highlight mode in your editor might not match the look of the output under (say) trac.

        If Gruber had been maintaining Markdown he might have grounds for concern, but since he's apparently left it to stagnate, then I don't see much of a problem with the standardising fork.

        • (Score: 2) by tonyPick on Monday October 06 2014, @11:25AM

          by tonyPick (1237) on Monday October 06 2014, @11:25AM (#102366) Homepage Journal

          Doh.

          output under (say) trac

          Of course trac isn't Markdown, it's wikiformatting. Bah. should have said "under the particular markdown parser"... Hit submit too early on that one...

        • (Score: 2) by c0lo on Monday October 06 2014, @11:42AM

          by c0lo (156) Subscriber Badge on Monday October 06 2014, @11:42AM (#102369) Journal

          You know wikiformatting as used in wiki, trac, etc?

          To my frustration, yes. Still feeling sorry for the time wasted with them (and with 4GL-es in my naive younger years).

          --
          https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
  • (Score: 3, Informative) by forsythe on Monday October 06 2014, @10:38AM

    by forsythe (831) on Monday October 06 2014, @10:38AM (#102357)

    By the time it's over, Markdown will be Lisp. Maybe that should be a corollary for Godwin's Law, except about format wars on the Internet.

    Greenspun's tenth rule [wikipedia.org] should suffice.

    • (Score: 2) by c0lo on Monday October 06 2014, @11:05AM

      by c0lo (156) Subscriber Badge on Monday October 06 2014, @11:05AM (#102360) Journal
      --
      https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
  • (Score: 2) by PizzaRollPlinkett on Monday October 06 2014, @11:06AM

    by PizzaRollPlinkett (4512) on Monday October 06 2014, @11:06AM (#102363)

    I put a restraining bolt on my computer to get control over those balky HTML tags.

    --
    (E-mail me if you want a pizza roll!)
  • (Score: 3, Insightful) by MrGuy on Monday October 06 2014, @12:14PM

    by MrGuy (1007) on Monday October 06 2014, @12:14PM (#102377)

    Wiki syntax predates Markdown. There have always been different flavors. Most have some similarity (*bold*, _italic_, |table|cells| seem fairly common in my experience), but "wiki" has never been consistent site to site.

    It would be interesting if there were a single standard wiki markup syntax that we could learn once and use everywhere, but that's likely a fool's errand. If I have a specific type of item that's useful in my specific context that I need to make easier (or isn't supported in the "common" syntax), I'm going out on my own. Consider a syntax that doesn't support superscripts or subscripts. I'm a chemistry site, and I need subscripts. I'm not submitting a request to the markdown syntax maintainers, waiting for it to be voted up, waiting for it to get coded and released in the "official" syntax. I'm creating my own syntax and overriding the parser locally.

    I guess I don't see what the big deal is. Markdown isn't ODF - it's not a general purpose "write somewhere, read consistently everywhere" syntax in its intent. People aren't exchanging markdown docs when they used to send MS Word docs. It's a "write here, render rich formatting here" format. There isn't a strong NEED to be consistent.

    Most of the formatting most people use is consistent from wiki syntax flavor to flavor (and certainly from Markdown flavor to Markdown flavor). The complicated or unusual stuff, you look up on the "formatting guide" when you need it. You don't commit it to long-term memory, because you don't need it.

    What do USERS gain from standardization that's significant?

    • (Score: 2, Informative) by schad on Monday October 06 2014, @12:41PM

      by schad (2398) on Monday October 06 2014, @12:41PM (#102381)

      Wiki syntax predates Markdown. There have always been different flavors. Most have some similarity (*bold*, _italic_, |table|cells| seem fairly common in my experience), but "wiki" has never been consistent site to site.

      Those didn't originate with wikis. They were in use in email and Usenet back in the 90s, at least, and probably earlier. It's actually _underline_, though, or /italics/, which makes a little more sense. Probably it became _italics_ because underlining was, in my experience, used mainly for emphasis (italics being reserved for foreign words, book titles, and so on), and in HTML the emphasis tag produces italics.

      I guess I don't see what the big deal is.

      I don't, either, but I don't understand why something like Markdown exists in the modern world, either. For people who just want simplicity, there are WYSIWYG editors that bypass the need for an intermediate language. The people who want finer control can be reasonably expected to learn HTML or something functionally similar.

      • (Score: 3, Interesting) by RobotMonster on Monday October 06 2014, @12:50PM

        by RobotMonster (130) on Monday October 06 2014, @12:50PM (#102383) Journal

        I don't understand why something like Markdown exists in the modern world

        I find it really useful when documenting C++ code with Doxygen; the generated documentation looks nice, and the source-code (which you end up looking at more) looks nice too, e.g. I can have an ASCII-style bulleted list in my comment block in a normal text editor, and in the generated documentation I get a nice HTML-style bulleted list. If I had to use HTML tags in the source-code comments, they'd look horrible.

      • (Score: 3, Informative) by TheRaven on Monday October 06 2014, @02:10PM

        by TheRaven (270) on Monday October 06 2014, @02:10PM (#102419) Journal
        The point of Markdown is to have a format that can be both read in its source form and pretty-printed. It's great for things like README(.md) files, code comments, and so on. You can read the raw text, and it's just nicely formatted plain text, or you can run it through something and generate HTML (or PDF or whatever).
        --
        sudo mod me up
    • (Score: 2) by VLM on Monday October 06 2014, @01:06PM

      by VLM (445) on Monday October 06 2014, @01:06PM (#102388)

      "I have a specific type of item that's useful in my specific context"

      Punctuation on keyboards might be an exciting problem.

      Maybe start with standardizing keyboards worldwide.

      Which would imply as a meta-meta-step standardizing world wide human language first.

      What I'm getting at is you could really F with people by having an extension to Markdown that puts in math square root symbols using the "key that makes umlauts" and unfortunately my keyboard layout doesn't have umlauts.

      Some things are likely pretty standard across all languages and all language keyboards, like underscore and slash and probably backslash. Releasing a markdown that changes color to green when you use a $ might not fly all around the world.

      If you limit yourself to "the subset of all keys that all keyboards have" you might be stuck with a rather small vocabulary for markdown making the argument about how to extend it pretty boring.

      • (Score: 2) by VLM on Monday October 06 2014, @01:08PM

        by VLM (445) on Monday October 06 2014, @01:08PM (#102390)

        Embarrassing that I came up with my best idea for Markdown seconds after clicking submit...

        To make a subscript for chemistry or math, you "pound the numbers half a line down" so markdown 2.0 should do signal subscripts by using the "english pound" symbol. That couldn't possibly be a problem worldwide, could it?

    • (Score: 2) by gidds on Monday October 06 2014, @02:12PM

      by gidds (589) on Monday October 06 2014, @02:12PM (#102422)

      There isn't a strong NEED to be consistent.

      Not for you, maybe; but not everyone is in your position.

      At work I have to use three different issue tracking systems and two different wikis, none of which I have control over, each with their own conflicting conventions and formats, and I'm forever having to look up how to set (or prevent) some formatting on whichever system I'm using at the time.  A single syntax would save me many headaches!

      If Markdown were powerful and consistent — and hence reliable — then sites might be more likely to use it in favour of wiki markup, BBcode, WikiFormatting, crippled HTML subsets, or all the other styles.  (For the basics, I prefer Markdown to most of those; but consistency is more important.)

      --
      [sig redacted]
      • (Score: 3, Insightful) by MrGuy on Monday October 06 2014, @03:28PM

        by MrGuy (1007) on Monday October 06 2014, @03:28PM (#102467)

        If Markdown were powerful and consistent — and hence reliable — then sites might be more likely to use it in favour of wiki markup, BBcode, WikiFormatting, crippled HTML subsets, or all the other styles.

        To quote Jayne Cobb, I'm smelling a lot of "if" coming off this plan.

        Where's the incentive for existing apps to switch to a different wiki markup (thus breaking all their existing content)? And where's the strong incentive for new apps to use "consistent with other apps" over "what seems expedient to us"? If I want to make it easy for customers to switch from existing app A to my cool new app B, I'm going to choose the syntax most familiar to A's existing users, whether it's standard or not, unless I feel like "I value consistency!" is truly more important for my long-term success than "I value my target customers!"

        I'm not doubting that there are advantages to having consistency between apps, but that's really not what they're judged on.

        Your situation is a case in point why I doubt this effort will accomplish anything. When the apps you use were chosen, no one considered "is the wiki markup format this app uses consistent with the wiki markup format of other apps" a compelling reason to choose different tools. Until and unless USERS are beating down vendors' doors for consistency, why would you expect app vendors to prioritize this?

  • (Score: 0) by Anonymous Coward on Monday October 06 2014, @12:37PM

    by Anonymous Coward on Monday October 06 2014, @12:37PM (#102380)

    This reminds me of the history of UML (Unified Modelling Language) as recounted by Martin Fowler in the first chapter of his book UML Distilled [amazon.com][BUY IT NOW!]. By the mid-90s, there were six or eight popular methods for drawing object models as an aid for designing code in languages such as C++, Java and Smalltalk. Most of the authors travelled around the country teaching classes to help programmers become more productive by using their method. Then three of them - the 'Three Amigos' - got together and decided to create a unified standard. Not everybody saw the need for it, especially not the other five guys and gals.

    For the OO methods community, the big news at OOPSLA '94 was the Jim Rumbaugh had left General Electric to join Grady Booch at Rational Software, with the intention of merging their methods.

    The next year was full of amusements.

    Grady and Jim proclaimed that "the mothods war is over -- we won", basically declaring that they were going to achieve standardization the Microsoft way. A number of other methodologists suggested forming an Anti-Booch coalition.

  • (Score: 1) by Bob The Cowboy on Monday October 06 2014, @03:07PM

    by Bob The Cowboy (2019) on Monday October 06 2014, @03:07PM (#102453)

    Atwood can come off as an ass sometimes, but you can tell he means well. He announced well over a year ago that he wanted to work with Gruber on clarifying some of the more vague parts of the Markdown "spec". He was completely ignored, so he worked with other people as mentioned in TFS. Eventually, the project sent an email asking if it was ok to call it "Standard Markdown", and was again met with silence. Two weeks later, after buying the domain, setting up the website, and announcing the whole thing, Gruber pitches a fit and claims it's being stolen from him and demands he changes the name to something inane like "Pedantic Markdown".

    Bottom line, the "spec" was unmaintained, and the author seemed to hold an active disinterest, so it got forked by people who cared. Individual web developers are free to continue using their own little homebrewed Markdown, but if sharing Markdown documents between projects is important to you, you'll appreciate the attempt to standardize.

    • (Score: 0) by Anonymous Coward on Monday October 06 2014, @07:53PM

      by Anonymous Coward on Monday October 06 2014, @07:53PM (#102593)

      I see a different narrative:

      - there was something simple[^1], and it worked well enough for what I used it for[^2], and I used it all the time and *never* had to think about it;

      - now someone is coming along saying that it was buggy and needs a spec and needs changes.

      Also, why doesn't SN support markdown?

      [^1]: simple = not intended for writing your dissertation, just for simple formatting of simple notes

      [^2]: what MD is good for = casual, text-based things like this comment

    • (Score: 2) by Non Sequor on Tuesday October 07 2014, @01:00AM

      by Non Sequor (1005) on Tuesday October 07 2014, @01:00AM (#102825) Journal

      I haven't used Markdown but the concept seems pretty self-explanatory: it translates informal plaintext formatting conventions to formal HTML conventions. If you cared about the precise translation being used, it would have been wise to use formally specified formatting conventions to begin with.

      If sharing Markdown documents between projects is important to you, maybe you should think long and hard about the decisions that have put you in this position.

      --
      Write your congressman. Tell him he sucks.
  • (Score: 2) by bzipitidoo on Monday October 06 2014, @05:06PM

    by bzipitidoo (4388) on Monday October 06 2014, @05:06PM (#102507) Journal

    I find this subject very interesting. HTML bungled badly at being human readable. It is, barely, but it could have been much better.

    All I saw was talk of Markdown to HTML. What about the other direction? Are there programs to go from HTML to Markdown?

  • (Score: 1) by hendrikboom on Tuesday October 07 2014, @01:30AM

    by hendrikboom (1125) on Tuesday October 07 2014, @01:30AM (#102829) Homepage Journal

    And there's Asciidoc, which is good for writing books, and also works for shorter things, like web pages.

    How's Markdown for books?

    -- hendrik